ArticlePDF Available

Insights into Internet of Medical Things (IoMT): Data fusion, security issues and potential solutions

September 2023
Information Fusion 102(4):102060

September 2023
102(4):102060

DOI:10.1016/j.inffus.2023.102060

License
CC BY 4.0

Authors:

Shams Forruque Ahmed

North South University

Md. Sakib Bin Alam

Asian Institute of Technology

Shaila Afrin

Asian University for Women

Sabiha Jannat Rafa

Asian University for Women

Show all 6 authorsHide

Content uploaded by Shams Forruque Ahmed

Content may be subject to copyright.

Information Fusion 102 (2024) 102060

Available online 29 September 2023

Insights into Internet of Medical Things (IoMT): Data fusion, security issues

and potential solutions

Shams Forruque Ahmed

, Md. Sakib Bin Alam

, Shaila Afrin

, Sabiha Jannat Rafa

Nazifa Rafa

, Amir H. Gandomi

Science and Math Program, Asian University for Women, Chattogram 4000, Bangladesh

Data Science and Articial Intelligence, Asian Institute of Technology, Chang Wat Pathum Thani 12120, Thailand

Department of Geography, University of Cambridge, Downing Place, Cambridge, CB2 3EN, UK

Faculty of Engineering & Information Technology, University of Technology Sydney, Sydney, NSW, 2007, Australia

University Research and Innovation Center (EKIK), ´

Obuda University, 1034 Budapest, Hungary

ARTICLE INFO

Keywords:

IoMT

Internet of medical things

Data fusion

Smart healthcare

IoT

Internet of Things

Blockchain

ABSTRACT

The Internet of Medical Things (IoMT) has created a wide range of opportunities for knowledge exchange in

numerous industries. The opportunities include patient empowerment, healthcare collaboration, medical edu-

cation and training, remote monitoring and telemedicine, customized treatment plans, data sharing for inno-

vation, continuous medical learning, supply chain management, public health initiatives, wearable health

devices, and quality improvement initiatives. However, the adoption of IoMT faces numerous challenges

regarding interoperability, data privacy, security, regulatory, and infrastructure costs. This paper aims to address

the implications of data fusion in IoMT, as well as the associated security challenges and their potential solutions,

which are lacking in the literature. Data collected from IoMT devices has a direct impact on the accuracy of

predictions because of its quality, quantity, and relevance. With an accuracy of 99.53 % to 99.99 %, the Epilepsy

seizure detector-based Naive Bayes (ESDNB) algorithm is found to be the most effective for detecting epileptic

seizures in IoMT networks. However, the way data are stored must also undergo a major revolution, and all

phases—collection, protection, and storage—need to be improved. The standardization of architecture and se-

curity measures may improve the detection of security threats and compromises. Methods to detect malware in

cross platforms is also an avenue for future research that can effectively tackle the heterogeneity of the IoMT

systems. Cryptography and blockchain technology have shown to be promising ways to increase the security of

an IoMT-based system. The ndings of this review will assist a wide variety of stakeholders in the healthcare

ecosystem.

1. Introduction

Current healthcare practices employ manual administration and

maintenance of patient data, drug stock, case histories, billing, di-

agnoses, and medications, which can run the risk of signicant human

errors that can affect patients. The healthcare system is supported by a

network of centralized agents that freely exchange raw data [1]. Internet

of Things (IoT)-based healthcare eliminates human error by connecting

all devices that monitor vital signs to a decision support system via a

network, thereby aiding physicians in making more accurate and timely

diagnoses [2]. Thus, IoT is poised to emerge as a major technological

advancement of our time [3]. As a result of the numerous potential uses

that IoT offers, it is expected to grow signicantly and reach more than

24.1 billion devices globally in 2030, an equivalent of almost four de-

vices per person [4]. As IoT technology has been rapidly adopted by the

medical industry, the Internet of Medical Things (IoMT), which collects,

processes, and analyzes the medical data produced by an extensive

number of IoT devices, has also advanced rapidly [5]. IoMT is a network

of connected medical devices that transfer data through the cloud [6].

These devices can share and collect data using a variety of standards and

technologies since they are connected and interconnected [7]. As a

result, IoMT has decreased hospital visits and is seen as a remedy for the

shortage of medical resources [8]. Thus, IoMT can successfully increase

disease treatment efciency and accessibility, decrease errors, enhance

patients’ experiences, and minimize costs [9]. The global IoMT market is

* Corresponding authors.

E-mail addresses: shams.ahmed@auw.edu.bd, shams.f.ahmed@gmail.com (S.F. Ahmed), gandomi@uts.edu.au (A.H. Gandomi).

Contents lists available at ScienceDirect

Information Fusion

journal homepage: www.elsevier.com/locate/inffus

https://doi.org/10.1016/j.inffus.2023.102060

Received 6 June 2023; Received in revised form 6 September 2023; Accepted 28 September 2023

Information Fusion 102 (2024) 102060

estimated to expand from $72.5 billion in 2020 to $188.2 billion in

2025, with the largest compound annual growth rate anticipated in Asia

Pacic as a result of policies that support globalization [10].

The initial stage for intelligent services to utilize IoMT is to perceive

and collect data about environments and physical items. The employ-

ment of many sensors is required to raise the depth of the data fusion

outcome and to broaden the range of information received, as a single

sensing modality often proves to be insufcient [11]. However,

distributing all of the data would be inefcient in terms of network

bandwidth and device power due to the heterogeneity of the several

sources and volume of sensory data. In order to increase data quality and

facilitate decision-making, data fusion becomes an important technique

to extract crucial data from widely sensed or gathered data. Data fusion

is the study of efcient methods to transform data collected at different

times and from different locations into a unied representation that can

be used to make decisions, either by humans or machines. Data fusion

has several specic applications, including optimization of data quan-

tity, reduction of data size and dimension, and information extraction

from data [11]. It assists in removing data anomalies and defeats the

collusion of detected data from several sensors. There is abundant sen-

sitive patient information in the heterogeneous data generated by the

data fusion process, which is at risk because neither the collecting ter-

minal nor the processing center can be trusted [12].

The disclosure of sensitive data may be caused by both active and

passive attacks. Due to their careless deployment into networks, IoMT

devices are more susceptible to hackers than those in any other sector.

Cyberattacks can affect nearly 50 % of all IoMT equipment [9,10]. A

new privacy-aware framework that effectively detects attacks while still

protecting users’ privacy on multiple levels was proposed by

Al-Hawawreh & Hossain [13]. The proposed method stored and

distributed the gathered IoMT data among several cloud nodes and

encrypted the more sensitive parts. A smart data fusion module was also

introduced to efciently combine IoMT from multiple sources and

protect user privacy. The proposed framework also used a differentially

private contractive deep autoencoder. With an overall accuracy and

detection rate of 99 %, the proposed framework outperformed existing

IoMT attack detection models while also protecting the privacy of the

collected data. Systems like IoMT stand apart from others because they

can affect the lives of patients and raise privacy concerns if patients’

identities are made public. Additionally, the average cost of compro-

mising healthcare data is 50 times greater than leaking credit card in-

formation [9]. As a result, one of the key prerequisites for a robust IoMT

system success is security. IoMT systems that deal with healthcare data

should exercise diligence at all times, especially when collecting,

transmitting, and storing data.

Due to the market’s inherent vendor competition, IoMT products

need to develop quickly. As a result, non-standard devices with disparate

data transfer standards and heterogeneous communication protocols are

created, which leads to security, privacy, and authentication problems

[14,15]. The weaknesses inherent in the IoMT infrastructure are often

exploited by adversaries as a launching pad for various attacks. Due to

the critical nature of security concerns in the IoMT realm, previous re-

view studies have predominantly focused on identifying security chal-

lenges and proposing strategies to mitigate risks associated with IoMT

products. However, the exploration of the implications of data fusion

within the IoMT domain remains largely unexplored. Table 1 highlights

and compares the topics covered in recent review works on IoMT in

relation to the subject matter of this paper.

The objectives of this review were established based on the identied

research gaps and the following research questions:

(i) In the context of the IoMT, what are the most important data

fusion techniques utilized to combine data from various medical

sensors and devices?

Acronyms

AAMI advancement of medical instrumentation

ABE attribute-based encryption

ACL access control list

AI articial intelligence

BGL blood glucose level

CNN convolutional neural network

DBP diastolic blood pressure

DNN deep neural network

DoS denial-of-service

DTMC discrete-time Markov chain

ECG electrocardiogram

ECNN enhanced convolutional neural network

EEG electroencephalogram

ESD epilepsy seizure detector

ESDNB epilepsy seizure detector based Naive Bayes

FRA Fletcher reeves algorithm

GDPR general data protection regulation

GPRS general packet radio service

H2B heartbeats-2-Bits

ICD implantable cardioverter-debrillators

ICU intensive care unit

iDDS integrated drug delivery system

IFTTT if this then that

IMD implantable medical device

IoMT Internet of Medical Things

IoT Internet of Things

IWD internet of wearable device

LSTM long short-term memory

MDLSTM modied deep long short-term memory

MEmoR multimodal emotion recognition

MitM man-in-the-middle

ML machine learning

MRI magnetic resonance imaging

PCTL probabilistic computational tree logic

PD Parkinson’s disease

PET positron emission tomography

PPG photoplethysmographs

PSPH premium seizure prediction horizon

PTP phase transition predictor

PTT pulse transit temporal

RBAC role-based access control

RCC remote clinical consultation

RFID radio frequency identication

RSK randomly-generated symmetric key

SBP systolic blood pressure

SHS smart healthcare surveillance

SNR signal-to-noise ratio

SoC strength of crowd

SVM support vector machine

TLS transport layer security

UAM user activity model

VPN virtual private network

WSN wireless sensor networks

XGBoosting extreme gradient boosting

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

(ii) How does data fusion in the IoMT contribute to the development

of useful knowledge and the enhancement of healthcare

outcomes?

(iii) How is the IoMT different from conventional healthcare IT sys-

tems, and what are the most pressing security concerns?

(iv) When it comes to patient privacy and safety, what may happen if

there were a security breach in the IoMT?

(v) When it comes to protecting information in the IoMT, what role

do cryptography, authentication, and other forms of access con-

trol play?

(vi) To what extent may articial intelligence and machine learning

be used to identify and counteract security risks in the IoMT?

(vii) How are issues of data privacy and security in the IoMT currently

being addressed, and what are the regulatory and legal consid-

erations involved?

(viii) When collecting and using private medical information, what are

the potential ethical implications and issues that may arise?

Based on the identied research gaps and research questions, the

present paper investigates IoMT architecture, data fusion on IoMT, and

security issues with IoMT and their potential solutions. Notably, this

paper distinguishes itself by incorporating valuable insights into the

impact of data fusion on IoMT. It commences by introducing the most

prevalent types of IoMT currently in use, as well as an overview of the

general IoMT architecture. The implications of data fusion in the context

of IoMT are highlighted. The paper also thoroughly explores and pre-

sents potential solutions to the persisting security issues encountered in

IoMT.

The ndings of this review will assist with advances in healthcare

technology and practices, thereby beneting a variety of people as well

as communities. For instance, researchers can gain access to useful in-

formation for medical advances and learning resources, while health-

care providers may reduce errors, streamline workows, and improve

patient care. Insights provided in the study can also help policymakers

and healthcare administrators lower expenses, make valuable policies,

and better allocate resources. There is also potential for growth and

improvement in the health technology industry.

The present paper is organized as follows: Methodological ap-

proaches implemented to collect, organize, and analyze relevant data for

this review are thoroughly discussed in Section 2. Different types of

IoMT are introduced and described in Section 3. Section 4 explores and

analyzes the architectures employed in the IoMT. Data fusion in IoMT is

explored in various contexts, and the results are summarized in Section

5. The security issues that arise with IoMT and their potential solutions

are explored in Section 6. Open issues and current challenges are sum-

marized in Section 7. Section 8 outlines the potential future research in

order to minimize the identied challenges in IoMT. Finally, Section 9

provides a concluding review of the study, wherein the key points are

summarized and emphasized.

2. Methodology for collecting, organizing, and analyzing

relevant data

This review study aims to shed light on IoMT, its applications, and

the associated security issues and solutions using an integrative litera-

ture method, including comprehensive collection, careful ltering, and

thorough evaluation of relevant and high-quality papers. Searches were

conducted using databases from well-known sources like Google Scholar

and Scopus, as well as the journals of esteemed publishers like Nature,

Elsevier, De Gruyter, Taylor & Francis, Wiley, Springer, Inderscience,

IEEE, ACM, and Sage. Publications were searched using the keywords

"IoMT", "Internet of medical things", "Data fusion", "Smart healthcare",

"IoT", and "Internet of Things" to nd those most relevant to this study.

After that, the extra relevant papers were uncovered by ltering and

collecting the aforementioned publications’ bibliographies and refer-

ences. The following criteria were used to thoroughly examine the ab-

stract, introduction, and conclusion of these chosen papers, and a nal

classication was arrived at:

(i) Concentrating mostly on peer-reviewed works from the afore-

mentioned reputable websites and publishers;

(ii) Accumulating published works of working researchers in the

eld;

(iii) Including an appropriate balance of both new and older studies;

Table 1

Comparative analyses of the topics covered in this review study with some of the recently published review studies.

Study Primary objective Architecture Data

fusion

Security

issues

Solutions to

security issues

This

study

Explores the importance of data fusion in IoMT and associated security challenges and mitigation

strategies to shed light on IoMT.

√ √ √ √

[16] Identies vulnerabilities in currently available medical equipment, along with potential solutions

and regulatory measures.

√ ×√ √

[17] Reviews state-of-the-art IoT-based sensors and sensor systems for taxonomic representation,

including privacy and security issues associated with sensor data and methods to resolve them.

× × √ √

[18] Discusses the applications, technologies and architecture of IoMT, and security improvements that

occurred to tackle COVID-19.

√ ×√ √

[9] Reviews existing techniques to improve the security of IoMT systems’ data during collection,

transmission, and storage by providing a comprehensive overview of potential attacks.

√ ×√ √

[19] Discusses the benets of IoMT applications in healthcare; provides an insight into technologies that

complement IoMT as well as the difculties with establishing a "smart" healthcare system.

√ ×√ (brief) ×

[20] Highlights the architecture and use of IoMT technology in the healthcare system. √ ×√ (brief) ×

[21] Provides a categorization of security threats as well as security counter-measures. √ ×√ √

[22] Analyzes the many digital system architectures already in-use in healthcare, including their

approaches, limits, and the present challenges facing the e-health sector.

√ (brief) ×√ √

[23] Systematically reviews articles published during COVID-19 by designing a taxonomy for the

categorization of various aspects of IoMT.

√ ×√ (brief) ×

[24] Comprehensively reviews IoMT and its architecture, discusses the obstacles and potential

solutions, and suggests future guidelines for the use and implementation of IoMT.

√ ×√ √

[4] Investigates the role of machine learning-based intrusion detection systems in resolving security

and privacy concerns in IoMT infrastructures.

√ ×√ √

[25] Analyses in-depth the security concerns associated with IoMT (and IoT). √ ×√ √

[26] Describes why certain security strategies, requirements, and design obstacles are essential, and

how to overcome them.

×√ √

√ available; ×not available.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

(iv) Including references to relevant commercial sites where the

aforementioned search terms can be found, as well as references

to cutting-edge technology that are relevant to the present

research;

(v) Returning to the original sources and retrieving critical papers

that were referenced in subsequent research, reviewing the prior

literature revealed several uncertainties about this report that

required additional research to explain the issues at hand and

strengthen the overall quality of the investigation. Some key-

words were carefully maintained to follow the thought process

and investigate the necessary publications to guarantee the ne-

cessity of the logical aspect of examining the literature in this

study. When brainstorming this paper’s topic, the terms "IoMT",

"Internet of Medical Things", "Data fusion", "Smart healthcare",

"IoT", and "Internet of Things" came to the top. It is essential to

have the criteria aiming to limit and lter the scope, as it is

thought that relevant publications have much larger importance

than the literature volume. Table 2 outlines the inclusion and

exclusion criteria used to choose the papers that formed the basis

of this study.

Despite being a part of the selection and limiting process for the

literature, the criteria for exclusion seemed quite judgmental. For this

reason, the authors implemented a test-retest procedure, in which the

retrieved data are checked for accuracy a second and even third time

following a random choice from primary research. To help readers

comprehend the primary material of this study, a holistic overview

based on pre-exploratory mapping is shown in Fig. 1a. Using the existing

publication categorization system, a comprehensive outline was devel-

oped. In the end, 1372 publications were collected for analysis, but only

137 were included in this review paper, as seen in the chronological

distribution of publications in Fig. 1b.

3. Types of IoMT

IoMT systems can help cure numerous medical diseases by providing

an effective pathway for better diagnoses and treatments. By enhancing

work efciency, enabling remote monitoring and diagnostics, and

improving patient care, IoMT has the potential to completely transform

the healthcare industry. For instance, some medical issues necessitate

the use of implantable devices like pacemakers to ensure that the heart

continues to beat normally [28]. Other devices that are wearables, like

smartwatches, provide a better, noninvasive healthcare experience to

monitor different organs of the body [29]. Due to these distinctions, the

IoMT systems can be divided into two groups: implantable medical de-

vices (IMDs) and Internet of Wearable Devices (IWDs). An overview of

these types is provided in Table 3.

3.1. Implantable medical devices

Implantable medical devices are gadgets that can be inserted into the

body to detect, monitor, and treat diseases. These devices are often

implanted by surgical treatment or inserted into the body utilizing

minimally invasive techniques. They may perform specic tasks or

provide current medical help, depending on the needs of the individuals.

For instance, a pacemaker is an IMD that helps with the treatment of

irregular heartbeat by stimulating the heart to beat normally when it is

beating too quickly or too slowly [30]. Wireless implantable medical

devices are important in modern healthcare because they offer a variety

of advantages and benets. By eliminating the need for cables or con-

nections, implantable wireless devices allow patients to move around

more easily and comfortably. Therefore, low power consumption,

limited storage space, and long lifespan, compact batteries are crucial

criteria for these devices to stay within a human body for an extended

duration. For example, pacemaker implants typically survive between 5

and 15 years [9]. Wireless implanted devices can be equipped with so-

phisticated safety measures, including automatic notications and

alarms, to alert medical professionals and patients of any serious acci-

dents or irregularities. Rapid medical attention and response, therefore,

Table 2

Criteria for identifying appropriate literature [27].

Inclusion guidelines Exclusion guidelines

- Publications chosen for this research

must be scientic and peer-reviewed

and must be both relevant and

capable of answering the research

questions.

- Gray literature is also thoroughly

researched because of its potential

usefulness and importance.

- Even if they were scholarly and peer-

reviewed, publications that omitted or

inadequately discussed the information

relevant to the aforementioned key-

words would be disqualied.

- Gray literature with redundant ndings

or lacking appropriate references or

contexts will be excluded. In addition,

the current analysis only considered

conference proceedings to a limited

extent.

Fig. 1. (a) Literature selection and ltering process owchart, where N is the

number of publications; (b) yearly publication distribution since 2010.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

improve patient outcomes and safety. Wireless connectivity enables

medical professionals to remotely set and customize implantable devices

according to patients’ needs and requirements. Without required inva-

sive procedures or device replacements, individualized treatment regi-

mens and adjustments are available because of this exibility.

Implantable medical devices include pacemakers, implantable

cardioverter-debrillators (ICDs), and neurostimulators [31].

A smartphone application was developed by Tarakji et al. [32] that

allows for immediate communication between Bluetooth low-energy

pacemaker devices for cardiac resynchronization treatment and smart

gadgets. A BlueSync eld evaluation was carried out to analyze both

patient and healthcare professional input regarding the technology and

to measure the efcacy rate of programmed remote monitoring trans-

missions. It was found that patient-controlled remote monitoring

transmissions utilizing an app on their own smartphone or tablet ach-

ieved a 94.6 % success rate, which is higher than that of earlier

console-based remote monitoring systems when comparing groups by

age, gender, and device types. Patients who continued to use the same

platform after the evaluation reported an identical transmission success

rate of 92.8 %. After one year of treatment, patients reported feeling

condent in the procedure’s safety and efcacy. These results

demonstrate that improved transmission is achieved by reduced patient

workload, increased automation, targeted system activation, and simple

interfaces. In addition, the use of the app for remote monitoring would

positively affect the care of patients with cardiovascular implantable

electronic devices by enhancing remote monitoring success and

improving patient and provider experiences. However, the study was

designed to permit some exibility in maintaining a conventional clin-

ical practice, so controls for enforcing the arrangement of transmissions

and survey completions were limited. Results reveal that 57 % of re-

spondents reported having four scheduled transmissions, while 2.9 %

reported having none.

Unpredictable and often prolonged seizures cause the deaths of

about 15 % of epileptic patients. Effective seizure prediction that occurs

before the onset of the seizure alerts epileptic patients to implement the

appropriate precautions to lessen the effects of these seizures, improving

their quality of life. Banu et al. [33] created ForeSeiz, a smart,

self-cognizant, as well as assertive seizure predictor that is intended to

anticipate seizures. The primary objective of this research was to accu-

rately forecast epileptic seizures in real-time before they commence.

This architecture was properly designed considering the IoMT founda-

tion. A seizure prediction headband with a total weight of 30 g was

Table 3

Overview of the types of IoMT.

IoMT type Devices Main Task Outcome Advantages Disadvantages Ref.

Implantable

medical

devices

Pacemaker Cardiac

resynchronization

treatment; immediate

communication between

Bluetooth low-energy

pacemaker devices.

Successful transmission

percentage of 92.8 %, which is

higher than that of prior

console-based remote

monitoring systems, and a

success rate of 94.6 %.

Enhanced transmission is

accomplished through less

patient burden, higher

automaticity, selective

initiation, and user-

friendly systems.

The controls for enforcing

the arrangement of

transmissions and survey

completions were limited.

Tarakji et al.

[32]

Seizure predictor Accurately forecast

epileptic seizures in real

time before they

commence

The model produced an

accuracy of 97 % and a

precision of 96.11 %.

Effective seizure

prediction for epileptic

patients; implementing

the appropriate

precautions to lessen the

effects of these seizures.

The seizure patterns, causes,

and characteristics may vary

widely between individuals.

Banu et al.

[33]

Seizure detector

with drug delivery

unit

Real-time seizure

detection and

administration of the

medication to the desired

location

Showed 100 % sensitivity and

a latency of 1.8 s on average;

delivered lower power usage

based on simulation results.

The system allowed for a

signicant decrease in

power consumption and

an improvement in

detection accuracy.

Data security and privacy

issues were a concern for

this study, especially in the

drug delivery system.

Sayeed et al.

[34]

Nanogenerator and

cardiac pacemaker

Internal energy

conversion from

mechanical to electrical;

ventricular pumping and

monitoring function

Increases in an active area and

surface charge density

resulted in better electrical

output after a few

modications were made to

the surface.

Energy harvesting was

successfully demonstrated

by Bluetooth monitoring

of live output voltage

data.; capacity of self-

recharging a lithium-ion

battery.

Device interference and

compatibility; for example,

a pacemaker might get

affected by electromagnetic

elds of MRI.

Ryu et al.

[40]

Internet of

wearable

devices

Cardiorespiratory

tracker

Cardiac signal processing

and data transmission

enable the wireless

transfer of the collected

information

Calculating breathing rate

from heart rate using only

pulse amplitude.

Simple to use, portable,

and smaller in size than

other gadgets.

Occasionally might set off

misleading alerts identifying

unusual heart rates or

breathing patterns in terms

of no such emergency; these

false alarms may trigger

tension, anxiety, or medical

treatments.

Sasidharan

et al. [36]

Wristwatch Measuring blood pressure Surpassed the AAMI standard

for the non-automated

sphygmomanometer and

demonstrated accuracy

equivalent to oscillometry-

based devices.

CareUp performed well in

the estimation of SBP and

even superior in the

estimation of DBP

compared to other

devices; the p-values

generated by CareUp were

consistently higher.

Limited user interface;

reduced battery life; data

privacy and security

concerns.

Lazazzera

et al. [37]

Heartbeats-2-Bits

(H2B) (piezo

sensor)

Measuring heartbeat

intervals at numerous

sites of the human body

The pairing has a robust

success rate of 95.6 %.

As evident from the power

tests, H2B is exceedingly

power-efcient.

The strong cardiac signal

might be collapsed into

physical movements.

Lin et al.

[38]

Smartwatch Record rest tremors in PD

patients and assess their

clinical association

The Spearman relationship

between the mean resting

tremor scores and the tremor

intensity readings was 0.81

(p<0.001).

Has the ability to export

wireless tremor intensity

records over time in order

to obtain clinically

appropriate data.

Low amplitude noise in the

signals recorded by the

gyroscopes may have an

impact on tremor

assessment.

Roberto

et al. [39]

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

constructed by integrating front-end electronics and a Seizure Predictor

Tag. In the proposed ForeSeiz predictor, the Fletcher Reeves Algorithm

(FRA) and Phase Transition Predictor (PTP) are incorporated for opti-

mizing an Enhanced Convolutional Neural Network (ECNN) classica-

tion model as well as a PTP for predicting the outcome of immediate

seizures. Transfer learning was employed to train and evaluate the

model on electroencephalogram (EEG) recordings. The model produced

a 97 % accuracy, 96.11 % precision, and a Premium Seizure Prediction

Horizon (PSPH) of 66.52 min before the initial occurrence of seizures.

The Firebase cloud was additionally integrated into the model to record

the conditions of epileptic patients. If a seizure is anticipated to start, the

caregivers would be notied right away to conduct further intervention

measures. However, since the seizure patterns, causes, and character-

istics of individuals vary, not all seizures may be predicted.

Around 1 % of the world’s population suffers from epilepsy, which

highlights the importance of wearable or implantable seizure control

devices. For automated seizure identication and management, Sayeed

et al. [34] presented an IoMT-based integrated drug delivery system

(iDDS) that incorporates a seizure-detecting unit and a medicine dis-

tribution unit. A deep neural network (DNN) classier and statistical

feature extraction are also implemented for the real-time detection of

seizures. A piezoelectric-operated double reservoir micropump is used

to administer the medication to the desired location when the detection

process is nished. Results show that the proposed system delivers lower

power usage and signicantly decreases latency, which is necessary for

efcient seizure control. Therefore, this system can be considered a

workable instrument for realistic biomedical applications because of its

dual reservoir mechanism that increases longevity. It also improved

sensitivity while decreasing latency, making it a possible option for use

as an implanted low-latency device. However, data security and privacy

issue persist as a concern for this study, especially in the drug delivery

system that might alter the rate of ow among devices.

3.2. Internet of wearable devices

Wearable devices that monitor vital signs like heart rate have the

potential to make a positive impact on people’s health. Some common

examples include ECG monitors, blood pressure trackers, smartwatches,

and fall detection bands [35]. When a user is not moving around,

monitoring can be utilized to identify slow and rapid heartbeats. The

new watches can also detect breaks and use ECG measurements to di-

agnose conditions like an irregular pulse. For non-critical patient

monitoring, these devices are utilized extensively but cannot likely

replace IMDs in life-or-death situations due to sensor accuracy and

battery life issues.

A wearable cardiorespiratory tracking device was developed by

Sasidharan et al. [36] that can simultaneously record, evaluate, and

display four different parameters, including temperature, respiration

rate, peripheral capillary oxygen saturation, and heart rate on a mobile

phone. The suggested system contains three noncontact sensors that

each measure one of the four district metrics in sequence. Numerous

cardiovascular, neurological, and even pulmonary conditions may be

rapidly and easily identied at an early stage with the aid of continuous

monitoring of various physiological markers. The body sensor network

of this health monitoring system includes signal processing and data

transmission modules, enabling wireless Internet transfer of the

collected information. The heart rate of healthy individuals generally

ranges between 70 and 80 bpm and stays constant during resting con-

ditions. The range of obtained temperatures was also considered

appropriate because it did not go above the normal human body tem-

perature of 37◦C. As monitored by the proposed system, the breathing

rates, pulse rates, and peripheral capillary oxygen saturations of persons

with a history of heart attacks were extremely uctuated over a week or

two. Furthermore, the mobile application stores and evaluates the data

collected from individuals, which can be used to help predict the

probability of a heart attack and alert users and medical professionals.

The proposed system is also simple to use, considering that mobile

phones have signicantly higher availability and credibility than other

electronic devices. The wearable device is more portable and smaller

than gadgets that are worn in shirt pockets.

Lazazzera et al. [37] developed the CareUp wristwatch, an innova-

tive wearable device capable of measuring blood pressure in real-time

that uses a pulse oximeter on the rear and another oximeter on the

front of the watch. The capture of two photoplethysmographs (PPG) is

initiated by placing the index nger on the front of the oximeter; then

the signals are processed and cross-correlated to obtain a temporal delay

between them, dened as the pulse transit temporal (PTT). The systolic

and diastolic blood pressure (SBP and DBP, respectively) are then

computed using the heart rate information from the nger PPG and PTT

measurements. Using a sphygmomanometer, the smartwatch’s capa-

bility to measure blood pressure was successfully veried. Measure-

ments obtained by CareUp were compared to those of two existing

oscillometry-based devices, particularly Thuasne and Magnien, during

the evaluation process. The Wilcoxon rank sum analysis was employed

in the statistical analysis to compare the standard deviation and mean of

the estimate errors. The results nearly surpassed the American Associ-

ation for the Advancement of Medical Instrumentation (AAMI) standard

for non-automated sphygmomanometers and demonstrated accuracy

equivalent to oscillometry-based devices. Only DBP was found to be

within AAMI’s acceptable range, whereas SBP’s standard deviation error

was two points higher. CareUp performed well in the estimation of SBP

and superior in the estimation of DBP compared to the other two de-

vices. The p-values generated by CareUp were consistently higher than

those obtained using the other two instruments. However, the model

faces a few limitations, including a limited user interface, reduced bat-

tery life, and data privacy and security.

Heartbeats-2-Bits (H2B) was developed by Lin et al. [38] for effec-

tively associating wearable devices by creating a shared secret key

derived from the skin vibrations generated through the heartbeat. The

demand for sophisticated heartbeat monitors like the electrocardiogram

was eliminated by detecting heartbeat points efciently using affordable

and energy-efcient piezo sensors, which served as the inspiration for

this research. In fact, the trials demonstrated that piezo sensors are

capable of measuring heartbeat intervals at numerous body sites,

including the chest, waist, neck, and ankle. Since piezo vibration sensors

were not intended to be precise heartbeat monitors, it was also found

that the heartbeat interval signal they recorded had a low

Signal-to-Noise Ratio (SNR). An exponential function-based quantica-

tion technique was used to completely derive the accessible entropy

from the disruptive piezo readings in order to solve this issue. The H2B

was prototyped using well-known piezo sensors, and its performance

was assessed using data obtained from various body positions of 23

volunteers. The results demonstrated that H2B had a successful pairing

rate of 95.6 %, and its resilience against three distinct types of attacks

was also evaluated. Additionally, it was evident from the power tests

that H2B is exceedingly power-efcient. However, H2B only succeeded

under the present design when a user was performing static actions like

sitting, standing, or lying down, which could be due to the sensitivity of

piezo sensors to various body motion artifacts. As a result, the strong

cardiac signal would be collapsed by the movements. Thus, there is a

need for more sophisticated signal processing techniques to solve this

issue.

The implementation of wearable technology in Parkinson’s disease

(PD) research has drawn increasing attention. In order to follow up on

PD patients, Roberto et al. [39] proposed a study to examine the viability

and dependability of utilizing a system based on smartwatches to record

rest tremors in PD patients and to assess their clinical association. The

gyroscopes of a smartwatch were utilized to collect raw data for an

Android application. A total of 22 PD patients were sequentially enrolled

and monitored for a full year. The root mean square of the angular speed

recorded by the smartwatch at the wrist serves as the tremor intensity

metric. In total, 64 smartwatch evaluations were performed. The

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

coefcient of reliability with a resting tremor to determine the smart-

watch’s ability to quantify tremors was 0.89, with a minimum detect-

able change of approximately 59.03 %. Smartwatches may not pick up

vibrations in the proximal ngers because of their location on the wrist.

Innovations in nger devices may one day allow us to bypass this re-

striction. The research also showed that tremor assessment could be

affected by low-amplitude noise in the signals acquired by the

gyroscopes.

4. IoMT architecture

IoMT provides a secure and efcient platform for collecting, pro-

cessing, and analyzing medical data to produce useful information and

decision-making systems. The architecture of IoMT consists of four

essential layers: sensor, edge (gateway/fog), cloud, and interface

(Visualization/Action) [41,42], as shown in Fig. 2. Wearable sensors are

used to continuously monitor a patient’s health problems. The sensor

layer remotely collects real-time health-related data from patients

through wearable sensors connected to a system, such as Raspberry Pi

[41]. The gateway layer performs fuzzication and makes decisions at

the edge to generate real-time notications about a patient’s high-risk

medical conditions. The cloud layer accumulates the collected data for

storage and secure access by authorized personnel for monitoring

healthcare and is also accountable for data storage and computing. Se-

curity of accessing data is implemented at the action layer using an

approval-based method. Moreover, the action layer acts as the direct

interface between people and the ecosystem [43]. The gateway layer,

which is comprised of local servers and gateway devices, operates be-

tween the cloud and action layer. IoMT-based individual health moni-

toring systems have resolved the issue with traditional health

monitoring by allowing sensors on the body to monitor health signals

and connect to family and physicians over the Internet [44–46].

4.1. Sensor layer

The sensor layer is the foundation of the IoMT system, which collects

data from patients via a variety of sensors and then transmits it to the

gateway/cloud for further processing. This layer is made up of hard-

ware, including sensors, controllers, and actuators [47], allowing the

accurate detection of the parameters associated with health concerns

[48]. The sensor layer is divided into the data-entry and data-processing

sublayers [49]. The data process sublayer’s main responsibility is data

understanding, for which it employs a variety of signal acquisition and

medical perception devices. General packet radio service (GPRS), radio

frequency identication (RFID), graphic code, and other popular signal

acquisition techniques are available [50]. The collected data are sub-

sequently sent to the next state through the data entrance sublayer using

short-ranged data transmission techniques, including Bluetooth, Wi-Fi,

4 G, and 5 G [51,52].

Wearable sensors are gaining popularity for numerous applications,

including IoMT, because of the precise and trustworthy information they

can offer on regular human activities [53]. Wireless and wearable are

the two categories of IoMT sensors [54–56]. Wireless cameras and

smartphones/smartwatches are the two main categories of wireless

sensors [57]. Patients’ cognitive signals are captured using wireless

cameras like smart video cameras, and the IoMT subsystem is connected

to the cloud infrastructure via smartphones and smartwatches with

built-in GPS or Bluetooth radios [42]. Besides, wearable gadgets are

intelligent sensing devices that can produce data, link to other gadgets,

and be worn as fashion accessories. Gestures, temperature, heartbeat,

and other signals are all measured by vital sign-measuring devices, such

as biosensors [5]. In addition, these devices are primarily utilized by

runners and sportsmen to assess blood pressure, respiration rate, elec-

trocardiogram (ECG), and sleep pattern [58,59]. Wearable trackers, like

wristwatches and tness trackers, are wrist-worn gadgets that can detect

an individual’s bodily activity, such as movement and heartbeat.

Furthermore, different types of smart clothing use integrated sensors to

Fig. 2. IoMT architecture.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

track various medical problems and collect data that can be pooled and

analyzed to offer more thorough health information [60,61].

4.2. Gateway layer

This layer, sometimes referred to as the fog layer or edge layer,

comes after the sensor layer and provides a variety of stages, interface-

related tasks, and data transfer techniques. This layer is composed of

local servers and gateway devices [49], as well as the transmission and

service sublayers. Real-time data transfer occurs between the sensor

layer and the transmission sublayer. The service layer is used to inte-

grate diverse networks, data warehouses, and data description formats

[50]. In edge computing, data preparation is executed in the gateway

layer as opposed to the cloud layer. Edge computing offers numerous

benets [62], which may be summed up as follows:

i) The burden on the cloud layer will be lessened since the data are

already processed and can be sent from the devices to the cloud

layer.

ii) The latency is signicantly decreased since the gateway server is

close to the IoT devices.

iii) Security and privacy are maintained properly.

The fog layer has been applied in several healthcare applications.

Almas et al. [63] suggested a trust solution for smart healthcare systems

that is adaptive and dependent on the context, using the principles of fog

computing. Many experiments intended to enhance a health monitoring

system by utilizing the advanced services available at gateways through

fog computation, e.g., distributed data storing, data processing, and

notication, which are located at the network’s edge. In order to

accomplish speed and latency, Farahani et al. [64] offered a

patient-focused e-healthcare framework based on cloud and fog layers.

4.3. Cloud layer

The cloud layer has the data storage and computing capabilities

required to evaluate the data and develop decision-making applications

based on the evaluations [42]. The cloud resources of this layer will

store the data generated by the medical devices, allowing for further

analytical processing as required. This layer carries out machine

learning activities, data warehousing, epidemiological, and statistical

medical research. It gives a graphical interface as a nishing touch for

feedback and visualization. This cloud architecture allows service pro-

viders and caregivers access to epidemics, illness patterns, medical

histories, and remote healthcare monitoring [16]. Services like chats,

data processing, data storage, and analysis are provided for IoMT sys-

tems by cloud platforms, such as Cloud, Microsoft Azure, Amazon Web

Services, IBM Cloud, and Google Cloud [65].

Large medical and healthcare systems may easily integrate into the

cloud to conduct their daily operations. Noor [62] proposed a technique

to identify epileptic seizures by creating a cloud layer to store patient

data together with daily updated EEG samples. The required data are

subsequently processed and sent to the approved hospitals. A heuristic

method, which was introduced by Fouad et al. [66], was used to

investigate vertebral tumors. The hospital used a spinal sensor and the

transformation software hosted in the cloud to evaluate the acquired

data and provide conclusions. Panja et al. [41] utilized a mobile appli-

cation to retrieve and store the medical information of patients in the

cloud. Since acquired medical data are extremely sensitive, a security

system that assures the authentication of users was designed to restrict

access only to the authorized group of individuals.

4.4. Action layer

The management of medical records is handled by the action/

application/visualization layer in IoMT using a variety of applications,

including apps, ambient sensors, remote diagnosis, and several non-

wearable healthcare devices [47]. The action layer is divided into two

sublayers, namely medical information and medical decision-making

layers, which are responsible for handling medical information and

making decisions, respectively [49]. To maintain patient information,

the medical information layer includes a variety of medical tools and

information-related materials, tracking systems, remote diagnosis sys-

tems, telemedicine, medical e-records, etc. This layer also includes re-

sources linked to medical diagnosis and treatment [47]. Exploration of

different information, such as patients, diseases, medications, diagnoses,

and treatments, is the focus of the medical data decision-making action

layer.

The key roles of the action layer are the interpretation of data and the

delivery of application-specic services. To make diagnoses and treat-

ment plans, multiple AI methods, especially deep learning, are employed

to analyze the collected data and draw conclusions [19]. Numerous

scientic applications are also used in the application layer, including

the development of drug activity, gene mutation, diabetes monitoring,

heart arrhythmia, and Alzheimer’s detection [19]. Overall, the archi-

tecture of IoMT systems may differ based on the implementation, the

technological choices made, and the healthcare context. The main layers

that have been explained above give a general framework that can be

used to comprehend the primary components that are involved in the

construction of IoMT solutions.

5. Data fusion applications in IoMT

Data fusion techniques are crucial in healthcare as they permit the

combination of various data sources to improve clinical practice, med-

ical research, and healthcare decision-making. Data fusion is essential in

the IoMT, where many connected devices and data sources revolutionize

healthcare. Sensor data fusion is used in IoMT to combine sensor inputs

for a complete patient health prole. Feature-level fusion improves

patient diagnostics by combining relevant features from multiple sour-

ces [67]. Decision-level fusion integrates device and algorithm results to

make healthcare decisions holistically. Training models with ML and AI

to interpret complex data relationships is used in medical imaging [68].

Temporal and context-aware fusion is essential for disease progression

monitoring and environmental health effects. IoMT data fusion is

enhanced by ensemble methods, semantic data integration, quality

assurance, and privacy-preserving techniques, ensuring data privacy,

accuracy, and actionable insights for better patient care and healthcare

system efciency.

Data fusion is used in medical imaging to improve diagnosis and

treatment planning for complex conditions by integrating information

from several modalities, including positron magnetic resonance imaging

(MRI) and emission tomography (PET) computed tomography [69].

Genomic and clinical data fusion integrates genetic and health records to

improve diagnosis, treatment, and prognosis [70]. Sensor fusion in

remote monitoring also integrates information from mobile apps and

wearables for more effective management of chronic conditions and

earlier initiation of treatment. These examples demonstrate how data

fusion can improve healthcare by providing a more complete picture of

patient information and medical research.

IoMT extensively relies on data fusion since it enables the integration

and analysis of diverse types of data from multiple sources. The term

"IoMT" refers to the grouping of internet-connected medical tools and

software that enable data and information sharing. With the IoMT, more

data are becoming available from several sources, including clinical

systems, wearables, electronic health records, and medical devices. To

better diagnose, treat, and manage diseases, data fusion includes

combining these many data sources to produce a more complete picture

of a patient’s health status [5]. By offering intelligent services tailored to

gathering and processing data produced by IoT, the cloud IoT, where IoT

and cloud mix, has emerged as an enabler to satisfy data fusion qualities

[71]. However, data fusion has many uses in IoMT, some of which

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

include: remote surgery, epilepsy seizure detection, digital biomarker,

teledentistry, Alzheimer’s detection, and heart disease (Fig. 3).

Distributed computing paradigms like mobile-edge computing,

transparent computing, and fog computing are replacing centralized

computing paradigms in computing models today [72]. By deploying

processing resources at the edge nearer to data sources, edge computing

and fog computing in this context bring additional features to the cloud

and narrow the distance of endpoint IoT gadgets from the cloud

[73–75]. The IoT and AI industries are collaborating, and different

companies have already included AI in IoT applications. The IoT’s full

potential could be realized by combining it with AI, which involves

machine learning and big data analysis [76].

5.1. Remote surgery

Remote surgery is a cutting-edge surgical technique that connects

patients and surgeons who are geographically separated using robotic

tools and networking technology. Telesurgery has emerged as a viable

alternative for patients in need of urgent and high-quality surgical care,

as well as those with a shortage of surgeons and practical restrictions on

physician schedules, due to its ability to overcome the limitations of

traditional surgery. Yang et al. [77] investigated the use of enhanced 5 G

robot-supported laparoscopic surgery in urology. Data from 30 patients

who had robot-assisted laparoscopic telesurgery utilizing 5 G technol-

ogies were retrospectively examined in this study. According to the

research, enhanced robot-assisted laparoscopic telesurgery using 5 G

technologies is a practical, secure, and successful method for urological

treatments, which can reduce operating room time and blood loss and

increase surgical accuracy and precision. The limited sample size and

retrospective design of this study, however, are some of its drawbacks.

An intensive care unit (ICU) teleultrasound diagnostic system was

proposed by Duan et al. [78]. On the patient’s end, there is an ultrasound

instrument that is aided by a robotic arm called MGIUSR3, and on the

doctor’s end, there is a control unit. Both the surgeon’s and patient’s end

include audio, video, and mainframe systems. The kidneys, gallbladders,

spleens, livers, and pancreas of 32 patients were imaged using ultra-

sound technology. Because of intestinal gases and poor image quality,

one subject was eliminated from the study. The experiment was suc-

cessful for the remaining patients, and high-quality images were

obtained.

5.2. Epilepsy seizure detection

The prevalence of impulsive seizures—a prominent paroxysmal

neurological disorder that is infrequently observed in medicine—is often

recognized as epilepsy. In the IoMT, automatic epileptic seizure identi-

cation from EEG signals is considered an efcient diagnosis. The EEG

signals of various patients are gathered from different locations to a

central server to create a reliable detection system in the IoMT archi-

tecture. Ding et al. [79] introduced Fed-ESD, a privacy-preserving

federated learning architecture that uses fog nodes to facilitate the ex-

change of location-oriented EEG data for the automatic detection of

epileptic seizures. The proposed architecture employs a spatiotemporal

transformer architecture to gain spatial and temporal presentations from

each participant’s data. The study demonstrated that the proposed

framework for deployment in IoMT was effective in terms of scalability,

detection, and resource efciency. However, the study lacks the inter-

pretability of epileptic seizure detection decisions. To produce visual or

verbal justications for the detection decisions, explainable AI may be a

promising path.

Idress et al. [80] proposed a method to detect epileptic seizures when

EEG data is being compressed without loss. The method has three pur-

poses: (i) reduces the amount of EEG data transferred to the cloud; (ii)

detects patients who are having epileptic seizures at the fog gateway

using the Naive Bayes algorithm; and (iii) compresses data by merging

k-means clustering and Huffman encoding from the edge to the fog

gateway. The compression capacity was four times that of previous

methods, and the accuracy ranged from 99.53 to 99.99 %. Latency rates

were reduced from 84.6 % to 88.2 % using the KCHE compared to the

non-compressed EEG data method for several different types of EEG data

recordings. This will enable the medical team to decide quickly on their

patients.

In order to gather information on heart rate, body temperature,

muscular spasms, and falls, Hassan et al. [81] developed a monitoring

system for Grand Mal Epilepsy Tonic-Clonic seizures employing a vari-

ety of sensors, including ECG, EMG, accelerometer, and Dallas sensor. In

this system, data are categorized into several seizure kinds using a fuzzy

logic algorithm, and the classications are shown graphically on an IoT

dashboard. Using the "If This Then That" (IFTTT) technology, abnormal

situations are recognized, and an SMS message is delivered to medical

staff. For body temperature, observing heart rate, muscular spasm, and

fall identication, the system showed average accuracies of 98.90 %,

95.49 %, 83.0 %, and 87.21 %, respectively. Clustering analysis might

be utilized to better categorize and tailor a patient’s symptoms.

5.3. Digital biomarker

Finding particular data signatures, also known as digital biomarkers,

that may be utilized for categorization or estimating the severity of the

underlying conditions is one of the important components of wearable

devices with machine learning (ML) techniques. Ahmed et al. [82]

described how non-invasive wearable devices may be used to measure

blood glucose levels (BGL) in diabetic patients using AI-based method-

ologies. Glycemic events may be tracked using digital biomarkers, which

is a signicant advancement in self-monitoring technology. The study

examined the effectiveness of AI techniques in calculating BGL among

diabetes patients utilizing non-invasive wearable device data. High ac-

curacy was achieved when estimating the link between glycemic mea-

sures and characteristics, but the study’s limited sample size of only 13

people was a major limitation.

Identifying human emotions in real-time is possible by utilizing data

from the Emotional IoT and a deep learning model called MEmoR, as

suggested by Kumar et al. [83]. MEmoR employs visual and psycho-

physiological data as its two data modalities. Video signals are used to

record the visual information, and a ResNet50 model that had already

been trained is adjusted for emotion categorization. Using a convolu-

tional neural network (CNN), the psychophysiological data are

Fig. 3. Applications of data fusion in IoMT.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

analyzed. The results from both approaches are integrated by utilizing

decision-level weighted fusion. MEmoR achieved accuracies of 83.79

and 81.54 % on the Bio Vid Emo DB multimodal dataset. However, the

study did not report how individuals might behave in a changing envi-

ronment for real-time emotion detection model creation.

A novel approach for diagnosing depression that combines the LSTM

and SVM algorithms was presented by Arora et al. [84]. Statistical

characteristics are integrated with the features extracted from activity

measures using the LSTM. The merged feature map was implemented to

train the SVM model, and the Depression dataset was used to test its

accuracy, which was 95.57 %. Depression was reportedly assessed using

overlapping sliding windows on recordings of motor activity and a mix

of statistics and deep learning-derived variables. Future research can

include transfer learning models and other behavioral health facets to

suggest a paradigm for behavioral health.

5.4. Teledentistry

Comprehensive oral healthcare requires the utilization of patient-

centered care, but there are obstacles, such as high treatment costs, an

aging population, chronic dental disorders, and difculties reaching

patients in distant places [85]. Telemedicine has been developed as a

means of lowering the frequency of dental visits, enabling at-home

self-care, and assisting in the identication of various oral disorders.

Additionally, telemedicine may bridge the gap in dental treatment dis-

crepancies between urban and rural locations and can monitor patients’

health problems.

In order to prevent and identify dental caries in its early stages,

Salagare and Prasad [86] presented a creative approach that is built on

the IoMT and teledentistry. This technologically-based model offers in-

formation on how to manage the oral cavity’s caries-causing variables,

such as biolm pH, biolm presence, and oral cavity temperature, using

intraoral sensors attached to appliances. This information is continu-

ously gathered from a patient’s mouth cavity and sent to a server via a

mobile app. The Internet of Dental Things, articial intelligence, and

telemedicine are used to analyze data. The method is easy to apply at the

community level.

In a study by Martin et al. [87], remote clinical consultations (RCCs)

were compared to in-person consultations to determine which is more

effective for restorative dentistry. A verication consultation took place

in-person after 23 patients had RCC performed remotely using

high-speed internet and audiovisual transmission. Results demonstrated

that RCC was as efcient and secure as in-person consultations,

regardless of the user’s gender or age. The research intervention team

and patients’ responses to a theme questionnaire revealed that the GDP,

nurse, and patient all made successful contributions to the RCC process.

This study, however, did not evaluate the intervention’s

cost-effectiveness, efcacy under all clinical conditions, or GDPs’

acceptability in a real primary care practice.

5.5. Alzheimer’s detection

Patients with Alzheimer’s disease are typically monitored using

wireless sensor networks (WSNs). These patients may experience sub-

stantial memory loss, and the independence and well-being of the pa-

tients may be impacted by this cognitive impairment. WSN may gather

patient activities, for instance, their postures and states, to assist health

studies as the fundamental framework for future healthcare systems. To

automate the detection of early-stage Alzheimer’s disease and early

identication of cognitive damage, Yin et al. [88] proposed an IoT ar-

chitecture employing an eye-tracker (ET) and cloud-based diagnostics

made possible by ML. The method uses multimodal information derived

from several oculomotor types for better accuracy of output classes, and

the custom eye-tracker nodes gather 3D oculomotor responses from a

variety of stereo video stimulation sessions. The suggested technique

showed an accuracy of 86 % for identifying Alzheimer’s patients. To

assess the learned models and enhance self-optimization of the training

methods, intelligent model evaluation techniques could be included in

future.

In order to help patients in regaining their condence, Gao et al. [89]

suggested an approach to identify anomalous behaviors associated with

moderate cognitive impairment and forecast patient actions. In this

approach, the home environment’s architecture is formalized as an ab-

stract grid. A user activity model (UAM) was built using a discrete-time

Markov chain (DTMC) based on sensor data, in which everyday be-

haviors are veried using probabilistic computational tree logic (PCTL).

A probabilistic model verication tool uses the UAM as input to compute

probability values and evaluate temporal behavior. By identifying ac-

tivities with unusual temporal characteristics or unexpected probability,

the approach may also be used to diagnose problems. However, the user

behavior model in this research was solely generated using DTMC and is

unable to consider variations in these tasks over time and interruptions

of some activities brought on by the forgetfulness of patients.

Three sensors and a Raspberry Pi were used in a smart cabinet to

track the frequency of doors opened, providing a measure of the user’s

memory [90]. After putting the smart cupboard to the test in a

controlled environment with 23 participants, a signicant correlation

was found between the results of the test and memory testing proced-

ures. Validated face-name association memory test ndings and a

self-reported test of perceived memory were used to evaluate the ac-

curacy of the memory tests. The proposed smart cabinet performed well

in the memory test. The fact that the user is assumed to be the system’s

sole consumer is a weakness of the existing approach. Therefore, a

system of identication that can locate a person is necessary.

5.6. Heart disease

Several heart monitoring apps have been used to test the IoMT sys-

tem’s data fusion capabilities. An electrocardiogram (ECG) can aid in

the diagnosis of heart diseases, such as arrhythmias and coronary artery

disease. For remote ECG monitoring, an IoT-authorized, cloud-oriented

system was proposed by Raheja and Manocha [91]. To remove noise and

identify ECG characteristic spots, the ECG signals are pre-processed

using Savitzky-Golay and maximum overlap discrete wavelet packet

transform. A triple data encryption standard is employed for encryption

and authentication, and the CNN algorithm was developed for classi-

cation tasks. Cardiologists were given access to encrypted and autho-

rized ECG data through the ThingSpeak platform for analysis. The

proposed CNN model showed an average accuracy of 99.12 % when

classifying heartbeats into ve distinct forms of arrhythmias.

The IoMT in heart disease screening systems enables individuals to

conduct self-examinations to evaluate the presence of irregularities in

their hearts, thereby facilitating early detection of heart disease. As

demonstrated by Su et al. [92], the screening system for valvular heart

disease effectively examines and evaluates the distinctive signal values

of individuals diagnosed with this condition. To enhance the prediction

of patients’ heart disease, Pan et al. [93] suggested an Enhanced Deep

Learning assisted CNN (EDCNN) model. The EDCNN’s hyperparameters

were tuned after being designed to be as adaptable as possible, allowing

for an accuracy of up to 99.1 %.

To develop a deep learning-based method for predicting heart dis-

eases, Rajkumar et al. [94] focused on a publically available dataset on

heart ailment in Hungary collected from Internet of Things (IoT) sensor

devices. Using a modied deep long short-term memory (MDLSTM)

method, the data are preprocessed, feature-selected, and categorized.

The output is then modied by an improved spotted-hyena optimization

technique. Python, where this method was implemented, has been

shown to be accurate to within 0.01 % across a range of metrics. Further

investigation into prenatal health and family history may aid in the

detection of heart disease.

Lalitha & Jinny [95] proposed a hybrid architectural framework and

ML model for predicting heart disease in medical settings based on

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

IoMT. Medical records were collected by utilizing IoMT devices in the

proposed architecture. Categorical information was encoded with a

one-hot mechanism and normalized with the help of a tried-and-true

scalar technique during the preprocessing stage. Additionally, a hybrid

sequence of Filter, Ensemble, and Wrapper approach was used to select

the most important features. Logistic Regression and the Extreme

Gradient Boosting (XGBoosting) algorithm both produced highly precise

predictions, with an accuracy of 86.47 % and 85.81 %, respectively. The

proposed model demonstrated a higher level of accuracy (80.53 %)

when utilizing the decision tree approach in comparison to the existing

approach (73.22 %).

A recent study by Basak and Chatterjee [96] discusses the develop-

ment of a smart healthcare surveillance (SHS) system for heart diseases

that integrates IoT and ML technologies to monitor and analyze vital

signs data. Using ML techniques like Random Forest, Gaussian NB, Lo-

gistic Regression, Decision Tree, KNN, and SVM, the study predicted

heart disease across all three layers of the proposed architecture, with

Random Forest obtaining the greatest accuracy of 92.4 %. Deep learning

models could be implemented to ensure the reliability of the method.

Integrating IoT, fog, and cloud, the FRIEND system was introduced

by Pati et al. [97] to provide real-time remote diagnostics for heart

patients. The system incorporated fog computing concepts and was

found to be both user- and energy-friendly. The system was further

evaluated using several machine learning algorithms and ensemble ap-

proaches and trained on a composite dataset comprised of several heart

disease datasets, yielding a high accuracy of 94.27 % and other positive

performance metrics. Despite its usefulness, the study has some draw-

backs, such as a high price tag, limited data, and reliance on just one

platform. The surveyed studies that were conducted on applications of

data fusion in IoMT are summarized in Table 4.

In general, IoMT applications for data fusion make it possible for

healthcare providers to employ large amounts of health data provided

by linked devices and systems. Integrating and analyzing this data al-

lows medical personnel to generate insights that can be put into action,

which in turn improves patient monitoring, enables predictive analytics,

and improves decision-making, all of which contribute to better patient

outcomes.

6. Security issues in IoMT and potential solution

The healthcare sector has been transformed by IoMT through the

interconnection of medical devices and systems to improve patient care

and outcomes. However, this connectivity also raises security issues that

must be addressed in order to protect individual patients’ information,

medical equipment, and the security of healthcare networks as a whole.

One of the primary hurdles is the vulnerability of IoMT devices to cyber

attacks, including unauthorized access, data breaches, and malicious

tampering. Security of protocols, bioinformatics, and health data has

become of the utmost importance due to the potential severity of an

attack on the healthcare system, particularly the loss of control over life-

supporting equipment [98]. The key security issues in IoMT include

authentication, authorization, data condentiality, availability, and

integrity, as illustrated in Fig. 4.

Various approaches have been investigated to protect the healthcare

system from being exploited through vulnerabilities, such as utilizing a

blockchain-based environment for the IoMT and edge cloud technology

[19,99]. Other potential solutions involve intense encryption imple-

mentation and authentication methods, regular software updates and

patches, thorough risk assessments, training healthcare professionals on

cybersecurity best practices, and establishing comprehensive regulatory

frameworks to ensure adherence and responsibility [100–102].

Furthermore, fostering collaboration among manufacturers, healthcare

providers, and cybersecurity experts is essential for resilient and secure

IoMT system development that prioritizes the privacy of patients.

Table 4

Surveyed studies on the applications of data fusion in IoMT.

Application Objective Outcome Remarks Ref.

Remote

Surgery

To investigate

the use of

enhanced 5 G

robot-

supported

laparoscopic

surgery in

urology.

Enhanced robot-

assisted

laparoscopic

telesurgery using

5 G technologies

was found to be a

practical, secure,

and successful

method for

urological

treatments.

The retrospective

nature of the

study and the

relatively small

sample size are

two issues that

need to be

addressed.

[77]

To propose an

intensive care

unit (ICU)

teleultrasound

diagnostic

system.

Kidneys,

gallbladders,

spleens, livers,

and pancreas

were imaged

using ultrasound

technology

performed on 32

patients. Success

with the

experiment was

achieved.

The image

quality was quite

high-quality.

[78]

Epilepsy

seizure

detection

Automated

identication of

epileptic

seizures in the

IoMT using EEG

data.

The efcacy of

the suggested

framework for

detection,

efciency of

resources, and

scalability was

demonstrated,

justifying its

implementation

in the IoMT.

Decisions made

during the

identication of

epileptic

episodes are not

easily

understood.

Possible benets

of explainable AI

include the

ability to

generate visual

or verbal

explanations for

detection

decisions.

[79]

To identify

epileptic

seizures while

EEG data is

compressed

without any

loss.

The compression

power was four

times that of

methods, and the

accuracy was

between 99.53 %

and 99.99 %.

When comparing

the KCHE to a

non-compressed

EEG data

method, the

latency rate is

reduced by

84.6–88.2 % for

different types of

EEG data

recordings.

[80]

To propose a

prototype

monitoring

system for

Grand Mal

Epilepsy Tonic-

Clonic (GTC)

seizures.

The system had

an average

accuracy of 83.0

% for detecting

muscle spasms,

98.1 % for

monitoring core

body

temperature,

95.4 % for

monitoring heart

rate, and 87.2 %

for detecting

falls.

Clustering

analysis might be

utilized to better

categorize and

tailor a patient’s

symptoms.

[81]

Digital

Biomarker

Analyze how

non-invasive

wearable

devices may be

used to measure

blood glucose

levels (BGL) in

diabetics.

High accuracy

was achieved

when estimating

the correlation

between

glycemic

measures and

characteristics.

Considering a

small number of

individuals may

not be reective

of the system’s

actual

performance as a

whole.

[82]

To identify

human

MEmoR achieved

accuracies of

A lack of

information on

[83]

(continued on next page)

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

6.1. Security issues in IoMT

IoMT is susceptible to numerous vulnerabilities due to resource

limitations in the devices, their diversity, and the sheer number of IoMT

users. IoMT security needs fall into three main categories: function se-

curity, access control security, and information security. These security

requirements are interconnected and reciprocally inuenced by one

another [103].

6.1.1. Authentication

Authentication challenges in IoMT refer to the difculties in veri-

fying the identity of devices, users, and data within the healthcare

ecosystem. Given the complexity and diverse perspectives of IoMT sys-

tems, proposing universal authentication solutions for different nodes

within these systems is challenging. Consequently, IoMT authentication

primarily focuses on three levels, namely the device, network, and user

levels [104]. Users at the application layer, such as patients and

healthcare providers, are the primary emphasis of user-level authenti-

cation, whereas devices within the IoMT system are the main focus of

device-level authentication [105–108]. Besides, network-level authen-

tication involves registering and authenticating users and devices to

ensure the inclusive security of the IoMT [109,110].

The main reason for using authentication techniques is to restrict

IoMT’s resources, features, facilities, and services to only those who are

authorized to use them. As a result, it is crucial to evaluate authenti-

cation solutions’ robustness in the face of potential attacks aimed at

gaining unauthorized access to IoMT systems. Common IoMT authen-

tication attacks include forgery attacks, smart card theft, insider attacks,

tracking attacks, sensor attacks, Denial-of-Service (DoS) attacks, session

key information breaches, Man-in-the-Middle (MitM) attacks, and dy-

synchronization, among others [104]. ECG-based authentication has

gained substantial attention in smart healthcare systems due to its

unique attributes, such as being inimitable, suitable, accessible, and

comfortable for users. However, enhancing authentication accuracy,

particularly in scenarios with massive users, poses a considerable

Table 4 (continued )

Application Objective Outcome Remarks Ref.

emotions in

real-time by

utilizing data

from the

Emotional IoT

and a deep

learning model

called MEmoR.

83.79 and 81.54

% in the

prediction of

valence-arousal

and discrete

emotion,

respectively.

individual

behavior in

changing

environments

prevents the

development of

real-time

emotion

detection

models.

To diagnose

depression

using LSTM and

SVM

algorithms.

Depression was

assessed by

overlapping

sliding windows

on recordings of

motor activity

and a mix of

statistics and

deep learning-

derived

variables.

The eld of

behavioral health

can benet from

further study that

incorporates

transfer learning

models and other

aspects.

[84]

Teledentistry To prevent and

identify dental

caries in its

early stages.

Cariogenic

parameters, such

as biolm pH,

biolm presence,

and oral cavity

temperature, can

be easily

monitored with

intraoral sensors

that can be tted

to a variety of

dental products.

The method is

easy to apply at a

community level.

[86]

To compare

remote clinical

consultations

(RCC) versus in-

person

consultations to

determine

which was more

effective for

restorative

dentistry.

RCC was as

efcient and

secure as an in-

person

consultation,

regardless of

gender or age.

The

intervention’s

cost-

effectiveness,

efcacy under all

clinical

circumstances,

and GDPs’

acceptability in a

real primary care

practice should

be evaluated.

[87]

Alzheimer’s

detection

To detect

Alzheimer’s by

employing eye-

tracker and

cloud-based

diagnostics

made possible

by ML.

The proposed

model

demonstrated an

accuracy of 86 %

for identifying

Alzheimer’s

patients.

To assess the

learned models

and enhance self-

optimization of

the training

methods,

intelligent model

evaluation

techniques could

be included in

future.

[88]

To identify

anomalous

behaviors

associated with

moderate

cognitive

impairment and

forecast patient

actions.

By identifying

activities with

unusual temporal

characteristics or

unexpected

probability, the

approach may be

used to diagnose

problems

associated with

Alzheimer’s

disease.

This model is

unable to

account for

variations in

several behavior

tasks over time

and interruptions

in some activities

caused by the

forgetfulness of

patients.

[89]

To measure a

user’s memory.

The proposed

smart cabinet

performed well

in a memory test.

The current

implementation

of the system

implies there is

only one user,

which is one of

the limitations.

[90]

Table 4 (continued )

Application Objective Outcome Remarks Ref.

Heart disease To propose an

IoT-authorized,

cloud-oriented

system for

monitoring ECG

remotely.

The proposed

CNN model had

an average

accuracy of

99.12 % when

classifying

heartbeats into

ve distinct

forms of

arrhythmias.

More categories

of arrhythmias

could be

analyzed in the

future.

[91]

To provide a

heart disease

prediction

system based on

deep learning.

The proposed

model showed

98.01 % accuracy

in predicting

heart disease.

The research can

be expanded to

examine early

hereditary and

health-related

characteristics

for the early

diagnosis of heart

disease.

[94]

To evaluate

vital signs data

in order to

develop an SHS

system for heart

disease.

The model used

ML algorithms

and achieved

92.4 % accuracy.

Deep learning

models could be

implemented to

verify the

reliability of the

method.

[96]

To provide real-

time remote

diagnostics of

heart patients.

The system was

found to be both

user- and energy-

friendly by

incorporating fog

computing

concepts.

Several issues

need addressing,

including high

prices, limited

data collection,

and reliance on a

single platform.

[97]

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

challenge. A parallel ECG-based authentication method called PEA was

introduced by Zhang et al. [111] to address this issue.

In addition to PEA, many models and strategies have been presented

to deal with authentication attacks in this domain. For instance, Fotouhi

et al. [112] devised a method that combines short-term and long-term

parameters to safely generate and transfer a communication session

key. These advancements aim to upgrade the security and effectiveness

of ECG-based authentication systems for smart healthcare applications.

Additionally, Hajian et al. [113] proposed an authentication strategy

that effectively defends against guessing and insider attacks. In this

scheme, the adversary is required to correctly guess three double-hashed

values to pass the authentication process, making guessing attacks

virtually impossible. Furthermore, Das [114] offered a method for

resilience against MitM attacks that employs a temporal credential for

secure communication between entities, preventing an adversary from

exploiting the communication even if they possess other private cre-

dentials. In addition, the verier in the proposed architecture is not

predicated on the credentials provided during authentication.

Kumar and Tripathi [115] developed and improved a system that

affects a transaction-based blockchain. By including time and identier

stamps on veried transactions, this scheme signicantly reduces the

likelihood of successful replay attempts. Sureshkumar et al. [116]

introduced a method that effectively thwarts DoS attacks, making them

difcult to execute. The scheme is based on request-response commu-

nication, in which the user must rst receive a time-stamped approval

from the sensor before a connection can be established. Bhuarya et al.

[117], Das et al. [118], Tahir et al. [119], and Wu et al. [120] explored

additional strategies and techniques to tackle the authentication and

security issues in smart healthcare systems.

6.1.2. Authorization

During attacks, there is a risk of unauthorized access or data alter-

ation to sensitive data that can result in the loss or unavailability of data

for authorized users and customers. Authorization plays a central role in

ensuring that solely authorized entities can access particular network

resources, such as medical IoT devices and patient medical information.

For example, committed entities are entitled to execute actions similar

to giving commands to IoMT or upgrading device software [121]. Ad-

versaries may exploit weak authorization techniques in an IoMT

network to gain resources without proper rights. As a result of users’

insufcient security training, knowledge, and awareness, social engi-

neering attacks can exploit vulnerabilities in IoMT devices, allowing

malicious actors to impersonate legitimate users and gain access to their

medical devices. Attackers can use compromised IoMT devices as bots to

initiate more attacks over the IoMT edge network and gain access to

network services, such as remote control of numerous IoMT devices and

important resources like patient medical data [122].

The consequences of inadequate authorization practices in IoMT can

have critical implications, including data breaches, identity theft, and

unauthorized use of sensitive medical information. Implementing robust

authorization practices is essential to protect patient data and maintain

the trust of individuals and organizations relying on IoMT systems. In

the case of medical devices that monitor vital signs, compromised

authorization can even harm a patient’s life. Access control is a widely

used security strategy that conrms proper authorization. An instance of

an access control mechanism is an access control list (ACL), which is

based on the concept of discretionary access control models using an

access matrix. ACLs determine the specic operations that a veried

user is authorized to access, retrieve, and execute [123]. These mecha-

nisms provide granular control over access rights, enabling adminis-

trators to dene and enforce permissions based on individual users or

groups.

The centralized authorization server in traditional access control

systems can pose challenges as being a single point of failure or a per-

formance bottleneck. Hence, Xu et al. [124] developed a decentralized

mechanism called BlendCAC using blockchain technology, which uti-

lizes token management for various actions, enabling decentralized

authorization decisions. This approach aligns with access control models

commonly used in distributed architectures, where the access control

logic is distributed and embedded within end devices rather than relying

on a central authority [125]. The proposed BlendCAC scheme introduces

a robust identity-based capability token management strategy that le-

verages smart contracts for propagation, revocation, and registration of

access authorization. Instead of depending on oversight from a central

authority, BlendCAC allows IoMT devices to manage their resources

autonomously. To assess the effectiveness of the strategy, BlendCAC was

deployed to Raspberry Pi and evaluated on a private blockchain network

in the area. The results showcase the viability and potential of BlendCAC

to provide a scalable, lightweight, decentralized, and ne-grained access

control solution for IoMT [124]. By leveraging blockchain technology

and decentralizing the authorization process, BlendCAC offers the po-

tential to enhance scalability, resilience, and performance in access

control systems.

Federated identity management facilitates the secure federation of

identities across various healthcare organizations, thereby reducing the

need for individual management of user identities. This can alleviate

Fig. 4. Key security issues in IoMT.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

administrative burdens and ensure the security of identity-related pro-

cesses within IoMT. Moreover, adopting safe protocols, such as Trans-

port Layer Security (TLS), guarantees the establishment of secure

communication channels between IoMT devices and systems. TLS safe-

guards data integrity and condentiality, effectively preventing any

unauthorized interception of data during transmission. By incorporating

federated identity management and robust protocols like TLS, the IoMT

environment can benet from enhanced identity security and protected

communication for data exchange [126].

6.1.3. Data condentiality

Ensuring the condentiality of data within IoMT is of utmost

importance due to the sensitive features of medical data. IoMT encom-

passes an interconnected network of medical systems, devices, and

sensors that collect and transmit health-related information. Conden-

tiality entails the safety of a patient’s medical data shared with thera-

pists, physicians, and medical personnel from unauthorized disclosure to

individuals who may misuse or harm the patient or exploit the data

inappropriately [21]. To illustrate, if the condentiality of transferred

information is compromised, an antagonist could intervene between the

sender and receiver, intercept the medical data being transmitted, and

gain access to restricted information. Preserving the integrity of

healthcare technologies and devices against potential attacks is crucial,

and there is a pressing need for improvements in traditional systems to

mitigate these risks. Consequently, extensive efforts are dedicated to

developing multiple solutions in this domain. Regarding data con-

dentiality, blockchains facilitate interested parties within a given

network to access information while enforcing stringent security mea-

sures, such as role-based and attribute-based policies. Regrettably, there

have been numerous instances where patient data has been left

vulnerable [127].

The IoMT edge network encompasses IoT devices with limited re-

sources that strike challenges for implementing resource-intensive

cryptographic solutions like data encryption/decryption. Therefore, it

becomes challenging to maintain a high data secrecy level, making the

network vulnerable to threats that aim to compromise the privacy of

transferred or stored information [121]. For instance, an attacker can

use eavesdropping techniques to monitor communications and examine

the contents of transferred data packages within the IoMT edge network.

The adversary can then passively capture the conversation between the

wearable sensor and IoMT gateway and, by trafc analysis or other

means, extract private information [128].

Interrogation attacks involving impersonation pose a signicant risk

to data condentiality [122]. In this scenario, a malicious actor mas-

querades as a legitimate system, redirecting requests to other entities

solely to expose private information about users. Numerous methods are

available to verify data condentiality, spanning from physical safe-

guards to cryptographic algorithms that obfuscate information. Cryp-

tography pertains to concealing communication practices to enhance the

condentiality of stored data, offering various encoding schemes that

add protection when transmitting information through open channels or

interconnected systems. However, relying solely on cryptography is

insufcient to guarantee comprehensive information security. Thus,

alternative approaches, such as steganography, are necessary to mitigate

risks. Steganography, closely related to cryptography, provides an

additional layer of protection. Different types of steganography algo-

rithms can be identied by the methods they use to embed and retrieve

information [129,130]. A novel Crypto-Stegno model was introduced by

Li et al. [130] to safeguard medical information within the IoMT envi-

ronment. The study validated the effectiveness of using healthcare in-

formation datasets and demonstrated exceptional outcomes in terms of

perceptibility quality, resilience against data loss, embedding capacity,

and overall security. These remarkable results establish the

Crypto-Stegno system as a reliable and efcient approach for securing

medical information within the IoMT platform.

Secure communication between low-power IoMT devices like

medical sensors and nodes is possible with the help of a variety of

lightweight cryptographic algorithms. These include symmetric key ci-

phers like block and stream ciphers, as well as hash functions [131].

However, key ciphers face challenges related to the distribution of keys,

which is a crucial aspect of cryptographic security. The evolving nature

of cyber threats demands ongoing monitoring, auditing, and incident

response plans to detect and address potential breaches promptly.

Achieving and maintaining data condentiality in IoMT is an ongoing

challenge that necessitates a holistic and proactive approach to protect

sensitive medical information and safeguard patient trust.

6.1.4. Availability

The interruption of correct operations and impeded access to medical

information signicantly impact the availability of data, potentially

resulting in life-threatening consequences. This feature indicates the

accessibility of IoMT services, whereby authorized users may encounter

unavailability of data when requesting access to these services. Partic-

ularly in healthcare systems where constant monitoring of a patient’s

health is paramount, the availability aspect assumes crucial importance

[128]. Specically, availability assures that the information is exclu-

sively accessible to authorized entities. Nevertheless, healthcare systems

that heavily rely on IoMT devices face resource and computational

constraints, thereby posing challenges to the preservation of service

availability.

In cases where an attacker cannot compromise the condentiality

and integrity of ongoing communication within the IoMT system, they

may resort to launching alternative attacks. These include jamming at-

tacks, congestion, collision attacks, battery drainage attacks, tampering

attacks, and IoT-botnet attacks, which can be directed at various

network layers impacting the IoMT edge network. A battery drainage

attack can be executed by a malicious person who intentionally sends

deceptive content to the targeted device, causing excessive power con-

sumption and depletion of the device’s battery [132]. Collision attacks

occur when two nodes concurrently transfer data on a shared frequency

channel, leading to an identication conict at the receiving end.

Consequently, corrupted received data packets are discarded, resulting

in the retransfer of the packets and the waste of network resources

within the IoMT [133].

Numerous research efforts have concentrated on centralized or

partially centralized approaches and solutions aimed at mitigating

jamming attacks [134,135]. Xuan et al. [136] introduced a trigger

identication service to defend against susceptible jammers, which

identies and distinguishes nodes that exhibit transmission behavior

similar to the jamming nodes. The authors utilized optimization prob-

lems to create a comprehensive framework for trigger identication in

unreliable wireless sensor networks. They also proposed an improved

algorithm to enhance the scheme’s robustness in various network sce-

narios, particularly when facing sophisticated jamming models. While

this scheme effectively counters malicious actions, it relies on cloud or

server-based decision-making processes.

Although pattern recognition algorithms applied to node trans-

missions demonstrate promising potential, using a centralized system is

essential to manage computational costs. Additionally, the Strength of

Crowd (SoC) procedure offers a distributed solution, which is particu-

larly suitable for IoMT devices with limited resources. It also assures

message distribution to the intended receiving nodes, despite the

blocking of a substantial portion of the available bandwidth. Specif-

ically, the SoC protocol relies on a strategy of deception, where legiti-

mate devices transmit deceptive packets into the network, confusing

potential jammers and making it challenging to differentiate between

genuine and false nodes [137]. Another approach by Liu and Sun [138]

involves a lightweight algorithm, which is capable of recognizing pat-

terns or behaviors, coupled with a notication system. This combination

enables the detection of unusual activities within the IoMT environment.

To guarantee the accessibility of various IoMT applications, several

case studies have implemented blockchain technology, which expedites

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

access to sources and services via forecasting and management [139].

Using hash functions for prole matching, Nie et al. [140] investigated

the authentication of private data in IoMT while addressing the issue of

physical layer availability. In order to facilitate safe data sharing, the

authors implemented encryption methods and outlined a secret key for

prole matching. The study also introduced an optimized pricing

mechanism to incentivize greater user participation in health data

sharing and maximize user protability. Furthermore, an inspection

demonstrated that the proposed approach effectively meets various se-

curity objectives within IoMT scenarios.

6.1.5. Integrity

Integrity ensures the authenticity and accuracy of data, guaranteeing

that received messages are free from false information, unauthorized

modications, and deletions during communication. Within the context

of IoMT, integrity concerns revolve around the accuracy, dependability,

and coherence of produced, transmitted, and processed data and infor-

mation by interconnected medical devices and systems. The productive

integration of IoMT networks within the medical eld also heavily relies

on upholding the integrity of the associated devices (e.g., wearable or

implantable sensors inside the human body). Nonetheless, because IoMT

devices commonly function within environments lacking trust, they are

susceptible to physical assaults aiming to compromise device integrity

[122]. These concerns may have serious effects on patient security and

the quality of services provided. Unauthorized tampering or manipula-

tion of medical data can result in erroneous diagnoses, inappropriate

treatments, and inadequate patient care. For instance, altering vital

signs or laboratory test results can mislead healthcare professionals,

leading to inaccurate medical decisions and potentially detrimental

outcomes.

A man-in-the-middle (MitM) attack is an assault that threatens the

integrity of IoMT networks. An intruder in this scenario could secretly

listen in on a conversation between two parties, manipulating infor-

mation without raising suspicion [121]. For instance, in IoMT edge

networks, the gathered medical data can be either stored locally or

transferred to a remote server within the devices’ internal memory.

During transmission, data becomes vulnerable to interception and

modication by a MitM attacker, which ultimately compromises the

integrity of the data [141,142].

Seliem and Elgazzar [143] proposed a method for protecting con-

dentiality that makes use of a network cluster, cloud server, smart

medical appliances, and medical facilities. All transactions within the

system are hashed, and only the data holders possess the corresponding

hash values to ensure integrity. This method provides robust security

measures, even in situations where communication channels are

compromised. However, using cryptography and transaction handling

in this approach causes increased power consumption due to additional

computational overhead. Another way to securely store health data is by

leveraging the characteristics of blockchain technology. Blockchain of-

fers features like distributed architecture (eliminating a single point of

failure), transparency, near-immutability, secure cryptography, and the

ability to utilize smart contracts. These characteristics make blockchain

the preferred mechanism for ensuring data integrity in the IoMT Cloud

[144].

The secure framework for IoMT presented by Rathnayake et al. [145]

focuses on the key privacy, security, and sensor data integrity issues. The

framework, operating within a cloud-mobile architecture, relies on three

encryption processes: Proven Data Possession (PDP), Attribute-Based,

and Advanced Data Encryption. Utilizing these encryption methods

and leveraging cloud technologies, the proposed model successfully

addresses the specic challenges associated with medical applications.

Notably, the PDP technique conrms the integrity of encrypted les

using AES and ABE. This approach facilitates users to verify the presence

of their information on the server without the retrieval of the entire

dataset. Additionally, key exchange is facilitated through in-band pro-

tocols in the suggested framework.

To protect data during transit over the IoMT’s edge network, Lounis

et al. [146] proposed a hybrid strategy utilizing symmetric cryptography

and attribute-based encryption (ABE). The shared messages undergo

initial encryption using a randomly generated symmetric key (RSK) and

are encrypted further using ABE. If an IoMT device possesses the

appropriate secret key that fullls the access policy of ABE, the message

and RSK can be decrypted. The private key is associated with the attri-

bute set of the device, representing the user’s advantages. Notably, by

legitimately modifying the system’s conguration, there is an opportu-

nity to encrypt only the downloaded RSK rather than the whole message,

resulting in improved communication efciency and reduced costs

[147].

There are many ways in which healthcare can benet from IoMT, but

the eld also faces new challenges. For instance, authentication and

authorization issues emerge in the context of IoMT devices and systems,

necessitating rigorous verication and control mechanisms for accessing

sensitive patient data and devices. Maintaining patients’ condence and

complying with strict privacy regulations are at the forefront of privacy

concerns related to protecting large amounts of private medical infor-

mation from unauthorized access and breaches. Since interruptions or

outages in the connectivity of medical devices can have potentially

serious consequences, availability issues necessitate that IoMT systems

remain consistently operational and accessible. Finally, data integrity

must be maintained to ensure that health data is accurate and un-

changed throughout the transmission and storage processes, which is

crucial for facilitating well-informed medical decision-making and

protecting patient safety. The effective and secure integration of IoMT

into healthcare ecosystems necessitates addressing these complex

challenges.

As the medical sector embraces IoT solutions, it is evident that some

manufacturers are rushing to adopt these technologies without priori-

tizing security measures. This lack of emphasis on security induces se-

curity concerns regarding data/software. Future studies should

prioritize addressing security, privacy, risk assessment, standardization,

interoperability, and ethical considerations in IoMT deployments. These

aspects are crucial for ensuring the authentication, authorization,

condentiality, integrity, and availability of medical data and systems

while promoting responsible and effective use of IoT technologies.

6.2. Potential solutions to the security issues in IoMT

Healthcare institutions must guarantee that their devices are secure

from current threats, given the various ways that hackers might

compromise security and negatively impact clinical operations by

gaining access to clinical equipment [148]. There are a number of po-

tential solutions that could be used to address the security issues in

IoMT, including securing communication protocols like TLS or datagram

TLS to secure and authenticate data transferred between IoMT devices

and systems. Strong authentication methods, such as multi-factor

authentication, can also improve security by conrming the identity of

users and devices [149]. Utilizing powerful encryption methods helps to

protect sensitive data both in transit and at rest.

A complete solution for collecting, protecting, and storing data in the

IoMT should include both technical solutions and best practices. To

provide patients with healthcare services, it is necessary to collect,

retain, and analyze private medical data to correspond with the General

Data Protection Regulation (GDPR). Key security solutions for collect-

ing, protecting and storing data in the IoMT are outlined below:

(i) Data backup and recovery: This is the process of saving copies of

data so that it can be restored in the event of data loss or a system

crash. Data loss can be prevented with the use of off-site storage

and redundant backup systems. Test the data restoration pro-

cedure to ensure it can be relied upon.

(ii) Secure protocols: Use encrypted connections between devices

and gateways by employing secure communication protocols like

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

TLS or datagram TLS. These protocols allow for authenticated

and veried delivery of data without compromising its security or

integrity.

(iii) Encryption: Use robust cryptographic techniques to encrypt data

from beginning to end. As part of this process, data are encrypted

at the sending device, sent securely through networks, and then

decrypted at the receiving device. Both the medium of trans-

mission and the content of the data being transmitted should be

encrypted.

(iv) Access control: Use robust mechanisms for access control to

manage data transfer within the IoMT ecosystem. Role-based

access control (RBAC) can be used to restrict data transmission

and reception to approved people and devices. Data transmission

can be isolated and protected by using network segmentation and

virtual private networks (VPNs).

(v) Mutual authentication: To ensure the integrity of all communi-

cations, it is important to implement mutual authentication

across all devices and gateways. This reduces the likelihood of

attacks from malicious software or hardware by making it

possible for all parties involved in a transaction to verify each

other’s identities before any data is exchanged.

(vi) Intrusion prevention and detection: Organize intrusion preven-

tion and detection systems to keep tabs on network activity and

spot signs of hacking. Unauthorised login attempts and unusual

data patterns are two examples of what these systems can detect

to send out alerts about or even take preventative measures

against.

(vii) Data loss prevention: Put preventative measures in place. To

prevent the loss of vital medical records, it is necessary to take

measures like keeping backups, establishing systems to handle

packet loss, and assuring data retransmission or redundancy.

(viii) Data lifecycle management: Dene the collection, storage, access,

and deletion of data inside the IoMT infrastructure and provide

unambiguous data lifecycle management policies. To protect user

privacy and remain in regulatory compliance, it is important to

implement data retention policies, data archiving protocols, and

secure data disposal techniques.

(ix) Trafc monitoring and analysis: Use network monitoring tools to

keep tabs on data ows and look for anything out of the ordinary

that could pose a security risk. This allows for continuous

tracking of data transfers and the early detection of any suspi-

cious activities.

(x) Security audits and testing: Perform routine security audits and

penetration testing to locate security holes in the IoMT platform.

Assessing the security posture of a system beforehand allows

businesses to pinpoint vulnerabilities and implement xes that

better protect sensitive information during transit.

(xi) Checks for data integrity: Implement procedures to ensure that

data is accurate. Checksums, digital signatures, and hash func-

tions are examples of techniques that can be used to protect data

from being altered without detection. In order to detect data

integrity breaches, regular integrity tests should be performed.

(xii) Security awareness and training: Inform medical staff, IT man-

agers, and end users of recommended practices for keeping sen-

sitive information safe throughout transmission. Regular training

sessions should be held to ensure that all employees know how to

handle sensitive information safely and securely while in transit.

Use encrypted connections between devices and gateways by

employing secure communication protocols like TLS or datagram

TLS. These protocols allow for authenticated and veried de-

livery of data without compromising its security or integrity.

(xiii) Firmware and software updates: Update the rmware and soft-

ware on your IoMT devices, gateways, and network infrastructure

on a regular basis. This protects the system from any new threats

that may arise by applying the latest security updates and xing

any known vulnerabilities.

By adopting these data transportation security solutions, healthcare

providers may better protect patient data, uphold the IoMT ecosystem,

and reduce the likelihood of data breaches caused by hacking, eaves-

dropping, or other forms of malicious activity. In conclusion, addressing

security concerns with IoMT calls for a multi-pronged strategy that in-

corporates robust authentication systems, encryption, privacy-

enhancing technology, standardized security procedures, patch man-

agement strategies, and employee education and training. It is essential

for manufacturers, healthcare organizations, regulatory agencies, and

cybersecurity specialists to work together to secure patient data and

ensure the security and privacy of IoMT systems.

7. Open issues and challenges in IoMT

Connecting medical devices, sensors, and healthcare systems

through the IoMT has the potential to revolutionize healthcare by

enhancing patient care and streamlining administrative tasks. However,

for the IoMT ecosystem to be widely adopted and successful, several

open issues and challenges must be resolved. Some of the most signi-

cant obstacles include:

■ Protecting the privacy and condentiality of patient information is of

utmost importance in the healthcare industry. The risk of data

breaches and unauthorised access is becoming increasingly impor-

tant as the number of connected devices and data transmission

increases.

■ The consequences of a healthcare system going down or a medical

device breaking can be devastating. It is crucial to guarantee the

dependability and accessibility of IoMT tools and infrastructure.

■ The quality and accuracy of the data collected by IoMT devices are

essential in enabling knowledgeable medical decision-making.

Maintaining trustworthy and accurate data is an ongoing issue.

■ The devices and systems that comprise IoMT frequently come from

various vendors and employ different communication protocols,

raising the issue of interoperability. However, ensuring that all of

these devices can communicate with one another and exchange data

without any difculties is no easy task.

■ As the number of connected devices and data sets increases, it can be

difcult to effectively manage and scale the underlying infrastruc-

ture supporting the IoMT ecosystem.

■ The proper use of IoMT devices and the interpretation of the data

they produce calls for proper training and education on the part of

healthcare professionals. It can be challenging to make sure everyone

working in healthcare has sufcient education.

■ Patients must have faith in IoMT tools and infrastructure for it to be

useful. Trust from patients must be earned and kept over time.

■ Delays in rolling out IoMT solutions can be caused by a lack of

suitable infrastructure and unreliable connectivity in some areas,

especially rural ones.

■ Although there have been some successful attempts to standardize

IoMT technologies, there is still a lack of universal standards that can

make it difcult for different IoMT technologies to work together

smoothly.

■Life-threatening consequences could result from hacking or unau-

thorised access to medical devices, making it imperative to ensure

their security.

■Some ethical issues arise when using IoMT, including patient con-

sent, data misuse, and the ethical relevance of AI-driven diagnoses

and treatment suggestions.

■ The issue of ascertaining ownership and control of data generated by

IoMT devices presents an ongoing challenge, with the question of

whether it belongs to the healthcare provider, patient, or device

manufacturer yet to be denitively resolved.

To overcome the issues mentioned above, the device manufacturers,

healthcare community, technologists, and regulators must work

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

together. New difculties and prospects are likely to arise as the eld of

IoMT develops further, highlighting the importance of continuing

research and development.

8. Future directions in IoMT research

With the advancement of new network security technologies and the

creation of additional digital medical devices, IoMT is constantly

evolving. Future studies should concentrate on the issues mentioned

below.

•IoMT has no set architecture that must be adhered to by everyone

who develops IoT applications. Some gadgets, including CCTV

cameras and eld sensors, lack the ability to update their software,

making them security-wise obsolete after a certain period of time. As

a result, newer, more modern versions must be used in their place.

Therefore, future studies should investigate the possibility of

upgrading these devices when they are connected to the IoMT to

reduce the implementation cost of visual surveillance over the IoMT.

•Protecting the IoMT application requires secure communication

channels, early identication of system faults, and the installation of

the right software and hardware. These regulations are essential to

the organization’s effective operation. Since IoMT involves a variety

of more private and sensitive data, it is crucial to identify which

patient data require the patient’s specic consent for access in order

to protect his or her fundamental rights. As a result, future adoption

of effective policy tailored specically for the IoMT is required.

•IoMT places a lot of importance on network research. Channel-based

attacks can damage the integrity of the original data and, thus, are

extremely dangerous because they could happen while the data is in

transit. The variability of the IoMT environment makes it challenging

to implement malware detection solutions. In a cross-platform

environment, it is challenging to build tools and procedures that

can detect malware. For instance, the system would become more

complex and malware would be tougher to discover if data and in-

formation were transferred from smart homes to smart healthcare

systems for patient monitoring. Consequently, continuous network-

based malware strategies must be created.

•IoMT is comprised of a wide range of devices, applications, networks,

and data transport paradigms, each with its own unique set of fea-

tures and requirements. As a consequence, nding malware is a

really difcult task. One method to address heterogeneity issues in

the IoMT is to utilize an "electronic health recorder" to store user data

over an IoT-based cloud for processing. In the IoMT setting, the BAN

also generates a tremendous amount of information. Consequently, a

more focused strategy needs to be developed in order to detect

malware in a setting this complicated. Moreover, each device has a

unique operating system, range of communication, storage capacity,

and power consumption. Therefore, a more specialised approach to

attack detection in such a system needs to be developed.

•IoMT-based systems contain a sizeable amount of condential and

sensitive data, which continues to grow with time. The security of a

system based on IoMT might be strengthened in several ways,

including through the use of cryptography and blockchain technol-

ogy. Data integrity is improved by using SHA-256 efciently.

Blockchain technology is being employed in a variety of technolog-

ical disciplines, and its application in the Internet of Things would be

very helpful in ensuring data integrity. Moreover, the IoMT-based

technology uses extremely sensitive data. Patients are the sole con-

trollers of such data, excluding healthcare professionals. As a result,

there is always a possibility of data leakage. The time-stamped

feature of blockchain technology could prove invaluable for veri-

fying the accuracy of such records. Once these records are stored on a

decentralized ledger, it will be easier to identify anomalies in patient

data.

•Detecting malware in cross-platform IoMT systems is a signicant

challenge that must be met to ensure the safety of patients, the

condentiality of personal information, and the reliability of

healthcare services as a whole. There needs to be more study done in

this area to prevent safety issues and other problems with IoMT in

healthcare from impeding its usefulness.

System software and application software play a crucial role in

facilitating a wide range of medical processes by acting as bridges be-

tween the software and hardware layers, respectively. Consequently, the

research and development in this domain emerge as a critical area of

focus for ensuring IoMT’s smooth functioning. It is critical to regularly

detect any defects in the codes of these programs; hence, a standardized

proper framework is required to design the security monitoring system

in the software. The operating system code in IoMT devices must also be

analysed to detect the presence of zero-day threats.

9. Conclusion

The applications of IoT and IoMT continue to expand daily in a wide

range of platforms for exchanging information in multiple industries.

IoMT is increasing the amount of available data from numerous sources,

including clinical systems, wearables, electronic health records, and

medical devices. Data fusion involves integrating these diverse data

sources, such as imaging scans and laboratory results, to generate a more

comprehensive picture of a patient’s health status. This allows health-

care providers to establish more precise diagnoses and treatment plans.

The accuracy of predictions is directly impacted by the quality, quantity,

and relevance of the data obtained from IoMT devices. While the

multimodal emotion recognition (MEmoR) model provided a minimum

accuracy of 81.54 % in predicting discrete emotion, the Epilepsy seizure

detector-based Naive Bayes (ESDNB) algorithm was found to be the most

effective for detecting epileptic seizures in IoMT networks, with an ac-

curacy of 99.53 % to 99.99 %. However, IoMT faces a number of chal-

lenges with data fusion. Data standardization is essential due to the fact

that data from numerous systems and devices may be stored in different

formats, making it difcult to integrate and analyze. Privacy and secu-

rity issues are major concerns since data sent via the Internet are

vulnerable to hacking and unauthorised access. There is also an

important need to revolutionize how data are collected, transported, and

stored.

CRediT authorship contribution statement

Shams Forruque Ahmed: Conceptualization, Writing – original

draft, Supervision. Md. Sakib Bin Alam: Methodology, Writing – orig-

inal draft. Shaila Afrin: Writing – original draft, Validation. Sabiha

Jannat Rafa: Writing – original draft, Data curation. Nazifa Rafa:

Writing – original draft, Formal analysis. Amir H. Gandomi: Writing –

review & editing, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing nancial

interests or personal relationships that could have appeared to inuence

the work reported in this paper.

Data availability

No data was used for the research described in the article.

Acknowledgments

The authors highly express their gratitude to the Asian University for

Women, Chattogram, Bangladesh, for their support in carrying out this

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

study.

References

[1] A. Rahman, M.S. Hossain, G. Muhammad, D. Kundu, T. Debnath, M. Rahman, M.

S.I. Khan, P. Tiwari, S.S. Band, Federated learning-based AI approaches in smart

healthcare: concepts, taxonomies, challenges and open issues, Cluster Comput. 26

(2023) 2271–2311.

[2] S. Vishnu, S.R. Jino Ramson, R. Jegan, Internet of Medical Things (IoMT)-an

overview, in: ICDCS 2020 - 2020 5th Int. Conf. Devices, Circuits Syst, 2020,

pp. 101–104, https://doi.org/10.1109/ICDCS48716.2020.243558.

[3] M.M. Islam, S. Nooruddin, F. Karray, G. Muhammad, Internet of Things: device

capabilities, architectures, protocols, and smart applications in healthcare

domain, IEEE Internet Things J 10 (4) (2022) 3611–3641.

[4] A. Si-Ahmed, M.A. Al-Garadi, N. Boustia, Survey of machine learning based

intrusion detection methods for Internet of Medical Things, Appl. Soft Comput.

140 (2022), 110227, https://doi.org/10.1016/j.asoc.2023.110227.

[5] G. Muhammad, F. Alshehri, F. Karray, A. El Saddik, M. Alsulaiman, T.H. Falk,

A comprehensive survey on multimodal medical signals fusion for smart

healthcare systems, Inf. Fusion. 76 (2021), https://doi.org/10.1016/j.

inffus.2021.06.007.

[6] L. Wang, Y. Ali, S. Nazir, M. Niazi, ISA evaluation framework for security of

internet of health things system using AHP-TOPSIS methods, IEEE Access 8

(2020) 152316–152332, https://doi.org/10.1109/ACCESS.2020.3017221.

[7] J.N.S. Rubí, P.R.de L. Gondim, Interoperable Internet of Medical Things platform

for e-Health applications, Int. J. Distrib. Sens. Netw. 16 (2020), https://doi.org/

10.1177/1550147719889591.

[8] X. Li, H. Dai, Q. Wang, M. Imran, D. Li, M. Ali, Securing Internet of Medical

Things with friendly-jamming schemes, Comput. Commun. 160 (2020) 431–442,

https://doi.org/10.1016/j.comcom.2020.06.026.

[9] A. Ghubaish, T. Salman, M. Zolanvari, D. Unal, A. Al-Ali, R. Jain, Recent advances

in the Internet-of-Medical-Things (IoMT) systems security, IEEE Internet Things

J. 8 (2021), https://doi.org/10.1109/JIOT.2020.3045653.

[10] B. Shanmugam, Risk Assessment of Heterogeneous IoMT Devices : a Review,

2023.

[11] W. Ding, X. Jing, Z. Yan, L.T. Yang, A survey on data fusion in internet of things:

towards secure and privacy-preserving fusion, Inf. Fusion. 51 (2019) 129–144,

https://doi.org/10.1016/j.inffus.2018.12.001.

[12] H. Lin, S. Garg, J. Hu, X. Wang, M. Jalil Piran, M.S. Hossain, Privacy-enhanced

data fusion for COVID-19 applications in intelligent Internet of Medical Things,

IEEE Internet Things J 8 (2021) 15683–15693, https://doi.org/10.1109/

JIOT.2020.3033129.

[13] M. Al-Hawawreh, M.S. Hossain, A privacy-aware framework for detecting cyber

attacks on internet of medical things systems using data fusion and quantum deep

learning, Inf. Fusion. (2023), 101889.

[14] Y. Sun, F.P.W. Lo, B. Lo, Security and privacy for the Internet of Medical Things

enabled healthcare systems: a survey, IEEE Access 7 (2019), https://doi.org/

10.1109/ACCESS.2019.2960617.

[15] M. Wazid, A.K. Das, J.J.P.C. Rodrigues, S. Shetty, Y. Park, IoMT malware

detection approaches: analysis and research challenges, IEEE Access 7 (2019)

182459–182476, https://doi.org/10.1109/ACCESS.2019.2960412.

[16] T. Yaqoob, H. Abbas, M. Atiquzzaman, Security vulnerabilities, attacks,

countermeasures, and regulations of networked medical devices-a review, IEEE

Commun. Surv. Tutorials. 21 (2019) 3723–3768, https://doi.org/10.1109/

COMST.2019.2914094.

[17] P. Pratim, D. Dash, N. Kumar, Sensors for internet of medical things : state-of-the-

art , security and privacy issues , challenges and future directions, Comput.

Commun. 160 (2020) 111–131, https://doi.org/10.1016/j.comcom.2020.05.029.

[18] A. Hazah, M. Aman, W. Haslina, S. Sameen, Journal of network and computer

applications IoMT amid COVID-19 pandemic : application , architecture ,

technology , and security, J. Netw. Comput. Appl. 174 (2021), 102886, https://

doi.org/10.1016/j.jnca.2020.102886.

[19] R. Dwivedi, D. Mehrotra, S. Chandra, Potential of Internet of Medical Things

(IoMT) applications in building a smart healthcare system: a systematic review,

J. Oral Biol. Craniofacial Res. 12 (2022), https://doi.org/10.1016/j.

jobcr.2021.11.010.

[20] P. Wal, A. Wal, N. Verma, R. Karunakakaran, A. Kapoor, Internet of Medical

Things – the future of healthcare, Open Public Health J. 15 (2022) 1–9, https://

doi.org/10.2174/18749445-v15-e221215-2022-142.

[21] M. Papaioannou, M. Karageorgou, G. Mantas, V. Sucasas, I. Essop, J. Rodriguez,

D. Lymberopoulos, A survey on security threats and countermeasures in Internet

of Medical Things (IoMT), Trans. Emerg. Telecommun. Technol. 33 (2022),

https://doi.org/10.1002/ett.4049.

[22] Z. Ashfaq, A. Rafay, R. Mumtaz, S. Mohammad, H. Zaidi, H. Saleem, S. Ali,

R. Zaidi, S. Mumtaz, A. Haque, A review of enabling technologies for Internet of

Medical Things ( IoMT ) ecosystem, Ain Sham. Eng. J. 13 (2022), 101660,

https://doi.org/10.1016/j.asej.2021.101660.

[23] A. Hemmati, A.M. Rahmani, Internet of Medical Things in the COVID-19 era: a

systematic literature review, Sustain 14 (2022), https://doi.org/10.3390/

su141912637.

[24] M.J. Sudha, S. Viveka, A comprehensive review of architecture, classication ,

challenges, and future of the Internet of Medical Things (IoMTs ), Med. J.

Babylon. 19 (2022) 311–317, https://doi.org/10.4103/MJBL.MJBL.

[25] F. Kamalov, B. Pourghebleh, M. Gheisari, Y. Liu, S. Moussa, Internet of Medical

Things privacy and security: challenges, solutions, and future trends from a new

perspective, Sustain 15 (2023), https://doi.org/10.3390/su15043317.

[26] B. Bhushan, A. Kumar, A.K. Agarwal, A. Kumar, P. Bhattacharya, A. Kumar,

Towards a secure and sustainable Internet of Medical Things (IoMT):

requirements, design challenges, security techniques, and future trends, Sustain

15 (2023), https://doi.org/10.3390/su15076177.

[27] S.F. Ahmed, N. Rafa, T. Mehnaz, B. Ahmed, N. Islam, M. Mojur, A.T. Hoang, G.

M. Shaullah, Integration of phase change materials in improving the

performance of heating, cooling, and clean energy storage systems: an overview,

J. Clean. Prod. 364 (2022), 132639, https://doi.org/10.1016/j.

jclepro.2022.132639.

[28] J.E. Ferguson, A.D. Redish, Wireless communication with implanted medical

devices using the conductive properties of the body, Expert Rev. Med. Dev. 8

(2011) 427–433, https://doi.org/10.1586/erd.11.16.

[29] A. Kos, V. Milutinovi´

c, A. Umek, Challenges in wireless communication for

connected sensors and wearable devices used in sport biofeedback applications,

Futur. Gener. Comput. Syst. 92 (2019) 582–592, https://doi.org/10.1016/j.

future.2018.03.032.

[30] T. Belkhouja, S. Sorour, M.S. Hefeida, Role-based hierarchical medical data

encryption for implantable medical devices, in: 2019 IEEE Glob. Commun. Conf.

GLOBECOM 2019 - Proc, 2019, pp. 1–6, https://doi.org/10.1109/

GLOBECOM38437.2019.9014192.

[31] E. Meng, R. Sheybani, Insight: implantable medical devices, Lab. Chip. 14 (2014)

3233–3240, https://doi.org/10.1039/c4lc00127c.

[32] K.G. Tarakji, A.M. Zaidi, S.L. Zweibel, N. Varma, S.F. Sears, J. Allred, P.

R. Roberts, N.A. Shaik, J.R. Silverstein, A. Maher, S. Mittal, A. Patwala,

J. Schoenhard, M. Emert, G. Molon, G. Augello, N. Patel, H. Seide, A. Porlio,

B. Maus, S.L. Di Jorio, K. Holloman, A.C. Natera, M.P. Turakhia, Performance of

rst pacemaker to use smart device app for remote monitoring, Hear. Rhythm O

2 (2021) 463–471, https://doi.org/10.1016/j.hroo.2021.07.008.

[33] B. Priya Prathaban, R. Balasubramanian, R. Kalpana, ForeSeiz: an IoMT based

headband for real-time epileptic seizure forecasting, Expert Syst. Appl. 188

(2022), 116083, https://doi.org/10.1016/j.eswa.2021.116083.

[34] M.A. Sayeed, S.P. Mohanty, E. Kougianos, H. Zaveri, IDDS: an edge-device in

IoMT for automatic seizure control using on-time drug delivery, in: Dig. Tech.

Pap. - IEEE Int. Conf. Consum. Electron. 2020-January, 2020, https://doi.org/

10.1109/ICCE46568.2020.9043143.

[35] B.A. Alzahrani, A. Irshad, A. Albeshri, K. Alsubhi, A provably secure and

lightweight patient-healthcare authentication protocol in wireless body area

networks, Wirel. Pers. Commun. 117 (2021) 47–69, https://doi.org/10.1007/

s11277-020-07237-x.

[36] P. Sasidharan, T. Rajalakshmi, U. Snekhalatha, Wearable cardiorespiratory

monitoring device for heart attack prediction, in: Proc. 2019 IEEE Int. Conf.

Commun. Signal Process. ICCSP 2019, 2019, pp. 54–57, https://doi.org/

10.1109/ICCSP.2019.8698059.

[37] R. Lazazzera, Y. Belhaj, G. Carrault, A newwearable device for blood pressure

estimation using photoplethysmogram, Sensors (Switzerland) 19 (2019) 1–18,

https://doi.org/10.3390/s19112557.

[38] Q. Lin, W. Xu, J. Liu, A. Khamis, W. Hu, M. Hassan, A. Seneviratne, H2B:

heartbeat-based secret key generation using piezo vibration sensors, in: IPSN

2019 - Proc. 2019 Inf. Process. Sens. Networks, 2019, pp. 265–276, https://doi.

org/10.1145/3302506.3310406.

[39] R. L´

opez-Blanco, M.A. Velasco, A. M´

endez-Guerrero, J.P. Romero, M.D. del

Castillo, J.I. Serrano, E. Rocon, J. Benito-Le´

on, Smartwatch for the analysis of rest

tremor in patients with Parkinson’s disease, J. Neurol. Sci. 401 (2019) 37–42,

https://doi.org/10.1016/j.jns.2019.04.011.

[40] H. Ryu, H. moon Park, M.K. Kim, B. Kim, H.S. Myoung, T.Y. Kim, H.J. Yoon, S.

S. Kwak, J. Kim, T.H. Hwang, E.K. Choi, S.W. Kim, Self-rechargeable cardiac

pacemaker system with triboelectric nanogenerators, Nat. Commun. 12 (2021)

1–9, https://doi.org/10.1038/s41467-021-24417-w.

[41] S. Panja, A.K. Chattopadhyay, A. Nag, J.P. Singh, Fuzzy-logic-based IoMT

framework for COVID19 patient monitoring, Comput. Ind. Eng. 176 (2023),

https://doi.org/10.1016/j.cie.2022.108941.

[42] S. Razdan, S. Sharma, Internet of Medical Things (IoMT): overview, emerging

technologies, and case studies, in: IETE Tech. Rev. (Institution Electron.

Telecommun. Eng. India), 2022, p. 39, https://doi.org/10.1080/

02564602.2021.1927863.

[43] K. Wei, L. Zhang, Y. Guo, X. Jiang, Health monitoring based on Internet of

Medical Things: architecture, enabling technologies, and applications, IEEE

Access 8 (2020), https://doi.org/10.1109/ACCESS.2020.2971654.

[44] M. Boumaiz, M. El Ghazi, S. Mazer, M. Fattah, A. Bouayad, M. El Bekkali,

Y. Balboul, Energy harvesting based WBANs: EH optimization methods, in:

Procedia Comput. Sci., 2019, https://doi.org/10.1016/j.procs.2019.04.147.

[45] M. Cicio˘

glu, A. Çalhan, SDN-based wireless body area network routing algorithm

for healthcare architecture, ETRI J. 41 (2019), https://doi.org/10.4218/

etrij.2018-0630.

[46] A.S. Abiodun, M.H. Anisi, M.K. Khan, Cloud-based wireless body area networks:

managing data for better health care, IEEE Consum. Electron. Mag. 8 (2019),

https://doi.org/10.1109/MCE.2019.2892244.

[47] A.H. Mohd Aman, W.H. Hassan, S. Sameen, Z.S. Attarbashi, M. Alizadeh, L.

A. Latiff, IoMT amid COVID-19 pandemic: application, architecture, technology,

and security, J. Netw. Comput. Appl. 174 (2021), https://doi.org/10.1016/j.

jnca.2020.102886.

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060

[48] R. Hireche, H. Mansouri, A.-S.K. Pathan, Security and privacy management in

Internet of Medical Things (IoMT): a synthesis, J. Cybersecur. Priv. 2 (2022),

https://doi.org/10.3390/jcp2030033.

[49] B.S Pritika, S. Azam, Risk Assessment of Heterogeneous IoMT Devices: A Review,

Technologies, 2023, p. 11, https://doi.org/10.3390/technologies11010031.

[50] J. Srivastava, S. Routray, S. Ahmad, M.M. Waris, Internet of Medical Things

(IoMT)-based smart healthcare system: trends and progress, Comput. Intell.

Neurosci 2022 (2022), https://doi.org/10.1155/2022/7218113.

[51] A. Ahad, M. Tahir, K.L.A. Yau, 5G-based smart healthcare network: architecture,

taxonomy, challenges and future research directions, IEEE Access 7 (2019),

https://doi.org/10.1109/ACCESS.2019.2930628.

[52] J. Sengupta, S. Ruj, S.Das Bit, A comprehensive survey on attacks, security issues

and blockchain solutions for IoT and IIoT, J. Netw. Comput. Appl. 149 (2020),

https://doi.org/10.1016/j.jnca.2019.102481.

[53] M.M. Islam, S. Nooruddin, F. Karray, G. Muhammad, Human activity recognition

using tools of convolutional neural networks: a state of the art review, data sets,

challenges, and future prospects, Comput. Biol. Med. (2022), 106060.

[54] M.S. Hossain, G. Muhammad, Emotion-aware connected healthcare big data

towards 5 G, IEEE Internet Things J. 5 (2018), https://doi.org/10.1109/

JIOT.2017.2772959.

[55] T. Hussain, K. Muhammad, S. Khan, A. Ullah, M.Y. Lee, S.W. Baik, Intelligent

baby behavior monitoring using embedded vision in IoT for smart healthcare

centers, J. Artif. Intell. Syst. 1 (2019), https://doi.org/10.33969/ais.2019.11007.

[56] J.H. Abawajy, M.M. Hassan, Federated Internet of Things and cloud computing

pervasive patient health monitoring system, IEEE Commun. Mag. 55 (2017),

https://doi.org/10.1109/MCOM.2017.1600374CM.

[57] M.M. Islam, S. Nooruddin, F. Karray, G. Muhammad, Multi-level feature fusion

for multimodal human activity recognition in Internet of Healthcare Things, Inf.

Fusion. 94 (2023) 17–31.

[58] F. Alam, R. Mehmood, I. Katib, N.N. Albogami, A. Albeshri, Data fusion and IoT

for smart ubiquitous environments: a survey, IEEE Access 5 (2017), https://doi.

org/10.1109/ACCESS.2017.2697839.

[59] R. Dautov, S. Distefano, R. Buyya, Hierarchical data fusion for smart healthcare,

J. Big Data. 6 (2019), https://doi.org/10.1186/s40537-019-0183-6.

[60] M. Chen, J. Yang, J. Zhou, Y. Hao, J. Zhang, C.H. Youn, 5G-smart diabetes:

toward personalized diabetes diagnosis with healthcare big data clouds, IEEE

Commun. Mag. 56 (2018), https://doi.org/10.1109/MCOM.2018.1700788.

[61] D. Naranjo-Hern´

andez, A. Talaminos-Barroso, J. Reina-Tosina, L.M. Roa,

G. Barbarov-Rostan, P. Cejudo-Ramos, E. M´

arquez-Martín, F. Ortega-Ruiz, Smart

vest for respiratory rate monitoring of copd patients based on non-contact

capacitive sensing, Sens. (Switzerl.) 18 (2018), https://doi.org/10.3390/

s18072144.

[62] H. Zhang, J. Li, B. Wen, Y. Xun, J. Liu, Connecting intelligent things in smart

hospitals using NB-IoT, IEEE Internet Things J. 5 (2018), https://doi.org/

10.1109/JIOT.2018.2792423.

[63] A. Almas, W. Iqbal, A. Altaf, K. Saleem, S. Mussiraliyeva, M.W. Iqbal, Context-

based adaptive fog computing trust solution for time-critical smart healthcare

systems, IEEE Internet Things J. (2023), https://doi.org/10.1109/

JIOT.2023.3242126.

[64] B. Farahani, F. Firouzi, V. Chang, M. Badaroglu, N. Constant, K. Mankodiya,

Towards fog-driven IoT eHealth: promises and challenges of IoT in medicine and

healthcare, Futur. Gener. Comput. Syst. 78 (2018), https://doi.org/10.1016/j.

future.2017.04.036.

[65] P. Pace, G. Aloi, R. Gravina, G. Caliciuri, G. Fortino, A. Liotta, An edge-based

architecture to support efcient applications for healthcare industry 4.0, IEEE

Trans. Ind. Informat. 15 (2019), https://doi.org/10.1109/TII.2018.2843169.

[66] H. Fouad, A.S. Hassanein, A.M. Soliman, H. Al-Feel, Internet of medical things

(IoMT) assisted vertebral tumor prediction using heuristic hock transformation

based gautschi model-A numerical approach, IEEE Access 8 (2020) 17299–17309,

https://doi.org/10.1109/ACCESS.2020.2966272.

[67] H. Cai, Z. Qu, Z. Li, Y. Zhang, X. Hu, B. Hu, Feature-level fusion approaches based

on multimodal EEG data for depression recognition, Inf. Fusion. 59 (2020)

127–138.

[68] M.J. Willemink, W.A. Koszek, C. Hardell, J. Wu, D. Fleischmann, H. Harvey, L.

R. Folio, R.M. Summers, D.L. Rubin, M.P. Lungren, Preparing medical imaging

data for machine learning, Radiology 295 (2020) 4–15.

[69] M. Illimoottil, D. Ginat, Recent advances in deep learning and medical imaging

for head and neck cancer treatment: MRI, CT, and PET scans, Cancer. (Basel) 15

(2023) 3267.

[70] M. Kang, E. Ko, T.B. Mersha, A roadmap for multi-omics data integration using

deep learning, Brief. Bioinform. 23 (2022) bbab454.

[71] F. Firouzi, B. Farahani, Architecting IoT Cloud. Intell. Internet Things from Device

to Fog Cloud, 2020, https://doi.org/10.1007/978-3-030-30367-9_4.

[72] F. Firouzi, B. Farahani, A. Marinˇ

sek, The convergence and interplay of edge, fog,

and cloud in the AI-driven Internet of Things (IoT), Inf. Syst. (2021), 101840,

https://doi.org/10.1016/J.IS.2021.101840.

[73] I. Martinez, A.S. Had, A. Jarray, Design, resource management, and evaluation

of fog computing systems: a survey, IEEE Internet Things J. 8 (2021), https://doi.

org/10.1109/JIOT.2020.3022699.

[74] A. Yousefpour, C. Fung, T. Nguyen, K. Kadiyala, F. Jalali, A. Niakanlahiji, J. Kong,

J.P. Jue, All one needs to know about fog computing and related edge computing

paradigms: a complete survey, J. Syst. Archit. 98 (2019) 289–330, https://doi.

org/10.1016/j.sysarc.2019.02.009.

[75] M. Aazam, S. Zeadally, K.A. Harras, Ofoading in fog computing for IoT: review,

enabling technologies, and research opportunities, Futur. Gener. Comput. Syst. 87

(2018), https://doi.org/10.1016/j.future.2018.04.057.

[76] F. Firouzi, S. Jiang, K. Chakrabarty, B. Farahani, M. Daneshmand, J. Song,

K. Mankodiya, Fusion of IoT, AI, edge-fog-cloud, and blockchain: challenges,

solutions, and a case study in healthcare and medicine, IEEE Internet Things J. 10

(2023), https://doi.org/10.1109/JIOT.2022.3191881.

[77] J. Li, X. Yang, G. Chu, W. Feng, X. Ding, X. Yin, L. Zhang, W. Lv, L. Ma, L. Sun,

R. Feng, J. Qin, X. Zhang, C. Gou, Z. Yu, B. Wei, W. Jiao, Y. Wang, L. Luo,

H. Yuan, Y. Chang, Q. Cai, S. Wang, P.C. Giulianotti, Q. Dong, H. Niu, Application

of improved robot-assisted laparoscopic telesurgery with 5 G technology in

urology, Eur. Urol 83 (2023), https://doi.org/10.1016/j.eururo.2022.06.018.

[78] S. Duan, L. Liu, Y. Chen, L. Yang, Y. Zhang, S. Wang, L. Hao, L. Zhang, A 5G-

powered robot-assisted teleultrasound diagnostic system in an intensive care unit,

Crit. Care 25 (2021), https://doi.org/10.1186/s13054-021-03563-z.

[79] W. Ding, M. Abdel-Basset, H. Hawash, S. Abdel-Razek, C. Liu, Fed-ESD: federated

learning for efcient epileptic seizure detection in the fog-assisted internet of

medical things, Inf. Sci. (Ny). 630 (2023), https://doi.org/10.1016/j.

ins.2023.02.052.

[80] A.K. Idrees, S.K. Idrees, R. Couturier, T. Ali-Yahiya, An edge-fog computing-

enabled lossless eeg data compression with epileptic seizure detection in IoMT

networks, IEEE Internet Things J. 9 (2022), https://doi.org/10.1109/

JIOT.2022.3143704.

[81] S. Hassan, E. Mwangi, P.K. Kihato, IoT based monitoring system for epileptic

patients, Heliyon 8 (2022), https://doi.org/10.1016/j.heliyon.2022.e09618.

[82] A. Ahmed, S. Aziz, U. Qidwai, A. Abd-Alrazaq, J. Sheikh, Performance of articial

intelligence models in estimating blood glucose level among diabetic patients

using non-invasive wearable device data, Comput. Method. Program. Biomed.

Updat. 3 (2023), https://doi.org/10.1016/j.cmpbup.2023.100094.

[83] A. Kumar, K. Sharma, A. Sharma, MEmoR: a Multimodal Emotion Recognition

using affective biomarkers for smart prediction of emotional health for people

analytics in smart industries, Image Vis. Comput. 123 (2022), https://doi.org/

10.1016/j.imavis.2022.104483.

[84] A. Arora, P. Chakraborty, M.P.S. Bhatia, Identifying digital biomarkers in

actigraph based sequential motor activity data for assessment of depression: a

model evaluating SVM in LSTM extracted feature space, Int. J. Inf. Technol. 15

(2023), https://doi.org/10.1007/s41870-023-01162-5.

[85] M.S. Alauddin, A.S. Baharuddin, M.I.M. Ghazali, The modern and digital

transformation of oral health care: a mini review, Healthc 9 (2021), https://doi.

org/10.3390/healthcare9020118.

[86] S. Salagare, R. Prasad, Internet of Dental Things (IoDT), intraoral wireless

sensors, and teledentistry: a novel model for prevention of dental caries, Wirel.

Pers. Commun. 123 (2022), https://doi.org/10.1007/s11277-021-09287-1.

[87] N. Martin, S. Shahrbaf, A. Towers, C. Stokes, C. Storey, Remote clinical

consultations in restorative dentistry: a clinical service evaluation study, Br. Dent.

J. 228 (2020), https://doi.org/10.1038/s41415-020-1328-x.

[88] Y. Yin, H. Wang, S. Liu, J. Sun, P. Jing, Y. Liu, Internet of Things for diagnosis of

Alzheimer’s disease: a multimodal machine learning approach based on eye

movement features, IEEE Internet Things J. 10, (13), (2023), 11476-1148510.110

9/JIOT.2023.3245067.

[89] H. Gao, L. Zhou, J.Y. Kim, Y. Li, W. Huang, Applying probabilistic model checking

to the behavior guidance and abnormality detection for A-MCI patients under

wireless sensor network, ACM Trans. Sens. Networks. 19 (2023), https://doi.org/

10.1145/3499426.

[90] F. Gonz´

alez-Landero, I. García-Magari˜

no, R. Amariglio, R. Lacuesta, Smart

cupboard for assessing memory in home environment, Sens. (Switzerl.) 19

(2019), https://doi.org/10.3390/s19112552.

[91] N. Raheja, A. Kumar Manocha, An IoT enabled secured clinical health care

framework for diagnosis of heart diseases, Biomed. Signal Process. Control. 80

(2023), https://doi.org/10.1016/j.bspc.2022.104368.

[92] Y.-S. Su, T.-J. Ding, M.-Y. Chen, Deep learning methods in internet of medical

things for valvular heart disease screening system, IEEE Internet Things J. 8

(2021) 16921–16932.

[93] Y. Pan, M. Fu, B. Cheng, X. Tao, J. Guo, Enhanced deep learning assisted

convolutional neural network for heart disease prediction on the internet of

medical things platform, IEEE Access 8 (2020) 189503–189512.

[94] G. Rajkumar, T. Gayathri Devi, A. Srinivasan, Heart disease prediction using IoT

based framework and improved deep learning approach: medical application,

Med. Eng. Phys. 111 (2023), https://doi.org/10.1016/j.

medengphy.2022.103937.

[95] P.R.N. Lalitha, S.V. Jinny, Internet of medical things-based multitiered and hybrid

architectural framework for effective heart disease prediction model, Concurr.

Comput. Pract. Exp. 34 (2022) e6953.

[96] S. Basak, K. Chatterjee, Smart healthcare surveillance system using IoT and

machine learning approaches for heart disease, in: Commun. Comput. Inf. Sci.,

2022, https://doi.org/10.1007/978-3-031-23092-9_24.

[97] A. Pati, M. Parhi, M.K. Al Alnabhan, M. Pattanayak, B.K. Habboush, A.

K. Nawayseh, An IoT-fog-cloud integrated framework for real-time remote

cardiovascular disease diagnosis, Informatics 10 (2023).

[98] S. Shrivastava, T.Q. Trung, N.E. Lee, Recent progress, challenges, and prospects of

fully integrated mobile and wearable point-of-care testing systems for self-testing,

Chem. Soc. Rev. 49 (2020) 1812–1866, https://doi.org/10.1039/C9CS00319C.

[99] S. Moin, A. Karim, Z. Safdar, K. Safdar, E. Ahmed, M. Imran, Securing IoTs in

distributed blockchain: analysis, requirements and open issues, Futur. Gener.

Comput. Syst. 100 (2019) 325–343, https://doi.org/10.1016/J.

FUTURE.2019.05.023.

[100] M.K. Hasan, S. Islam, R. Sulaiman, S. Khan, A.H.A. Hashim, S. Habib, M. Islam,

S. Alyahya, M.M. Ahmed, S. Kamil, M.A. Hassan, Lightweight encryption

technique to enhance medical image security on Internet of Medical Things

S.F. Ahmed et al.

Information Fusion 102 (2024) 102060
20
applications, IEEE Access 9 (2021) 47731–47742, https://doi.org/10.1109/ 
ACCESS.2021.3061710. 
[101] A. Avinashiappan, B. Mayilsamy, Internet of Medical Things: security threats, 
security challenges, and potential solutions, Internet of Thing.  (2021) 1–16, 
https://doi.org/10.1007/978-3-030-63937-2_1/COVER. 
[102] W. Raque, L. Qi, I. Yaqoob, M. Imran, R.U. Rasool, W. Dou, Complementing IoT 
services through software dened networking and edge computing:  a 
comprehensive survey, IEEE Commun. Surv. Tutorials. 22 (2020)  1761–1804, 
https://doi.org/10.1109/COMST.2020.2997475. 
[103] G. Hatzivasilis, O. Soultatos,  S. Ioannidis, C. Verikoukis, G. Demetriou, 
C. Tsatsoulis, Review of  security and privacy for  the internet of medical  things 
(IoMT): resolving the protection  concerns for the novel  circular economy 
bioinformatics, in: Proc. - 15th Annu. Int. Conf.  Distrib. Comput. Sens. Syst. 
DCOSS 2019, 2019, https://doi.org/10.1109/DCOSS.2019.00091. 
[104] N. Alsaeed, F. Nadeem,  Authentication in the Internet of Medical Things: 
taxonomy, review, and open issues, Appl. Sci. 12  (2022) 7487, https://doi.org/ 
10.3390/APP12157487. 
[105] A. Arfaoui, A. Kribeche, S.M. Senouci, Context-aware anonymous authentication 
protocols in the internet  of things dedicated to  e-health applications, Comput. 
Netw. 159 (2019) 23–36, https://doi.org/10.1016/J.COMNET.2019.04.031. 
[106] U. Chatterjee, D. Sadhukhan, S. Ray, An improved authentication and key 
agreement protocol for smart healthcare system in the  context of internet of 
things using elliptic curve cryptography, Lect. Notes Netw.  Syst. 116 (2020) 
11–22, https://doi.org/10.1007/978-981-15-3020-3_2/COVER. 
[107] G. Xu, F. Wang, M. Zhang, J. Peng, Efcient and provably secure anonymous user 
authentication scheme for patient monitoring using wireless medical  sensor 
networks, IEEE Access 8 (2020) 47282–47294, https://doi.org/10.1109/ 
ACCESS.2020.2978891. 
[108] M.F.A Minahil, K. Mahmood, S. Kumari, A.K. Sangaiah, Lightweight 
authentication protocol for e-health clouds in IoT-based applications through 5 G 
technology, Digit. Commun. Netw. 7 (2021) 235–244, https://doi.org/10.1016/J. 
DCAN.2020.06.003. 
[109] S.S. Sahoo, S. Mohanty,  B. Majhi, A secure three factor based authentication 
scheme for health care systems using IoT enabled  devices, J. Ambient Intell. 
Humaniz. Comput. 12 (2021) 1419–1434, https://doi.org/10.1007/S12652-020- 
02213-6/TABLES/10. 
[110] H. Khalid, S.J. Hashim,  S.M.S. Ahmad, F. Hashim, M.A. Chaudhary, Cross-SN: a 
lightweight authentication scheme for  a multi-server platform using  IoT-based 
wireless medical sensor network,  Electron 10 (2021) 790,  https://doi.org/ 
10.3390/ELECTRONICS10070790. 
[111] Y. Zhang, R. Gravina, H. Lu, M. Villari, G. Fortino, PEA: parallel 
electrocardiogram-based authentication for smart healthcare systems, J. Netw. 
Comput. Appl. 117 (2018) 10–16, https://doi.org/10.1016/J.JNCA.2018.05.007. 
[112] M. Fotouhi, M. Bayat,  A.K. Das, H.A.N. Far, S.M. Pournaghi, M.A. Doostari, 
A lightweight and secure two-factor authentication scheme for wireless body area 
networks in health-care IoT, Comput. Netw. 177 (2020), 107333, https://doi.org/ 
10.1016/J.COMNET.2020.107333. 
[113] R. Hajian, S. ZakeriKia,  S.H. Erfani, M. Mirabi, SHAPARAK: scalable healthcare 
authentication protocol with attack-resilience and anonymous key-agreement, 
Comput. Netw. 183 (2020), 107567, https://doi.org/10.1016/J. 
COMNET.2020.107567. 
[114] A.K. Das, A secure  and robust temporal credential-based three-factor user 
authentication scheme for wireless sensor networks, Peer-to-Peer Netw.  Appl. 9 
(2016) 223–244, https://doi.org/10.1007/S12083-014-0324-9/TABLES/7. 
[115] R. Kumar, R. Tripathi,  Towards design and implementation of security and 
privacy framework for Internet of Medical Things (IoMT)  by leveraging 
blockchain and IPFS technology, J. Supercomput. 77 (2021) 7916–7955, https:// 
doi.org/10.1007/S11227-020-03570-X/TABLES/3. 
[116] V. Sureshkumar, R. Amin,  V.R. Vijaykumar, S.R. Sekar, Robust secure 
communication protocol for smart healthcare system with FPGA implementation, 
Futur. Gener. Comput. Syst.  100 (2019) 938–951, https://doi.org/10.1016/J. 
FUTURE.2019.05.058. 
[117] P. Bhuarya, P. Chandrakar, R. Ali, A. Sharaff, An enhanced authentication scheme 
for Internet of Things  and cloud based on  elliptic curve cryptography, Int.  J. 
Commun. Syst. 34 (2021) e4834, https://doi.org/10.1002/DAC.4834. 
[118] A.K. Das, A.K. Sutrala, V. Odelu, A. Goswami, A secure smartcard-based 
anonymous user authentication scheme for healthcare applications using wireless 
medical sensor networks, Wirel. Pers. Commun. 94 (2017)  1899–1933, https:// 
doi.org/10.1007/S11277-016-3718-6/TABLES/9. 
[119] M. Tahir, M. Sardaraz, S. Muhammad, M.S. Khan, A lightweight authentication 
and authorization framework for  blockchain-enabled IoT network in  health- 
informatics, Sustain. 12 (2020)  6960, https://doi.org/10.3390/SU12176960. 
[120] F. Wu, X. Li, A.K. Sangaiah, L. Xu, S. Kumari, L. Wu, J.  Shen, A lightweight and 
robust two-factor authentication scheme for personalized healthcare systems 
using wireless medical sensor  networks, Futur. Gener. Comput.  Syst. 82 (2018) 
727–737, https://doi.org/10.1016/J.FUTURE.2017.08.042. 
[121] A. Alrawais, A. Alhothaily,  C. Hu, X. Cheng, Fog computing for the Internet of 
Things: security and privacy issues, IEEE Internet Comput.  21 (2017) 34–42, 
https://doi.org/10.1109/MIC.2017.37. 
[122] I. Makhdoom, M. Abolhasan, J. Lipman, R.P. Liu, W. Ni, Anatomy of threats to the 
Internet of Things, IEEE  Commun. Surv. Tutor. 21  (2019) 1636–1675, https:// 
doi.org/10.1109/COMST.2018.2874978. 
[123] C. Camara, P. Peris-Lopez,  J.E. Tapiador, Security and privacy issues in 
implantable medical devices: a comprehensive survey, J. Biomed.  Inform. 55 
(2015) 272–289, https://doi.org/10.1016/J.JBI.2015.04.007. 
[124] R. Xu, Y. Chen, E. Blasch, G. Chen, BlendCAC: a smart contract enabled 
decentralized capability-based access control  mechanism for the IoT,  Comput 7 
(2018) 39, https://doi.org/10.3390/COMPUTERS7030039. 
[125] S. Gusmeroli, S. Piccione,  D. Rotondi, A capability-based security approach to 
manage access control in the Internet of Things, Math. Comput. Model. 58 (2013) 
1189–1205, https://doi.org/10.1016/J.MCM.2013.02.006. 
[126] S.Y. Lim, P.T. Fotsing, A. Almasri, O. Musa, M.L.M. Kiah, T.F. Ang, R.  Ismail, 
Blockchain technology the identity  management and authentication service 
disruptor: a survey, Int.  J. Adv. Sci. Eng.  Inf. Technol. 8 (2018) 1735–1745, 
https://doi.org/10.18517/IJASEIT.8.4-2.6838. 
[127] R.M. Aileni, G. Suciu, IoMT: a blockchain perspective, Stud. Big Data. 71 (2020) 
199–215, https://doi.org/10.1007/978-3-030-38677-1_9/COVER. 
[128] F. Alsubaei, A. Abuhussein, S. Shiva, Security and privacy in the Internet of 
Medical Things: taxonomy and risk assessment, in: Proc. - 2017 IEEE 42nd Conf. 
Local Comput. Networks Work.  LCN Work, 2017, pp.  112–120, https://doi.org/ 
10.1109/LCN.WORKSHOPS.2017.72, 2017. 
[129] O.N. Akande, O.C. Abikoye, A.A. Kayode, O.T. Aro, O.R. Ogundokun, A dynamic 
round triple data encryption standard  cryptographic technique for data security, 
in: Lect. Notes Comput. Sci. (Including Subser. Lect. Notes Artif. Intell. Lect. Notes 
Bioinformatics), 12254 LNCS, 2020, pp. 487–499, https://doi.org/10.1007/978- 
3-030-58817-5_36/COVER. 
[130] T. Li, H. Li,  L. Hu, H. Li, A reversible steganography method with statistical 
features maintained based on the difference value, IEEE  Access 8 (2020) 
12845–12855, https://doi.org/10.1109/ACCESS.2020.2964830. 
[131] J. Ni, K. Zhang, X.  Lin, X.S. Shen, Securing fog computing for Internet of Things 
applications: challenges and solutions, IEEE Commun. Surv. Tutorials. 20 (2018) 
601–628, https://doi.org/10.1109/COMST.2017.2762345. 
[132] D.R. Raymond, S.F. Midkiff, Denial-of-service in wireless sensor networks: attacks 
and defenses, IEEE Pervas.  Comput. 7 (2008) 74–81,  https://doi.org/10.1109/ 
MPRV.2008.6. 
[133] T. Borgohain, U. Kumar,  S. Sanyal, Survey of  Security and Privacy Issues of 
Internet of Things, 2015. 
[134] W. Xu, T. Wood, W. Trappe, Y. Zhang, Channel Surng and Spatial Retreats, 
2004, pp. 80–89, https://doi.org/10.1145/1023646.1023661. 
[135] A.D. Wood, J.A. Stankovic, S.H. Son, JAM: a Jammed-area mapping service for 
sensor networks, in: Proc.  - Real-Time Syst. Symp,  2003, pp. 286–297, https:// 
doi.org/10.1109/REAL.2003.1253275. 
[136] Y. Xuan, Y. Shen, N.P. Nguyen, M.T. Thai, A trigger identication service for 
defending reactive jammers in  WSN, IEEE Trans. Mob.  Comput. 11 (2012) 
793–806, https://doi.org/10.1109/TMC.2011.86. 
[137] S. Sciancalepore, G. Oligeri, R. Di Pietro, Strength of crowd (SOC)—defeating a 
reactive jammer in IoT with decoy messages, Sensors 18 (2018) 3492, https://doi. 
org/10.3390/S18103492. 
[138] J. Liu, W. Sun,  Smart attacks against intelligent wearables in people-centric 
Internet of Things, IEEE  Commun. Mag. 54 (2016)  44–49, https://doi.org/ 
10.1109/MCOM.2016.1600553CM. 
[139] A. Mubashar, K. Asghar, A.R. Javed, M. Rizwan, G. Srivastava, T.R. Gadekallu, 
D. Wang, M. Shabbir,  Storage and Proximity Management  for Centralized 
Personal Health Records Using an IPFS-Based Optimization Algorithm, 31, 2021, 
https://doi.org/10.1142/S0218126622500104. 
[140] X. Nie, A. Zhang, J. Chen, Y. Qu, S. Yu, Blockchain-empowered secure and 
privacy-preserving health data sharing  in edge-based IoMT, Secur.  Commun. 
Networks. 2022 (2022), https://doi.org/10.1155/2022/8293716. 
[141] J. Deogirikar, A. Vidhate, Security attacks in IoT: a survey, in: Proc. Int. Conf. IoT 
Soc. Mobile, Anal. Cloud,  I-SMAC 2017, 2017, pp.  32–37, https://doi.org/ 
10.1109/I-SMAC.2017.8058363. 
[142] H. Ning, H. Liu, L.T. Yang, Cyberentity security in the internet of things, Comput. 
(Long. Beach. Calif 46  (2013) 46–53, https://doi.org/10.1109/MC.2013.74. 
[143] M. Seliem, K. Elgazzar, BIoMT: blockchain for the internet of medical things, in: 
2019 IEEE Int. Black Sea Conf. Commun. Networking, BlackSeaCom 2019, 2019, 
https://doi.org/10.1109/BLACKSEACOM.2019.8812784. 
[144] M. Mettler, Blockchain technology in healthcare: the revolution starts here, in: 
2016 IEEE 18th Int. Conf. e-Health Networking, Appl. Serv. Heal, 2016, https:// 
doi.org/10.1109/HEALTHCOM.2016.7749510, 2016. 
[145] R.M.P.H.K. Rathnayake, M.S. Karunarathne,  N.S. Na, M.A. Gregory, Cloud 
Enabled Solution for Privacy  Concerns in Internet of  Medical Things, in: 2018 
28th Int. Telecommun. Networks Appl. Conf. ITNAC 2018, 2019, https://doi.org/ 
10.1109/ATNAC.2018.8615361. 
[146] A. Lounis, A. Hadjidj, A. Bouabdallah, Y. Challal, Secure and scalable cloud-based 
architecture for e-Health wireless  sensor networks, in: 2012  21st Int. Conf. 
Comput. Commun, Networks, 2012,  https://doi.org/10.1109/ 
ICCCN.2012.6289252. 
[147] K. Lorincz, D.J. Malan, T.R.F. Fulford-Jones, A. Nawoj, A. Clavel, V. Shnayder, 
G. Mainland, M. Welsh,  S. Moulton, Sensor networks  for emergency response: 
challenges and opportunities, IEEE Pervas. Comput. 3 (2004) 16–23, https://doi. 
org/10.1109/MPRV.2004.18. 
[148] J.P.A. Yaacoub, M. Noura, H.N. Noura, O. Salman, E. Yaacoub, R. Couturier, 
A. Chehab, Securing internet  of medical things systems:  limitations, issues and 
recommendations, Futur. Gener. Comput.  Syst. 105 (2020), https://doi.org/ 
10.1016/j.future.2019.12.028. 
[149] S. Ibrokhimov, K.L. Hui, A.A. Al-Absi, M. Sain,  Multi-factor authentication in 
cyber physical system: a  state of art survey,  in: 2019 21st Int. Conf. Adv. 
Commun. Technol., IEEE, 2019,  pp. 279–284. 
S.F. Ahmed et al.                                                                                                                                                                                                                                

Cybersecurity for Sustainable Smart Healthcare: State of the Art, Taxonomy, Mechanisms, and Essential Roles

Article

Full-text available

May 2024

Cutting-edge technologies have been widely employed in healthcare delivery, resulting in transformative advances and promising enhanced patient care, operational efficiency, and resource usage. However, the proliferation of networked devices and data-driven systems has created new cybersecurity threats that jeopardize the integrity, confidentiality, and availability of critical healthcare data. This review paper offers a comprehensive evaluation of the current state of cybersecurity in the context of smart healthcare, presenting a structured taxonomy of its existing cyber threats, mechanisms and essential roles. This study explored cybersecurity and smart healthcare systems (SHSs). It identified and discussed the most pressing cyber threats and attacks that SHSs face, including fake base stations, medjacking, and Sybil attacks. This study examined the security measures deployed to combat cyber threats and attacks in SHSs. These measures include cryptographic-based techniques, digital watermarking, digital steganography, and many others. Patient data protection, the prevention of data breaches, and the maintenance of SHS integrity and availability are some of the roles of cybersecurity in ensuring sustainable smart healthcare. The long-term viability of smart healthcare depends on the constant assessment of cyber risks that harm healthcare providers, patients, and professionals. This review aims to inform policymakers, healthcare practitioners, and technology stakeholders about the critical imperatives and best practices for fostering a secure and resilient smart healthcare ecosystem by synthesizing insights from multidisciplinary perspectives, such as cybersecurity, healthcare management, and sustainability research. Understanding the most recent cybersecurity measures is critical for controlling escalating cyber threats and attacks on SHSs and networks and encouraging intelligent healthcare delivery.

The Internet of Medical Things (IoMT): opportunities and challenges

Article

Full-text available

May 2024
WIREL NETW

The Internet of Medical Things (IoMT) is a transformative concept in healthcare, leveraging the power of connected devices and technology to improve patient care and health outcomes. These devices are typically connected to the internet or a network and can communicate with each other to exchange data and provide insights to healthcare providers, patients, and other stakeholders. Advanced digital technologies like artificial intelligence, machine learning, and Blockchain integrated into the Internet of Medical Things (IoMT) can better mitigate the impact of pandemics and protect public health. IoMT utilizes medical sensors to capture real-time physiological data of patients and is available to a medical professional to diagnose, recognize, analyze, and make appropriate decisions. Data breaches or cyberattacks could compromise the security and integrity of IoMT devices and data, exposing sensitive information or causing malfunctions or disruptions. Consequently, to make IoMT systems reliable, data protection and secure communication must conform to security standards. Blockchain technology is being used in the healthcare industry to ensure the security of patient records and to streamline the sharing of information among healthcare providers, laboratories, pharmaceutical firms, and other healthcare providers. Overall, digital technologies have been instrumental in managing the COVID-19 pandemic, enabling more efficient surveillance, response, and care delivery. By leveraging these technologies, public health authorities and healthcare providers have been able to better mitigate the impact of the pandemic and protect public health.

An Interval-valued Spherical Fuzzy CIMAS-WISP Group Decision-analytic Model for Blockchain Platform Selection in Digital Projects

Article

Sep 2024
APPL SOFT COMPUT

Digital projects aspiring to reach target audiences are executed through decentralized and trustworthy blockchain platforms (BPs). Once the objectives and target audience of a digital project are defined, the selection of suitable BPs is undertaken. The primary objective of this research is to develop a decision support system that aids in the selection of BPs for transferring digital data and assets. Numerous quantitative parameters determine the performance of BPs, alongside qualitative parameters indicating their performance. In this study, the aim is to determine the performance of BPs based on both qualitative and quantitative parameters. Within this scope, a multi-criteria decision-making approach and interval-valued spherical fuzzy (IVSF) sets are adopted. IVSF sets are utilized to determine expert importance levels. The IVSF-criteria importance assessment (CIMAS) method is introduced for the weighting of criteria. IVSF-CIMAS enables the determination of reliability levels in calculating criterion weights. The IVSF-simple weighted sum product (WISP) method is formulated to obtain the performance ranking of BPs. Thus, in this research, the IVSF-CIMAS-WISP hybrid model is developed, and an algorithm for this novel decision-analytic model is presented. A case study is developed focusing on BP selection for a digital project to demonstrate the applicability of the proposed hybrid model. The robustness of IVSF-CIMAS-WISP is tested through extensive sensitivity analysis scenarios. According to the research results, the applicability of the IVSF-CIMAS-WISP hybrid method is supported and its robustness is confirmed. The research findings provide numerous insights for project managers and practitioners.

Trusted Composition of Internet of Medical Things Over Imperfect Networks

Preprint

Full-text available

May 2024

The Internet of Medical Things (IoMT) represents a specialized domain within the Internet of Things, focusing on medical devices that require regulatory approval to ensure patient safety. Trusted composition of IoMT systems aims to ensure high assurance of the entire composed system, despite potential variability in the assurance levels of individual components. Achieving this trustworthiness in IoMT systems, especially when using less-assured, commercial, off-the-shelf networks like Ethernet and WiFi, presents a significant challenge. To address this challenge, this paper advocates a systematic approach that leverages the Architecture Analysis & Design Language (AADL) along with Behavior Language for Embedded Systems with Software (BLESS) specification and behavior. This approach aims to provide high assurance on critical components through formal verification, while using less-assured components in a manner that maintains overall system determinism and reliability. A clinical case study involving an automated opioid infusion monitoring IoMT system is presented to illustrate the application of the proposed approach. Through this case study, the effectiveness of the systemic approach in achieving trusted composition of heterogeneous medical devices over less-assured networks is demonstrated.

IoMT Landscape: Navigating Current Challenges and Pioneering Future Research Trends

Preprint

Full-text available

May 2024

Smart electronic devices and telemedicine are widely used in humans' daily lives. This provides remote treatment of patients using information and communications technology. New telemedicine technologies, such as the Internet of Medical Things (IoMT), make it easier for medical and computing devices to communicate regularly and effectively. Critical motivations for adapting the IoMT are reduced cost, increased quality of life, and timely medical intervention. IoMT is significant because it enables continuous, real-time patient monitoring during routine everyday activities using a variety of wearables and sensors. With big data, the IoMT technology makes excellent use of Artificial Intelligence (AI) to support disease detection and health condition prediction, alerting patients and healthcare providers. Many research studies have been conducted to explore various aspects of IoMT and its applications in the real world. However, it is challenging to comprehend all the techniques and solutions proposed by the research community. Therefore, this survey sheds light on some crucial aspects of IoMT technologies and explores the potential research gaps and directions the research community could tackle. The survey examines and discusses the characteristics of IoMT standards, protocols, and types. It then delves into the layers of IoMT and distinguishes them into Fog and edge. The studies published under each type were explored, and the limitations of these works were highlighted. The research gaps and directions on IoMT approaches and technologies were also highlighted. With such findings and research directions, further research endeavors could be carried out to address the issues and existing limitations in the IoMT.

Cyber Security Risk Assessment Methods for Smart Healthcare

Article

Full-text available

Jun 2024

The perverseness of the Internet of Things (IoT) has reached the healthcare sector, where interconnected medical devices are transforming how we deliver and manage health. These devices, linked wirelessly, create a vast network known as Internet of Medical Things (IoMT), that integrates seamlessly with the broader healthcare system. This interconnected infrastructure facilitates the exchange of massive amounts of patient data, paving the way for a more distributed and data-driven approach to healthcare. In healthcare’s ICT environment rely heavily on electronic health records (EHR), e-prescribing systems and other systems. Protecting this sensitive data necessitates robust cybersecurity measures. Even though data security is an indirect cost, it is crucial for healthcare systems. In general, patient trust in a healthcare system depends on the infrastructure’s ability to protect data from security and privacy threats. A critical component of this strategy is security risk assessment. A risk is an indirect cost borne by those designing these systems. The risk assessment process identifies, evaluates, and prioritizes potential threats to the organization’s assets including hardware and software. By assessing risks before acting, healthcare organizations can effectively allocate resources to mitigate the most critical vulnerabilities. While broadly applicable cyber risk assessment frameworks like NIST, ISO, and OCTAVE exist, they may lack a strong risk between assets, threats and impacts, and controls. Therefore, they do not provide a comprehensive picture for healthcare specifically. A more thorough examination is required to establish a strong association between protected systems, potential threats, and the resulting risks. This paper critically reviews these frameworks and their methodologies and limitations for healthcare cybersecurity. Even though data security is an indirect cost, it is crucial for healthcare systems.

Securing and Sustaining the Internet of Medical Things (IoMT): Needs, Design Hurdles, Security Methods, and Future Directions

Chapter

May 2024

Recent advances in M2M communications, miniaturized hardware, and microcomputing have spurred the rise of IoT. In healthcare, IoT integration with medical devices has transformed treatment, offering cost-effective solutions, improved monitoring, and personalized care. This has led to complex IoMT systems with unique operating systems and protocols. However, security vulnerabilities threaten their adoption. Research is addressing these challenges by integrating robust security measures. This paper emphasizes the need for security in IoMT, discussing key aspects and state-of-the-art techniques for securing data transmission, collection, and storage. It explores various security requirements, design complexities, and methodologies to enhance IoMT resilience. Additionally, it surveys current research and suggests future directions for innovation.

Navigating the Future of Healthcare: A User-Centric Approach to Designing Lucrative Business Models for the IoMT

Chapter

May 2024

Rita Komalasari

The internet of medical things (IoMT) is a rapidly evolving technology that is set to revolutionize patient care, diagnosis, and monitoring. However, its success relies on the strategic design of user-centric business models. This research explores the relationship between technological innovation and business model design in IoMT, focusing on user perspectives and expectations. Findings reveal that trust-building mechanisms and subscription services are pivotal factors in IoMT adoption, demonstrating the importance of integrating successful digital technology business models to close the gap between technical progress and user expectations.

IoMT Tsukamoto Type-2 fuzzy expert system for tuberculosis and Alzheimer’s disease

Article

May 2024

Accurate disease monitoring is an extremely time-consuming task for medical experts and technocrats involved, requiring technical support for diagnostic systems. To overcome this situation, we developed an Internet of Medical Things (IoMT) based on Tsukamoto Type 2 Fuzzy Inference System (TT2FIS) that can easily handle diagnostic and predictive aspects in the medical field. In the proposed system, we developed a Tsukamoto type 2 fuzzy inference system that takes the patient’s symptoms as input factors and the medical device as the output factor of the result. The aim of this work is to demonstrate the usefulness of type 2 fuzzy sets in Tuberculosis and Alzheimer’s disease diagnostic system. Numerical calculations are also performed to illustrate the applicability of the proposed method. A validation of the proposed derivation of the proposed IoMT model is also discussed in the results and conclusions section.

Mechanics 4.0 and Mechanical Engineering Education

Article

Full-text available

May 2024

Industry 4.0 is an industrial paradigm that is causing changes in form and substance in factories, companies and businesses around the world and is impacting work and education in general. In fact, the disruptive technologies that frame the Fourth Industrial Revolution have the potential to improve and optimize manufacturing processes and the entire value chain, which could lead to an exponential evolution in the production and distribution of goods and services. All these changes imply that the fields of engineering knowledge must be oriented towards the concept of Industry 4.0, for example, Mechanical Engineering. The development of various physical assets that are used by cyber-physical systems and digital twins is based on mechanics. However, the specialized literature on Industry 4.0 says little about the importance of mechanics in the new industrial era, and more importance is placed on the evolution of Information and Communication Technologies and artificial intelligence. This article presents a frame of reference for the importance of Mechanical Engineering in Industry 4.0 and proposes an extension to the concept of Mechanics 4.0, recently defined as the relationship between mechanics and artificial intelligence. To analyze Mechanical Engineering in Industry 4.0, the criteria of the four driving forces that defined mechanics in the Third Industrial Revolution were used. An analysis of Mechanical Engineering Education in Industry 4.0 is presented, and the concept of Mechanical Engineering 4.0 Education is improved. Finally, the importance of making changes to the educational models of engineering education is described.

Recent Advances in Deep Learning and Medical Imaging for Head and Neck Cancer Treatment: MRI, CT, and PET Scans

Article

Full-text available

Jun 2023

Simple Summary Deep learning techniques have significant potential in head and neck cancer imaging, particularly in tumor detection, segmentation, and outcome prediction using magnetic resonance imaging (MRI), computed tomography (CT), and positron emission tomography (PET) scans. Advanced deep learning methods, such as convolutional autoencoders, generative adversarial networks (GANs), and transformer models, have further enhanced imaging capabilities. Comparing deep learning and traditional techniques and the advantages and limits of each reveals their complementary roles in cancer management. Integrating radiogenomics with deep learning models promises further advancements in personalized care. However, challenges such as standardization, data quality, model overfitting, and computational requirements persist. Addressing these issues, integrating multimodal and temporal information, enhancing explainability, and conducting clinical validation are crucial for implementing deep learning models in head and neck cancer diagnosis and treatment. Overcoming these obstacles will pave the way for improved patient outcomes and personalized treatment strategies in head and neck cancer management. Abstract Deep learning techniques have been developed for analyzing head and neck cancer imaging. This review covers deep learning applications in cancer imaging, emphasizing tumor detection, segmentation, classification, and response prediction. In particular, advanced deep learning techniques, such as convolutional autoencoders, generative adversarial networks (GANs), and transformer models, as well as the limitations of traditional imaging and the complementary roles of deep learning and traditional techniques in cancer management are discussed. Integration of radiomics, radiogenomics, and deep learning enables predictive models that aid in clinical decision-making. Challenges include standardization, algorithm interpretability, and clinical validation. Key gaps and controversies involve model generalizability across different imaging modalities and tumor types and the role of human expertise in the AI era. This review seeks to encourage advancements in deep learning applications for head and neck cancer management, ultimately enhancing patient care and outcomes.

Towards a Secure and Sustainable Internet of Medical Things (IoMT): Requirements, Design Challenges, Security Techniques, and Future Trends

Article

Full-text available

Apr 2023

Recent advances in machine-to-machine (M2M) communications, mini-hardware manufacturing, and micro computing have led to the development of the Internet of Things (IoT). The IoT is integrated with medical devices in order to enable better treatment, cost-effective medical solutions, improved patient monitoring, and enhanced personalized healthcare. This has led to the development of more complex and heterogeneous Internet of Medical Things (IoMT) systems that have their own operating systems and protocols. Even though such pervasive and low-cost sensing devices can bring about enormous changes in the healthcare sector, these are prone to numerous security and privacy issues. Security is thus a major challenge in these critical systems, one that inhibits their widespread adoption. However, significant inroads have been made by the on-going research, which powers the IoMT applications by incorporating prevalent security measures. In this regard, this paper highlights the significance of implementing key security measures, and essential aspects of the IoMT that make it useful for interconnecting various internal and external working domains of healthcare. This paper presents state-of-the-art techniques for securing IoMT systems, in terms of data transmission, collection, and storage. Furthermore, the paper also explores various security requirements, inherent design challenges, and various security techniques that could make the IoMT more secure and sustainable. Finally, the paper gives a panoramic view of the current status of research in the field and outlines some future research directions in this area.

Risk Assessment of Heterogeneous IoMT Devices: A Review

Research

Full-text available

Feb 2023

The adaptation of the Internet of Medical Things (IoMT) has provided efficient and timely services and has transformed the healthcare industry to a great extent. Monitoring patients remotely and managing hospital records and data have become effortless with the advent of IoMT. However, security and privacy have become a significant concern with the growing number of threats in the cyber world, primarily for personal and sensitive user data. In terms of IoMT devices, risks appearing from them cannot easily fit into an existing risk assessment framework, and while research has been done on this topic, little attention has been paid to the methodologies used for the risk assessment of heterogeneous IoMT devices. This paper elucidates IoT, its applications with reference to in-demand sectors, and risks in terms of their types. By the same token, IoMT and its application area and architecture are explained. We have also discussed the common attacks on IoMT. Existing papers on IoT, IoMT, risk assessment, and frameworks are reviewed. Finally, the paper analyzes the available risk assessment frameworks such as NIST, ISO 27001, TARA, and the IEEE213-2019 (P2413) standard and highlights the need for new approaches to address the heterogeneity of the risks. In our study, we have decided to follow the functions of the NIST and ISO 270001 frameworks. The complete framework is anticipated to deliver a risk-free approach for the risk assessment of heterogeneous IoMT devices benefiting its users.

Internet of Medical Things Privacy and Security: Challenges, Solutions, and Future Trends from a New Perspective

Article

Full-text available

Feb 2023

The Internet of Medical Things (IoMT), an application of the Internet of Things (IoT) in the medical domain, allows data to be transmitted across communication networks. In particular, IoMT can help improve the quality of life of citizens and older people by monitoring and managing the body's vital signs, including blood pressure, temperature, heart rate, and others. Since IoMT has become the main platform for information exchange and making high-level decisions, it is necessary to guarantee its reliability and security. The growth of IoMT in recent decades has attracted the interest of many experts. This study provides an in-depth analysis of IoT and IoMT by focusing on security concerns from different points of view, making this comprehensive survey unique compared to other existing studies. A total of 187 articles from 2010 to 2022 are collected and categorized according to the type of applications, year of publications, variety of applications, and other novel perspectives. We compare the current studies based on the above criteria and provide a comprehensive analysis to pave the way for researchers working in this area. In addition, we highlight the trends and future work. We have found that blockchain, as a key technology, has solved many problems of security, authentication, and maintenance of IoT systems due to the decentralized nature of the blockchain. In the current study, this technology is examined from the application fields' points of view, especially in the health sector, due to its additional importance compared to other fields.

An IoT-Fog-Cloud Integrated Framework for Real-Time Remote Cardiovascular Disease Diagnosis

Article

Full-text available

Feb 2023

Recently, it has proven difficult to make an immediate remote diagnosis of any coronary illness, including heart disease, diabetes, etc. The drawbacks of cloud computing infrastructures, such as excessive latency, bandwidth, energy consumption, security, and privacy concerns, have lately been addressed by Fog computing with IoT applications. In this study, an IoT-Fog-Cloud integrated system, called a Fog-empowered framework for real-time analysis in heart patients using ENsemble Deep learning (FRIEND), has been introduced that can instantaneously facilitate remote diagnosis of heart patients. The proposed system was trained on the combined dataset of Long-Beach, Cleveland, Switzerland, and Hungarian heart disease datasets. We first tested the model with eight basic ML approaches, including the decision tree, logistic regression, random forest, naive Bayes, k-nearest neighbors, support vector machine, AdaBoost, and XGBoost approaches, and then applied ensemble methods including bagging classifiers, weighted averaging, and soft and hard voting to achieve enhanced outcomes and a deep neural network, a deep learning approach, with the ensemble methods. These models were validated using 16 performance and 9 network parameters to justify this work. The accuracy, PPV, TPR, TNR, and F1 scores of the experiments reached 94.27%, 97.59%, 96.09%, 75.44%, and 96.83%, respectively, which were comparatively higher when the deep neural network was assembled with bagging and hard-voting classifiers. The user-friendliness and the inclusion of Fog computing principles, instantaneous remote cardiac patient diagnosis, low latency, and low energy consumption, etc., are advantages confirmed according to the achieved experimental results.

A privacy-aware framework for detecting cyber attacks on internet of medical things systems using data fusion and quantum deep learning

Article

Jun 2023
INFORM FUSION

Internet of Medical Things (IoMT) devices and systems are often designed without adequate security, leaving them highly susceptible to cyber threats. Unlike other IoT applications and devices, cyber attacks against IoMT devices will pose significant risks to patient safety. As a result of the progress made in attack detection through machine learning, numerous applications of machine learning are now involved in IoMT. However, how to securely collect and fuse the data from heterogeneous IoMT devices and systems and ensure efficient cyber-attack detection while avoiding the privacy disclosure of the dataset remains challenging for IoMT systems. To fill this gap, a new privacy-aware framework is advised for storing the collected IoMT data in distributed cloud nodes and providing data fusion of heterogeneous IoMT data using differential privacy and deep learning and efficient attack detection using quantum deep learning with maintaining data privacy. Specifically, the proposed privacy-aware framework provides sensitive data privacy protection at multi-levels starting from storing data in distributed cloud nodes with privacy-preserving and then fusing IoMT heterogeneous data using deep contractive autoencoder with differential privacy for protecting the sensitive data through the learning process and then using the output of data fusion which is reduced data dimension as input for a quantum neural network for identifying cyber-attacks. Our experimental results indicate that the proposed framework is highly effective in detecting cyber-attacks.

Survey of Machine Learning based intrusion detection methods for Internet of Medical Things

Article

Mar 2023
APPL SOFT COMPUT

Internet of Things for Diagnosis of Alzheimer’s Disease: A Multimodal Machine Learning Approach Based on Eye Movement Features

Article

Jul 2023

Alzheimer’s disease (AD) is a degenerative neurological disease that occurs in the elderly with typical symptoms of decline in cognition, manifested by eye movement behaviors. The key to AD treatment requires early detection of cognitive impairment, which relies on frequent medical screening. This paper proposes an Internet of Things (IoT) architecture constructed with eye-tracker (ET) nodes and cloud-based diagnosis enabled by machine learning (ML), which can provide convenient screening of oculomotor abnormalities and automatic identification of early-stage AD. The bespoke ET nodes collect three-dimensional (3-D) oculomotor responses from diverse stereo video stimulation trials and transmit data into a dedicated multimodal machine learning (MMML) algorithm in the cloud. The algorithm incorporates multimodal features extracted from diverse oculomotor types to enhance the classification accuracy and optimized data dimension reduction in feature fusion to improve the classifier’s performance. From evaluation, the proposed method can distinguish AD patients from the control group with 86% accuracy (ACC), 78% true positive rate (TPR), and 90% positive predictive value (PPV). The results confirm the effectiveness of our MMML algorithm in AD diagnosis with the fusion of multimodal oculomotor features and prove the feasibility of our IoT-powered eye-tracking solution for AD screening.

Fed-ESD: Federated Learning for Efficient Epileptic Seizure Detection in the Fog-Assisted Internet of Medical Things

Article

Feb 2023
INFORM SCIENCES

Identifying digital biomarkers in actigraph based sequential motor activity data for assessment of depression: a model evaluating SVM in LSTM extracted feature space

Article

Feb 2023

This research puts forward a methodology for depression assessment using actigraph recordings of motor activity. High level features of motor activity are extracted using Long-Short Term Memory (LSTM) which are paired with statistical features to deliver valuable digital biomarkers. Overlapping sliding window is used to input sequences into LSTM to capture superior features in activity recordings. The predictive ability of these digital biomarkers is evaluated using Support Vector Machine (SVM). The hybrid framework is validated on benchmark, Depresjon dataset and achieves accuracy of 95.57%. Effectiveness of overlapping sliding window and statistical features is evaluated, and their significance is validated. It is validated that the concept of overlapping sliding window improves performance accuracy by 3.51% and the use of discriminative statistical features improves model performance by 1.08%. It is concluded that the proposed methodology based on feature extraction, statistical features and overlapping sliding window outperforms state-of-the-art techniques as well as baseline deep learning algorithms for depression detection.

Insights into Internet of Medical Things (IoMT): Data fusion, security issues and potential solutions

Recommended publications

Secure blockchain assisted Internet of Medical Things architecture for data fusion enabled cancer wo...

The Power of Internet of Things (IoT): Connecting the Dots with Cloud, Edge, and Fog Computing

Toward a Secure 5G-Enabled Internet of Things: A Survey on Requirements, Privacy, Security, Challeng...

Optimizing Trust and Security in Healthcare 4.0: Human Factors in Lightweight Secured IoMT Ecosystem...

Integration of AI, Digital Twin and Internet of Medical Things (IoMT) For Healthcare 5.0: A Bibliome...