FIGURE 1 - available via license: CC BY
Content may be subject to copyright.
| Schematic illustration of the network structures of deep learning algorithms. (A) Multi-layer perceptron model architecture. (B) Convolutional neural network (CNN) model architecture. (C) Network structure of three recurrent neural networks (RNN) models.
Source publication
With its low-cost, label-free and non-destructive features, Raman spectroscopy is becoming an attractive technique with high potential to discriminate the causative agent of bacterial infections and bacterial infections per se. However, it is challenging to achieve consistency and accuracy of Raman spectra from numerous bacterial species and phenot...
Context in source publication
Context 1
... learning methods included two models, CNN and RNN, in which MLP had a special CNN network structure while SimpleRNN and GRU were simplified LSTM model (Figure 1). In particular, CNN model consisted of one input layer, six convolutional layers with convolution kernel sizes of 5 * 1 and 3 * 1, three maximum pooling layers, one fully connected layer and a softmax output layer that achieved 15-dimensional outputs. ...
Similar publications
The quantitative assessment of the musical timbre in an audio record is still an open-ended issue. Evaluating the musical timbre allows not only to establish precise musical parameters but also the recognition, classification of musical instruments, and assessment of the musical quality of a sound record. In this paper, we present a minimum set of...
Citations
... The collection method of aerosol particles has been used for laboratory static FTIR spectral classification research [21]. By combining this method with a classification model, spectral features were extracted and classified to identify unknown biological particles [22,23]. The classification performance of the model was related to spectral features, which improved as the spectral features of the input model increased. ...
The rapid and sensitive detection of pathogenic and suspicious bioaerosols are essential for public health protection. The impact of pollen on the identification of bacterial species by Raman and Fourier-Transform Infrared (FTIR) spectra cannot be overlooked. The spectral features of the fourteen class samples were preprocessed and extracted by machine learning algorithms to serve as input data for training purposes. The two types of spectral data were classified using classification models. The partial least squares discriminant analysis (PLS-DA) model achieved classification accuracies of 78.57% and 92.85%, respectively. The Raman spectral data were accurately classified by the support vector machine (SVM) algorithm, with a 100% accuracy rate. The two spectra and their fusion data were correctly classified with 100% accuracy by the random forest (RF) algorithm. The spectral processed algorithms investigated provide an efficient method for eliminating the impact of pollen interference.
... However, little work on ensemble learning algorithms has been employed for SCRS analysis until recently, where Baladehi et al. combined EMC learning and Ramanome to rapidly identify and metabolically profile microalgal single cells and found that the accuracy of classifying species and states can be above 97%, further highlighting the promising potential of the technique for classification and identification of microorganisms (Heidari et al., 2021). It is interesting to note that in recent studies, the deep learning algorithm LSTM did not perform well as did CNN, and in some cases, it was not as good as classical algorithms, such as RF and KNN (Yu et al., 2021;Tang et al., 2022). However, in our study, the LSTM method was almost as accurate as the CNN method, with an accuracy of 92.07%. ...
Rapid and accurate identification of lactic acid bacteria (LAB) species would greatly improve the screening rate for functional LAB. Although many conventional and molecular methods have proven efficient and reliable, LAB identification using these methods has generally been slow and tedious. Single-cell Raman spectroscopy (SCRS) provides the phenotypic profile of a single cell and can be performed by Raman spectroscopy (which directly detects vibrations of chemical bonds through inelastic scattering by a laser light) using an individual live cell. Recently, owing to its affordability, non-invasiveness, and label-free features, the Ramanome has emerged as a potential technique for fast bacterial detection. Here, we established a reference Ramanome database consisting of SCRS data from 1,650 cells from nine LAB species/subspecies and conducted further analysis using machine learning approaches, which have high efficiency and accuracy. We chose the ensemble meta-classifier (EMC), which is suitable for solving multi-classification problems, to perform in-depth mining and analysis of the Ramanome data. To optimize the accuracy and efficiency of the machine learning algorithm, we compared nine classifiers: LDA, SVM, RF, XGBoost, KNN, PLS-DA, CNN, LSTM, and EMC. EMC achieved the highest average prediction accuracy of 97.3% for recognizing LAB at the species/subspecies level. In summary, Ramanomes, with the integration of EMC, have promising potential for fast LAB species/subspecies identification in laboratories and may thus be further developed and sharpened for the direct identification and prediction of LAB species from fermented food.
... For example, Wang et al. recently demonstrated an interesting SERS approach in combination with different machine learning methods for the rapid identification of Mycobacterium tuberculosis in sputum samples [24]. In general, there have been major recent advances in clinical SERS applications for the rapid and accurate detection of bacterial pathogens [25][26][27], whereby the studies confirm the great potential of integrating SERS techniques and machine learning algorithms for the rapid and reliable identification of various bacterial pathogens, but the authors always point out the still existing wide gap between basic research and the clinical application of SERS technology. Likewise, in the food industry [28][29][30], time is crucial for product safety and a long shelf life. ...
In a previous publication, we trained predictive models based on Raman bulk spectra of microorganisms placed on a silicon dioxide protected silver mirror slide to make predictions for new Raman spectra, unknown to the models, of microorganisms placed on a different substrate, namely stainless steel. Now we have combined large sections of this data and trained a convolutional neural network (CNN) to make predictions for single cell Raman spectra. We show that a database based on microbial bulk material is conditionally suited to make predictions for the same species in terms of single cells. Data of 13 different microorganisms (bacteria and yeasts) were used. Two of the 13 species could be identified 90% correctly and five other species 71%–88%. The six remaining species were correctly predicted by only 0%–49%. Especially stronger fluorescence in bulk material compared to single cells but also photodegradation of carotenoids are some effects that can complicate predictions for single cells based on bulk data. The results could be helpful in assessing universal Raman tools or databases.
... This study proposed two unsupervised ML techniques, Agglomerative Nesting (AGNES), and K-means Clustering were conducted for clustering investigation. Also, eight supervised ML techniques were approximated in terminology of bacterial detection through Raman spectra, which demonstrated that CNN reached the most acceptable accuracy (99.86%) including the highest area (0.9996) under the curve ROC [11]. ...
Surface-enhanced Raman spectroscopy (SERS) is a powerful technique for molecular sensing and has gained significant attention due to its high sensitivity and selectivity. SERS based on deep learning technology have been used in this study of materials, biological recognition, food safety, and intelligence. Deep learning techniques have shown tremendous potential in various scientific fields, including spectroscopy-based detection methodologies. In this study, we propose an effective deep learning approach for SERS detection using an artificial neural network (ANN). Then, the results are compared using two datasets batch1 and batch2. This study used Rhumamine 6G (R6G) as an aim molecule in this study. The experimental develops show the effectiveness of proposed ANN.
... Using SERS in combination with unsupervised and supervised machine learning algorithms, Tang et al., successfully identified 15 different bacterial pathogens isolated from clinical samples [129]. "Using this machine learning-assisted techniques, SERS was able to accurately identify bacterial pathogens at the general level with high specificity and sensitivity. ...
The enormous economic burden and increasing public health threats associated with pathogenic organisms have necessitated the development of rapid bacterial identification platforms. Surface-enhanced Raman spectroscopy (SERS) is considered one of the most promising pathogen diagnostic techniques. The method generates a unique and sensitive biomarker fingerprint that may be utilized for quick and multiplexed analysis on a portable platform. Herein, we have made a comprehensive review on the advances made in SERS. First, we briefly discussed the specific and non-specific capture of pathogens using the label-free method. Then, various label-based immunosensors and aptasensors that have been developed over the years for the detection of pathogenic organisms have been summarized. This review then focused on the combination of CRISPR/Cas assays and SERS techniques in the detection of pathogenic organisms. Next, we discussed recent applications of machine learning techniques in Raman signal analysis and their application in clinical medicine. Furthermore, various conditions that affect Raman signal effects and how to deal with such conditions were discussed. Finally, challenges, key points, and future outlooks for SERS application have been summarized.
... Raman spectroscopy is an attractive method in the recognition of the causative agents of bacterial infections, and thereby of bacterial infections themselves [32]. Preparations of silver and copper nanoparticles are widely used in medical practices as an alternative to traditional antimicrobial, antifungal, and disinfectant methods. ...
... Raman spectroscopy is an attractive method in the recognition of the causative agents of bacterial infections, and thereby of bacterial infections themselves [32]. ...
In the present study, copper and silver nanoparticles with a concentration of 20 µg/cm 2 were synthesized using the method of laser-induced forward transfer (LIFT). The antibacterial activity of the nanoparticles was tested against bacterial biofilms that are common in nature, formed by several types of microorganisms (mixed-species bacteria biofilms): Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa. The Cu nanoparticles showed complete inhibition of the bacteria biofilms used. In the course of the work, a high level of antibacterial activity was demonstrated by nanoparticles. This activity manifested in the complete suppression of the daily biofilm, with the number of bacteria decreasing by 5-8 orders of magnitude from the initial concentration. To confirm antibacterial activity, and determine reductions in cell viability, the Live/Dead Bacterial Viability Kit was used. FTIR spectroscopy revealed that after Cu NP treatment, there was in a slight shift in the region, which corresponded to fatty acids, indicating a decrease in the relative motional freedom of molecules.
... However, these methods suffer from multi-step complexity and reliance on commercially-available kits, which makes the differentiation procedure time-consuming, sophisticated, and expensive, calling for the development of novel techniques and methods in the field (Pizzato et al., 2022). Recently, as a rapidly developed emerging technique, surface enhanced Raman spectrometry (SERS) has been extensively explored for its potential and promising application in bacterial pathogen detection and antibiotic resistance profiling Wang et al., 2021;Tang et al., 2022). There are currently multiple studies using SERS technique combined with machine learning models to GRAPHICAL ABSTRACT Highlights -Shigella spp., and E. coli SERS spectra had unique characteristic peaks. ...
... The preparation procedures for silver nanoparticles (AgNPs) have been well documents in previous studies Tang et al., 2022;Wang et al., 2022a,b). Therefore, we only briefly stated the preparation steps in terms of the key reactions. ...
... As an easy-to-learn, low-cost, non-invasive and label-free method, Raman spectroscopy, especially the surface enhanced Raman spectroscopy due to the significantly enhanced Raman signals, has great application potential for rapid and accurate bacterial pathogen identification in clinical settings, though huge challenges exist and the method has been officially used in real-world situation yet (Wang et al., 2021). Previously, the promising SERS technique has been widely used for both genus/species discrimination and antibiotic resistance profiling in clinically important bacterial pathogens with rapid speed and high accuracy (Ho et al., 2019;Tang et al., 2021;Tang et al., 2022;Wang et al., 2022a,b). However, there is currently no reported of using label-free SERS technique coupled with machine learning algorithms for the differentiation of Shigella spp., and E. coli, which is thoroughly investigated in this study from the comparison of characteristic peaks for unique molecular structure and components, to clustering analysis of distinct spectral groups, and then to species predictions based on supervised machine learning algorithms. ...
Shigella and enterotoxigenic Escherichia coli (ETEC) are major bacterial pathogens of diarrheal disease that is the second leading cause of childhood mortality globally. Currently, it is well known that Shigella spp., and E . coli are very closely related with many common characteristics. Evolutionarily speaking, Shigella spp., are positioned within the phylogenetic tree of E . coli . Therefore, discrimination of Shigella spp., from E . coli is very difficult. Many methods have been developed with the aim of differentiating the two species, which include but not limited to biochemical tests, nucleic acids amplification, and mass spectrometry, etc. However, these methods suffer from high false positive rates and complicated operation procedures, which requires the development of novel methods for accurate and rapid identification of Shigella spp., and E . coli . As a low-cost and non-invasive method, surface enhanced Raman spectroscopy (SERS) is currently under intensive study for its diagnostic potential in bacterial pathogens, which is worthy of further investigation for its application in bacterial discrimination. In this study, we focused on clinically isolated E . coli strains and Shigella species (spp.), that is, S . dysenteriae , S . boydii , S . flexneri , and S . sonnei , based on which SERS spectra were generated and characteristic peaks for Shigella spp., and E . coli were identified, revealing unique molecular components in the two bacterial groups. Further comparative analysis of machine learning algorithms showed that, the Convolutional Neural Network (CNN) achieved the best performance and robustness in bacterial discrimination capacity when compared with Random Forest (RF) and Support Vector Machine (SVM) algorithms. Taken together, this study confirmed that SERS paired with machine learning could achieve high accuracy in discriminating Shigella spp., from E . coli , which facilitated its application potential for diarrheal prevention and control in clinical settings.
... compared with all of the other machine learning algorithms (23). Recently, Tang et al. investigated the effects of machine learning algorithms on 15 bacterial species, which also showed that the CNN could achieve as high as a 99.86% prediction accuracy with the highest area (0.9996) under the receiver operating characteristic curve (32). It is interesting to note that another Rapid Diagnosis of Bacterial Pathogens Microbiology Spectrum deep learning algorithm, namely long short-term memory (LSTM), did not perform as well as did CNN, and, in some cases, it was not even as good as the classical machine learning algorithms, such as the random forest and KNN (31,32). ...
... Recently, Tang et al. investigated the effects of machine learning algorithms on 15 bacterial species, which also showed that the CNN could achieve as high as a 99.86% prediction accuracy with the highest area (0.9996) under the receiver operating characteristic curve (32). It is interesting to note that another Rapid Diagnosis of Bacterial Pathogens Microbiology Spectrum deep learning algorithm, namely long short-term memory (LSTM), did not perform as well as did CNN, and, in some cases, it was not even as good as the classical machine learning algorithms, such as the random forest and KNN (31,32). However, a recent study by Yu et al. showed that the LSTM method was actually faster and more accurate than was the normal CNN model, achieving an average isolation-level accuracy of greater than 94%, which requires further exploration (33). ...
In this study, we investigated 30 bacterial species belonging to 9 different bacterial genera that were isolated from clinical samples via surfaced enhanced Raman spectroscopy (SERS). A total of 17,149 SERS spectra were harvested from a Raman spectrometer and were further analyzed via machine learning approaches, the results of which showed that the convolutional neural network (CNN) deep learning algorithm could achieve the highest prediction accuracy for recognizing pathogenic bacteria at both the genus and species levels.
... Previously, SERS technique has been widely applied to detect many infection-causing bacterial pathogens, such as Escherichia coli, Staphylococcus aureus and Pseudomonas aeruginosa, etc.; these studies normally focused on analyzing SERS spectra and corresponding characteristic peaks via statistical methods such as partial least square-discriminant analysis (PSL-DA), principal component analysis (PCA), and hierarchical cluster analysis (HCA), etc. [26][27][28][29]. However, due to the complexity of the SERS spectral data, these classical statistical methods are insufficient for data analysis and pattern recognition, which requires the assistance of advanced computational algorithms such as machine learning methods [21,30]. In fact, a variety of studies have already applied machine learning algorithms on SERS spectroscopy for the rapid detection of bacterial pathogens, which include but not limited to support vector machine (SVM) and random forest (RF), etc. [31][32]. ...
... Ho et al. conducted the pioneering study in which the state-of-the-art CNN technique was applied to 60,000 SERS spectra for rapid identification of 30 common bacterial pathogens with the accuracy of 82% [33]. A series of studies then compared the prediction accuracies of various machine learning algorithms including deep learning algorithms in different types of bacterial pathogens [22,30,34]. In specificity, Tang et al. performed the comparative analysis of ten machine learning algorithms on 2752 SERS spectra of nine Staphylococcus species, which confirmed that the convolutional neural network (CNN) algorithm was the best model with an accuracy of 98.21% [22]. ...
... When the data arrives at the fully connected layer, a multi-class neural network is performed through the Softmax activation function to obtain the final output. [30]. Therefore, combination of SERS spectrometry with deep learning algorithm provides an advanced method with sufficient accuracy in identifying bacterial species that holds the application potential in clinical settings. ...
Over the past decades, conventional methods and molecular assays have been developed for the detection of tuberculosis (TB). However, these techniques suffer limitations in the identification of Mycobacterium tuberculosis (Mtb), such as long turnaround time and low detection sensitivity, etc., not even mentioning the difficulty in discriminating antibiotics-resistant Mtb strains that cause great challenges in TB treatment and prevention. Thus, techniques with easy implementation for rapid diagnosis of Mtb infection are in high demand for routine TB diagnosis. Due to the label-free, low-cost and non-invasive features, surface enhanced Raman spectroscopy (SERS) has been extensively investigated for its potential in bacterial pathogen identification. However, at current stage, few studies have recruited handheld Raman spectrometer to discriminate sputum samples with or without Mtb, separate pulmonary Mtb strains from extra-pulmonary Mtb strains, or profile Mtb strains with different antibiotic resistance characteristics. In this study, we recruited a set of supervised machine learning algorithms to dissect different SERS spectra generated via a handheld Raman spectrometer with a focus on deep learning algorithms, through which sputum samples with or without Mtb strains were successfully differentiated (5-fold cross-validation accuracy=94.32%). Meanwhile, Mtb strains isolated from pulmonary and extra-pulmonary samples were effectively separated (5-fold cross-validation accuracy=99.86%). Moreover, Mtb strains with different drug-resistant profiles were also competently distinguished (5-fold cross-validation accuracy=99.59%). Taken together, we concluded that, with the assistance of deep learning algorithms, handheld Raman spectrometer has a high application potential for rapid point-of-care diagnosis of Mtb infections in future.
