Figure 6 - uploaded by Róisín Howard
Content may be subject to copyright.
Source publication
A method of estimating force using an accelerometer is presented. This model is based on estimating the resultant acceleration of a body at its centre of mass using a tri-axial accelerometer. A data set of ground reaction forces are gathered using a force platform, which is used as the control for this experiment. Signal processing techniques for r...
Context in source publication
Context 1
... flight time was also calculated whereby a flight phase began when the force dropped below 10 N and the landing phase began when the force advanced above 10 N. The flight time was used to decide which trial was the best jump from each participant, to be used in the results. Similarly one of the DJ force traces from the force platform can be seen in Figure 6. The peak impulse on initial and final landing can be identified in this plot. ...
Similar publications
Sports biomechanics enables thorough examination of athletic movements to enhance athletic performance and/or reduce injury risk. Few studies have looked at the possibilities of cutting-edge technology in fencing, even though it presents an intriguing scenario for sports biomechanics due to the significant demands it places on the body in terms of...
Experimental research in sports biomechanics in Zagreb dates to the beginning of 1980-ies. In course of these years one has developed and standardised, among other, a certain laboratory diagnostic methodology suited primarily to sports disciplines such as sport gymnastics and track-and-field (see 1, for instance). It was based on surface EMG and gr...
This study aimed to investigate the effects of the external load of elastic bands attached to the waist and heels to enhance the pre‐activation of leg extensor muscles on drop jumps (DJs). Twelve male college athletes volunteered for this study. Eight cameras and two force platforms were used to collect data. Each subject performed DJs with elastic...
Force platforms are the most commonly used sensory instrument used in the measurement of ground reaction forces for sports biomechanical evaluation. This paper documents the development and validation of such a force platform to be used by national level athletes. The developed force platform consists of four load cells, and an aluminium plate of d...
The purpose of this study was to analyse the kinetic and kinematic parameters of
starting movements in swimming using a portable smart platform with Android
pie for sport biomechanics applications. A TAL220 load cell sensor,
microcontroller, an Arduino Nano, Wi-Fi module, and Bluetooth HC-05 were
used to support the various components of the protot...
Citations
... For several years, the scientific community has been interested in the relationship between GRFs and wearable accelerometer signals during walking, running, and jumping [6][7][8][9][10]. With regards to walking, this relationship may allow the detection of muscle weakness in patients with osteogenesis imperfecta in their living conditions [6]. ...
... With regards to walking, this relationship may allow the detection of muscle weakness in patients with osteogenesis imperfecta in their living conditions [6]. With regards to jumping, this relationship can be used to evaluate performance during sports training and conditioning in an ecologically valid environment rather than in a laboratory [9]. With regards to running, it has been shown that there are correlations between the value of the peak measured by an accelerometer located on the tibia and the loading rate or the amplitude of the passive peak during running [11]. ...
The estimation of vertical ground reaction forces (VGRFs) during running is necessary to understand running mechanisms. For this purpose, the use of force platforms is fundamental. However, to extend the study of VGRFs to real conditions, wearable accelerometers are a promising alternative to force platforms, whose use is often limited to the laboratory environment. The objective of this study was to develop a VGRF model using wearable accelerometers and a stepwise regression algorithm. Several models were developed and validated using the VGRFs and acceleration signals collected during 100 stances performed by one participant. The validated models were tested on eight participants. In a sensitivity study, the strongest correlations were observed at cut-off frequencies of ≤25 Hz and in models developed with 30 to 90 stances. After the validation phase, the 10 best models had, on average, low relative differences (≤10%) in the estimation of discrete VGRF parameters, i.e., the passive peak (εpp=6.26%), active peak (εap=2.22%), and loading rate (εlr=2.17%). The results indicate that the development of personalized models is more suitable for achieving the best estimates. The proposed methodology opens many perspectives for monitoring VGRFs under real conditions using a limited number of wearable sensors.
... All four accelerometers recorded the data simultaneously with a force plate. The initiation of the vertical jump was defined as the time point at which the derived ground reaction force decreased by >20 N from the baseline (i.e., system mass), after which peak concentric force and airtime were easily detectable [25,26]. In order to minimize the possible influence of fatigue, each set was separated by a 1-2-minute rest, and each repetition was separated by a 30-60-second rest interval. ...
... Previous research has reported mixed findings regarding the use of accelerometer technology for the estimation of PCF during a countermovement vertical jump with no arm swing [19,24,26]. When compared to the force plate as a criterion measure, Hojka et al. [19] found that the Myotest accelerometer placed on the hip tended to underestimate PCF on average by 167 N. In a similar investigation, Howard et al. [26] discovered that the Shimmer accelerometer placed at the same anatomical location overestimated PCF on average by 619 N. The aforementioned findings are contradictory to our results, where no statistically significant differences were observed between each of the four accelerometer placements (AB, CH, HP, and UB) and the force plate as a gold standard testing modality when performing a countermovement jump without an arm swing. ...
... Previous research has reported mixed findings regarding the use of accelerometer technology for the estimation of PCF during a countermovement vertical jump with no arm swing [19,24,26]. When compared to the force plate as a criterion measure, Hojka et al. [19] found that the Myotest accelerometer placed on the hip tended to underestimate PCF on average by 167 N. In a similar investigation, Howard et al. [26] discovered that the Shimmer accelerometer placed at the same anatomical location overestimated PCF on average by 619 N. The aforementioned findings are contradictory to our results, where no statistically significant differences were observed between each of the four accelerometer placements (AB, CH, HP, and UB) and the force plate as a gold standard testing modality when performing a countermovement jump without an arm swing. However, these results seem to be in agreement with the findings of a recently conducted study that used an identical accelerometer device (StriveTech) and found similar PCF values derived from AB and HP anatomical placements (1753 N and 1776 N) [24]. ...
With rapid technological development over recent years, the use of wearable athlete monitoring devices has substantially gained popularity. Thus, the purpose of the present study was to examine the impact of the anatomical placement of an accelerometer on biomechanical characteristics of countermovement vertical jump with and without an arm swing when compared to the force plate as a criterion measure. Seventeen recreationally active individuals (ten males and seven females) volunteered to participate in the present study. Four identical accelerometers sampling at 100 Hz were placed at the following anatomical locations: upper-back (UB), chest (CH), abdomen (AB), and hip (HP). While standing on a uni-axial force plate system sampling at 1000 Hz, each participant completed three non-sequential maximal countermovement vertical jumps with and without an arm swing. All devices recorded the data simultaneously. The following variables of interest were obtained from ground reaction force curves: peak concentric force (PCF), peak landing force (PLF), and vertical jump height (VJH). The findings of the present study reveal that the most appropriate anatomical locations to place the accelerometer device when attempting to estimate PCF, PLF, and VJH during a countermovement vertical jump with no arm swing are CH, AB, and UB, and during a countermovement vertical jump with an arm swing are UB, HP, and UB, respectively. Overall, these findings may help strength and conditioning professionals and sports scientists to select appropriate anatomical locations when using innovative accelerometer technology to monitor vertical jump performance characteristics.
... A triaxial accelerometer consistently overestimated peak force, rate of force development, peak power, flight time, and vertical displacement when compared to a force plate and linear position transducer (Ruben et al., 2011). Additionally, peak concentric forces from an accelerometer were consistently higher when compared to a force plate, despite a good agreement between the devices on minimum eccentric force (Howard et al., 2014). Castagna et al. (2013) conducted a study on an elite cohort of rugby players and found that flight times reported by a wearable accelerometer during maximal countermovement vertical jumps were notably greater when compared to an optical jump height system and a laboratory-based force plate system as a criterion measure. ...
While force plate technology is the gold standard for assessment of many aspects of vertical jump performance, its cost is prohibitive to a broad spectrum of the population. Accelerometry may be more practical, inexpensive, and provide a simple solution that allows hands-on practitioners to readily assess vertical jump performance acutely and over time. Thus, the purpose of this study was to examine the accuracy of an experimental accelerometer for testing vertical jump heights derived from flight times when compared to a laboratory-based force plate system as a criterion measure. Fifteen subjects performed three sets of three nonconsecutive maximal countermovement vertical jumps while standing on the force plate. The accelerometer device sampling at 100 Hz was placed on the anterior abdomen immediately inferior to the umbilicus and secured with an elastic band. Both devices recorded the data simultaneously. The experimental accelerometer was an appropriate tool for the assessment of vertical jump height; however, it significantly overestimated actual vertical jump heights by an average of 3.1 cm. This consistent discrepancy in the measurement may be easily fixed by a simple algorithm correction and should not present an issue in the practical setting where ease of use and the ability to provide immediate feedback regarding an athlete’s performance is of critical importance.
... In the literature, there are many protocols to prove or validate the proposed systems. Among the different kind of jumps performed in those protocols, there are jumps with and without countermovement [1,4,5,[8][9][10][11][12][13][14][15], jumps with and without arm swing [12,16], drop jumps [1,8,17], single and double leg jump [6], continuous jumps [4,17], squat jumps [1,2,4,12], and loaded squat jumps [7]. With any of these types of jumps, height reached by the user can be analyzed, but the jumps most commonly used in all related work are the countermovement and squat jumps. ...
... In the literature, there are many protocols to prove or validate the proposed systems. Among the different kind of jumps performed in those protocols, there are jumps with and without countermovement [1,4,5,[8][9][10][11][12][13][14][15], jumps with and without arm swing [12,16], drop jumps [1,8,17], single and double leg jump [6], continuous jumps [4,17], squat jumps [1,2,4,12], and loaded squat jumps [7]. With any of these types of jumps, height reached by the user can be analyzed, but the jumps most commonly used in all related work are the countermovement and squat jumps. ...
... However, these methods often show overestimation on jump height, and this could be due to arm stretching performed unconsciously by the user. Among systems developed in the literature, different kind of sensors are used like force-sensitive resistors (FSRs) [3,16], capacitive sensors [5], inertial measurement units [2,4,8,10,17], electromyography sensors [1,6], kinematic sensors [6], ultrasonic sensors [20], microswitches [9], video cameras, [11] and optical sensors [4,12]. ...
Background
Vertical jump height is widely used in health care and sports fields to assess muscle strength and power from lower limb muscle groups. Different approaches have been proposed for vertical jump height measurement. Some commonly used approaches need no sensor at all; however, these methods tend to overestimate the height reached by the subjects. There are also novel systems using different kind of sensors like force-sensitive resistors, capacitive sensors, and inertial measurement units, among others, to achieve more accurate measurements.
Objective
The objective of this study is twofold. The first objective is to validate the functioning of a developed low-cost system able to measure vertical jump height. The second objective is to assess the effects on obtained measurements when the sampling frequency of the system is modified.
Methods
The system developed in this study consists of a matrix of force-sensitive resistor sensors embedded in a mat with electronics that allow a full scan of the mat. This mat detects pressure exerted on it. The system calculates the jump height by using the flight-time formula, and the result is sent through Bluetooth to any mobile device or PC. Two different experiments were performed. In the first experiment, a total of 38 volunteers participated with the objective of validating the performance of the system against a high-speed camera used as reference (120 fps). In the second experiment, a total of 15 volunteers participated. Raw data were obtained in order to assess the effects of different sampling frequencies on the performance of the system with the same reference device. Different sampling frequencies were obtained by performing offline downsampling of the raw data. In both experiments, countermovement jump and countermovement jump with arm swing techniques were performed.
Results
In the first experiment an overall mean relative error (MRE) of 1.98% and a mean absolute error of 0.38 cm were obtained. Bland-Altman and correlation analyses were performed, obtaining a coefficient of determination equal to R2=.996. In the second experiment, sampling frequencies of 200 Hz, 100 Hz, and 66.6 Hz show similar performance with MRE below 3%. Slower sampling frequencies show an exponential increase in MRE. On both experiments, when dividing jump trials in different heights reached, a decrease in MRE with higher height trials suggests that the precision of the proposed system increases as height reached increases.
Conclusions
In the first experiment, we concluded that results between the proposed system and the reference are systematically the same. In the second experiment, the relevance of a sufficiently high sampling frequency is emphasized, especially for jump trials whose height is below 10 cm. For trials with heights above 30 cm, MRE decreases in general for all sampling frequencies, suggesting that at higher heights reached, the impact of high sampling frequencies is lesser.
... The task of vertical jump was further investigated by Howard et al. [67] by means of a tri-axial accelerometer placed close to the centre of mass. GRF were simultaneously measured by means of a FP while the subjects were performing some countermovement and drop jumps. ...
... Therefore the force obtained from the measurement of acceleration could not be used interchangeably with the force measured by the FP. Thus, it was recommended to use gyroscopes to increase accuracy of datasets [67]. ...
In the last few years, estimating ground reaction forces by means of wearable sensors has come to be a challenging research topic paving the way to kinetic analysis and sport performance testing outside of labs. One possible approach involves estimating the ground reaction forces from kinematic data obtained by inertial measurement units (IMUs) worn by the subject. As estimating kinetic quantities from kinematic data is not an easy task, several models and protocols have been developed over the years. Non-wearable sensors, such as optoelectronic systems along with force platforms, remain the most accurate systems to record motion. In this review, we identified, selected and categorized the methodologies for estimating the ground reaction forces from IMUs as proposed across the years. Scopus, Google Scholar, IEEE Xplore, and PubMed databases were interrogated on the topic of Ground Reaction Forces estimation based on kinematic data obtained by IMUs. The identified papers were classified according to the methodology proposed: (i) methods based on direct modelling; (ii) methods based on machine learning. The methods based on direct modelling were further classified according to the task studied (walking, running, jumping, etc.). Finally, we comparatively examined the methods in order to identify the most reliable approaches for the implementation of a ground reaction force estimator based on IMU data.
... The advantage of this over the vision-based system is that it does not require any additional hardware. The methodology for force calculation through acceleration measurement has been discussed in [9]. ...
... The error in the calculation of force is of the order of 10 −3 g which is about 0.001%. The methodology for the estimation of force using the accelerometer data is provided in [9]. The data is acquired from both x and y axes of the accelerometer. ...
... Time-synchronization is therefore a matter of applying the calculated lag to the time stamps of one of the two data sets. This method is similar to what Howard et al. [83] use to synchronize their IMU and force platform signals. ...
In this paper, we describe and validate the EquiMoves system, which aims to support equine veterinarians in assessing lameness and gait performance in horses. The system works by capturing horse motion from up to eight synchronized wireless inertial measurement units. It can be used in various equine gait modes, and analyzes both upper-body and limb movements. The validation against an optical motion capture system is based on a Bland–Altman analysis that illustrates the agreement between the two systems. The sagittal kinematic results (protraction, retraction, and sagittal range of motion) show limits of agreement of ±2.3 degrees and an absolute bias of 0.3 degrees in the worst case. The coronal kinematic results (adduction, abduction, and coronal range of motion) show limits of agreement of −8.8 and 8.1 degrees, and an absolute bias of 0.4 degrees in the worst case. The worse coronal kinematic results are most likely caused by the optical system setup (depth perception difficulty and suboptimal marker placement). The upper-body symmetry results show no significant bias in the agreement between the two systems; in most cases, the agreement is within ±5 mm. On a trial-level basis, the limits of agreement for withers and sacrum are within ±2 mm, meaning that the system can properly quantify motion asymmetry. Overall, the bias for all symmetry-related results is less than 1 mm, which is important for reproducibility and further comparison to other systems.
... This study has shown accelerometry can be used to very accurately and reliably estimate force production during ambulation at varying speeds (r=0.98) to within <1% of force platform recorded values. This has also been demonstrated for counter-movement jumping, Howard, et al. 233 , found both minimum eccentric force and maximum concentric force were accurately estimated, in comparison to force platforms although there were higher intra-class correlations in the minimum eccentric force (r=0.93), compared to maximum concentric force (r=0.6). ...
There is a paucity of research demonstrating objective methods to empirically derive quality of movement measurements and its subsequent relationship to health, performance and risk factors. Therefore, the overarching aim of this thesis was to characterise and profile children’s physical activity movement (and gait) quality.
Following laboratory based work to confirm the feasibility and development of assessing movement quality in children, ankle mounted accelerometers were used for all experimental studies. Classic and novel temporal and frequency domain analyses were conducted in all studies. All data underwent hierarchical clustering based on normalised Euclidean distances. Further inferential statistics were conducted to investigate differences and correlations, accordingly.
Experimental chapter 1 consisted of three smaller studies to, first, test the technical specifications of the SlamTracker accelerometer and that the data output were valid and reliable. Second, to verify the validity of using raw accelerometry to estimate movement characteristics during ambulation compared to gold standard methods, and third, to characterise the relationship between overall integrated acceleration and three-dimensional kinematic variables whilst performing fundamental movement skills. The accuracy, suitability and validity of the SlamTracker raw accelerometer data (absolute variance: <0.001 g, CV: 0.004%, in all axes) was confirmed. Following this, the ability to accurately capture complex movement characteristics, compared to gold standard methods, such as joint angle (r=0.98, P=0.001) and force production (r=0.98, P=0.001) was demonstrated. Finally, there were no differences found in overall activity (integrated acceleration) in children who completed the same fundamental movements, whilst a large variance was detected in the kinematics of children’s movement (CV: up to 65%). We concluded that quality of movement, whilst evidently important and diverse even in standardised tasks, needed a specific operational definition in the context of this thesis. Following pilot work, the term quality was therefore defined as, and derived from, the purity of the fundamental frequency spectra (signal) during human movement, specifically relating to gait, otherwise termed, spectral purity.
Experimental chapter 2 was necessary to establish a credible base for the combination of raw accelerometry and novel analytics, outside of laboratory settings. This was the first empirical study to draw upon frequency domain analysis and hierarchical clustering in the characterisation of movement in children. The aims of this study were; first, to characterise the movement quality of children during a standardised fitness assessment and second, to report how movement quality characteristics cluster according to body mass indices in 9-11y children. One hundred and three children (10.3±0.6y, 1.42±0.08m, 37.8±9.3kg, body mass index; 18.5±3.3 kg.m2) volunteered for this study, had anthropometric recordings taken and took part in the twenty-metre multistage fitness test. This study found that children with high BMI had significantly lower spectral purity and time to exhaustion. Moreover, BMI was hierarchically clustered with stride profile, and time to exhaustion was clustered with spectral purity. BMI was negatively correlated with time to exhaustion, spectral purity, integrated acceleration, stride angle and stride variability. In conclusion, spectral purity was representative of children’s performance during a standardised fitness test, and significantly negatively correlated with body mass index.
Experimental chapter 1 and 2 both utilised more controlled environments, whereas Experimental chapter 3 moved away from controlled into more free-form, uncontrolled movement, i.e. recess. The aims of this study were to characterise children’s recess physical activity, and investigate how movement quality characteristics cluster during school recess. Twenty-four children (18 boys) (10.5±0.6y, 1.44±0.09m, 39.6±9.5kg, body mass index; 18.8±3.1 kg.m2) who were a representative sub-sample of 822 children (10.5±0.6y, 1.42±0.08m, 27.3±9.6kg, body mass index; 18.7±3.5 kg.m2), took part in a normal school-time recess for one school week (five days). This study found that integrated acceleration (overall physical activity) during recess was invariant day-to-day, yet significant daily differences were found for spectral purity. Integrated acceleration was clustered with spectral purity, in addition to a significant positive correlation between integrated acceleration and spectral purity (P<0.05), whilst body-mass index percentile was negatively correlated with integrated acceleration and spectral purity. This study highlighted that movement quality measurement was achievable and robust in an uncontrolled environment (i.e. recess). Given the link established in previous chapters between spectral purity and movement quality, the tenuous literature on motor competency development through childhood, and, the evidence motor competence may track though the life course; it was deemed appropriate to examine movement quality characteristics in early years’ children, in conjunction with traditional motor competency assessment.
The aims of Experimental chapter 4 were two-fold; to characterise children’s free-play physical activity and investigate how movement quality characteristics cluster in children (3-5y). Sixty-one children (39 boys, 4.3±0.7y, 1.04±0.05m, 17.8±3.2kg, body mass index; 16.2±1.9 kg.m2) took part in free-play and completed the movement assessment battery for children, second edition, using standardised procedures. There were significant differences between motor competency classifications for spectral purity and integrated acceleration (P<0.001). Spectral purity was hierarchically clustered with motor competence and overall physical activity. Additional significant positive correlations were found between spectral purity, integrated acceleration and motor competence (P<0.001).
In conclusion, children’s movement quality can be reliably computed using novel analytics in laboratory and in-field. The novel quality measure coined in this thesis, spectral purity, was shown to be hierarchically clustered with, and indicative of, performance, physical activity and motor competence. This thesis has expanded the current evidence base on children’s physical activity and movement quality and demonstrated that raw accelerometry can be used, in conjunction with novel analytics, to provide innovation in movement quality assessment across ages.
... Others have rotated the acceleration vector measured in the sensor reference frame such that it is expressed in the world reference frame to accurately assess center of mass kinematics during jumping [28,29] and walking [30] tasks. If the orientation of an IMU relative to a force plate is known, then IMU estimates of F using Newton's 2 nd Law may be compared to that measured by the force plate [31]. To the author's knowledge, no studies have assessed the ability of a trunk mounted IMU to perform kinetic analyses of accelerative running tasks. ...
... For example, accelerometers have shown the ability to deterct running fatigue [74] and contact times during steady state jogging, running, and sprinting, as well as the first, third, and fifth steps of accelerative running [75]. Other studies have estimated running speed in free living conditions [76], spatiotemporal data of ice hockey skating [77] and sprinting [78], biomechanical variables of countermovement and drop jumps [31], joint angles [79], and kinematics of the barbell high pull [80]. ...
... These two prediction models by Neugebauer et al. [17] and Neugebauer et al. [16] are dependent on curve fitting accelerometer data with force plate data from preliminary trials as opposed to using Newton's 2 nd Law. The latter may be more generalizable and has been used more recently for kinetic analysis of hopping and heel-rise tests [83], the development of a smartphone application to estimate kinetic and kinematic variables during a sit-to-stand task [84], and to estimate eccentric and concentric forces during drop jumps and countermovement jumps [31]. The first two of these three studies suggest valid estimates ...
