Inertial Sensor-Based Gait Recognition: A Review

Locomotion in virtual environments predicts cardiovascular responsiveness to subsequent stressful challenges

Individuals differ in their physiological responsiveness to stressful challenges, and stress potentiates the development of many diseases. Heart rate variability (HRV), a measure of cardiac vagal break, is emerging as a strong index of physiological stress vulnerability. Thus, it is important to develop tools that identify predictive markers of individual differences in HRV responsiveness without exposing subjects to high stress. Here, using machine learning approaches, we show the strong predictive power of high-dimensional locomotor responses during novelty exploration to predict HRV responsiveness during stress exposure. Locomotor responses are collected in two ecologically valid virtual reality scenarios inspired by the animal literature and stress is elicited and measured in a third threatening virtual scenario. Our model's predictions generalize to other stressful challenges and outperforms other stress prediction instruments, such as anxiety questionnaires. Our study paves the way for the development of behavioral digital phenotyping tools for early detection of stress-vulnerable individuals.

Agreement of Gait Events Detection during Treadmill Backward Walking by Kinematic Data and Inertial Motion Units

Backward walking (BW) is being increasingly used in neurologic and orthopedic rehabilitation as well as in sports to promote balance control as it provides a unique challenge to the sensorimotor control system. The identification of initial foot contact (IC) and terminal foot contact (TC) events is crucial for gait analysis. Data of optical motion capture (OMC) kinematics and inertial motion units (IMUs) are commonly used to detect gait events during forward walking (FW). However, the agreement between such methods during BW has not been investigated. In this study, the OMC kinematics and inertial data of 10 healthy young adults were recorded during BW and FW on a treadmill at different speeds. Gait events were measured using both kinematics and inertial data and then evaluated for agreement. Excellent reliability (Interclass Correlation > 0.9) was achieved for the identification of both IC and TC. The absolute differences between methods during BW were 18.5 ± 18.3 and 20.4 ± 15.2 ms for IC and TC, respectively, compared to 9.1 ± 9.6 and 10.0 ± 14.9 for IC and TC, respectively, during FW. The high levels of agreement between methods indicate that both may be used for some applications of BW gait analysis.

Responsiveness of Objective vs. Clinical Balance Domain Outcomes for Exercise Intervention in Parkinson's Disease

Background: Balance deficits in people with Parkinson's disease (PD) are often not helped by pharmacological or surgical treatment. Although balance exercise intervention has been shown to improve clinical measures of balance, the efficacy of exercise on different, objective balance domains is still unknown.

Objective: To compare the sensitivity to change in objective and clinical measures of several different domains of balance and gait following an Agility Boot Camp with Cognitive Challenges (ABC-C) intervention.

Methods: In this cross-over, randomized design, 86 individuals with PD participated in 6-week (3×/week) ABC-C exercise classes and 6-week education classes, consisting of 3–6 individuals. Blinded examiners tested people in their practical off state. Objective outcome measures from wearable sensors quantified four domains of balance: sway in standing balance, anticipatory postural adjustments (APAs) during step initiation, postural responses to the push-and-release test, and a 2-min natural speed walk with and without a cognitive task. Clinical outcome measures included the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) Part III, the Mini Balance Evaluation Systems Test (Mini-BESTest), the Activities of Balance Confidence (ABC), and the Parkinson's Disease Questionnaire (PDQ-39). The standardized response means (SRM) of the differences between before and after each intervention compared responsiveness of outcomes to intervention. A linear mixed model compared effects of exercise with the active control—education intervention.

Results: The most responsive outcome measures to exercise intervention with an SRM > 0.5 were objective measures of gait and APAs, specifically arm range of motion, gait speed during a dual-task walk, trunk coronal range of motion, foot strike angle, and first-step length at step initiation. The most responsive clinical outcome measure was the patient-reported PDQ-39 activities daily living subscore, but all clinical measures had SRMs <0.5.

Conclusions: The objective measures were more sensitive to change after exercise intervention compared to the clinical measures. Spatiotemporal parameters of gait, including gait speed with a dual task, and APAs were the most sensitive objective measures, and perceived functional independence was the most sensitive clinical measure to change after the ABC-C exercise intervention. Future exercise intervention to improve gait and balance in PD should include objective outcome measures.

Gait Phase Recognition Using Deep Convolutional Neural Network with Inertial Measurement Units

Gait phase recognition is of great importance in the development of assistance-as-needed robotic devices, such as exoskeletons. In order for a powered exoskeleton with phase-based control to determine and provide proper assistance to the wearer during gait, the user’s current gait phase must first be identified accurately. Gait phase recognition can potentially be achieved through input from wearable sensors. Deep convolutional neural networks (DCNN) is a machine learning approach that is widely used in image recognition. User kinematics, measured from inertial measurement unit (IMU) output, can be considered as an ‘image’ since it exhibits some local ‘spatial’ pattern when the sensor data is arranged in sequence. We propose a specialized DCNN to distinguish five phases in a gait cycle, based on IMU data and classified with foot switch information. The DCNN showed approximately 97% accuracy during an offline evaluation of gait phase recognition. Accuracy was highest in the swing phase and lowest in terminal stance.

Gait Recognition as an Authentication Method for Mobile Devices

With the rate at which smartphones are currently evolving, more and more of human life will be contained in these devices. At a time when data privacy is extremely important, it is crucial to protect one’s mobile device. In this paper, we propose a new non-intrusive gait recognition based mechanism that can enhance the security of smartphones by rapidly identifying users with a high degree of confidence and securing sensitive data in case of an attack, with a focus on a potential architecture for such an algorithm for the Android environment. The motion sensors on an Android device are used to create a statistical model of a user’s gait, which is later used for identification. Through experimental testing, we prove the capability of our proposed solution by correctly classifying individuals with an accuracy upwards of 90% when tested on data recorded during multiple activities. The experiments, conducted on a low sampling rate and at short time intervals, show the benefits of our solution and highlight the feasibility of an efficient gait recognition mechanism on modern smartphones.

Evaluation of spinal posture during gait with inertial measurement units

The increasing number of postural disorders emphasizes the central role of the vertebral spine during gait. Indeed, clinicians need an accurate and non-invasive method to evaluate the effectiveness of a rehabilitation program on spinal kinematics. Accordingly, the aim of this work was the use of inertial sensors for the assessment of angles among vertebral segments during gait. The spine was partitioned into five segments and correspondingly five inertial measurement units were positioned. Articulations between two adjacent spine segments were modeled with spherical joints, and the tilt-twist method was adopted to evaluate flexion-extension, lateral bending and axial rotation. In total, 18 young healthy subjects (9 males and 9 females) walked barefoot in three different conditions. The spinal posture during gait was efficiently evaluated considering the patterns of planar angles of each spine segment. Some statistically significant differences highlighted the influence of gender, speed and imposed cadence. The proposed methodology proved the usability of inertial sensors for the assessment of spinal posture and it is expected to efficiently point out trunk compensatory pattern during gait in a clinical context.

A review of foot pose and trajectory estimation methods using inertial and auxiliary sensors for kinematic gait analysis

The use of foot mounted inertial and other auxiliary sensors for kinematic gait analysis has been extensively investigated during the last years. Although, these sensors still yield less accurate results than those obtained employing optical motion capture systems, the miniaturization and their low cost have allowed the estimation of kinematic spatiotemporal parameters in laboratory conditions and real life scenarios. The aim of this work was to present a comprehensive approach of this scientific area through a systematic literature research, breaking down the state-of-the-art methods into three main parts: (1) zero velocity interval detection techniques; (2) assumptions and sensors' utilization; (3) foot pose and trajectory estimation methods. Published articles from 1995 until December of 2018 were searched in the PubMed, IEEE Xplore and Google Scholar databases. The research was focused on two categories: (a) zero velocity interval detection methods; and (b) foot pose and trajectory estimation methods. The employed assumptions and the potential use of the sensors have been identified from the retrieved articles. Technical characteristics, categorized methodologies, application conditions, advantages and disadvantages have been provided, while, for the first time, assumptions and sensors' utilization have been identified, categorized and are presented in this review. Considerable progress has been achieved in gait parameters estimation on constrained laboratory environments taking into account assumptions such as a person walking on a flat floor. On the contrary, methods that rely on less constraining assumptions, and are thus applicable in daily life, led to less accurate results. Rule based methods have been mainly used for the detection of the zero velocity intervals, while more complex techniques have been proposed, which may lead to more accurate gait parameters. The review process has shown that presently the best-performing methods for gait parameter estimation make use of inertial sensors combined with auxiliary sensors such as ultrasonic sensors, proximity sensors and cameras. However, the experimental evaluation protocol was much more thorough, when single inertial sensors were used. Finally, it has been highlighted that the accuracy of setups using auxiliary sensors may further be improved by collecting measurements during the whole foot movement and not only partially as is currently the practice. This review has identified the need for research and development of methods and setups that allow for the robust estimation of kinematic gait parameters in unconstrained environments and under various gait profiles.

Exploration and Research of Human Identification Scheme Based on Inertial Data

The identification work based on inertial data is not limited by space, and has high flexibility and concealment. Previous research has shown that inertial data contains information related to behavior categories. This article discusses whether inertial data contains information related to human identity. The classification experiment, based on the neural network feature fitting function, achieves 98.17% accuracy on the test set, confirming that the inertial data can be used for human identification. The accuracy of the classification method without feature extraction on the test set is only 63.84%, which further indicates the need for extracting features related to human identity from the changes in inertial data. In addition, the research on classification accuracy based on statistical features discusses the effect of different feature extraction functions on the results. The article also discusses the dimensionality reduction processing and visualization results of the collected data and the extracted features, which helps to intuitively assess the existence of features and the quality of different feature extraction effects.

Sensor-to-Segment Calibration Methodologies for Lower-Body Kinematic Analysis with Inertial Sensors: A Systematic Review

Kinematic analysis is indispensable to understanding and characterizing human locomotion. Thanks to the development of inertial sensors based on microelectronics systems, human kinematic analysis in an ecological environment is made possible. An important issue in human kinematic analyses with inertial sensors is the necessity of defining the orientation of the inertial sensor coordinate system relative to its underlying segment coordinate system, which is referred to sensor-to-segment calibration. Over the last decade, we have seen an increase of proposals for this purpose. The aim of this review is to highlight the different proposals made for lower-body segments. Three different databases were screened: PubMed, Science Direct and IEEE Xplore. One reviewer performed the selection of the different studies and data extraction. Fifty-five studies were included. Four different types of calibration method could be identified in the articles: the manual, static, functional, and anatomical methods. The mathematical approach to obtain the segment axis and the calibration evaluation were extracted from the selected articles. Given the number of propositions and the diversity of references used to evaluate the methods, it is difficult today to form a conclusion about the most suitable. To conclude, comparative studies are required to validate calibration methods in different circumstances.

State of the Field of waist-mounted sensor algorithm for gait events detection: A scoping review

BACKGROUND

A waist-mounted sensor is an attractive option for detecting initial and end of foot contacts during gait in a clinical setting without disturbing the subject's natural gait.

RESEARCH QUESTION

To examine the current state of the field regarding waist-mounted sensor algorithms for gait event detection during locomotion in adults.

METHODS

A scoping review design was used to search peer-reviewed literature or conference proceedings published through October 2018 for algorithms for gait event detection. We analyzed data from the studies in a descriptive manner.

RESULTS

In total, 588 potentially relevant articles were selected, of which 14 (171 participants, mean age: 44.0 years) met the inclusion criteria. We identified 15 algorithms developed using biomechanical theories including the inverted pendulum model that represents gait during level walking. Most algorithms estimated gait events using triaxial acceleration data with an absolute error of approximately 50-100 ms in healthy adults. However, there was a large amount of inter-trial heterogeneity, and only a few algorithms were validated in patients with neurological diseases. Lower gait speed reduced the accuracy of gait event estimation.

SIGNIFICANCE

There was no algorithm that showed outstanding performance in the estimation of gait events during level walking using the waist-mounted sensor. More comparisons of all available algorithms with an established reference standard for one data-set are needed to identify the best algorithms. As patients with pathological conditions display altered trunk acceleration and slower gait speeds, the development of an algorithm that does not rely on particular signal characteristics and is robust for a wide range of gait speeds is needed before a specific algorithm can be recommended as a valid strategy for clinical practice.

Using an Accelerometer-Based Step Counter in Post-Stroke Patients: Validation of a Low-Cost Tool

Monitoring the real-life mobility of stroke patients could be extremely useful for clinicians. Step counters are a widely accessible, portable, and cheap technology that can be used to monitor patients in different environments. The aim of this study was to validate a low-cost commercial tri-axial accelerometer-based step counter for stroke patients and to determine the best positioning of the step counter (wrists, ankles, and waist). Ten healthy subjects and 43 post-stroke patients were enrolled and performed four validated clinical tests (10 m, 50 m, and 6 min walking tests and timed up and go tests) while wearing five step counters in different positions while a trained operator counted the number of steps executed in each test manually. Data from step counters and those collected manually were compared using the intraclass coefficient correlation and mean average percentage error. The Bland–Altman plot was also used to describe agreement between the two quantitative measurements (step counter vs. manual counting). During walking tests in healthy subjects, the best reliability was found for lower limbs and waist placement (intraclass coefficient correlations (ICCs) from 0.46 to 0.99), and weak reliability was observed for upper limb placement in every test (ICCs from 0.06 to 0.38). On the contrary, in post-stroke patients, moderate reliability was found only for the lower limbs in the 6 min walking test (healthy ankle ICC: 0.69; pathological ankle ICC: 0.70). Furthermore, the Bland–Altman plot highlighted large average discrepancies between methods for the pathological group. However, while the step counter was not able to reliably determine steps for slow patients, when applied to the healthy ankle of patients who walked faster than 0.8 m/s, it counted steps with excellent precision, similar to that seen in the healthy subjects (ICCs from 0.36 to 0.99). These findings show that a low-cost accelerometer-based step counter could be useful for measuring mobility in select high-performance patients and could be used in clinical and real-world settings.

Motor style at rest and during locomotion in human

Humans exhibit various motor styles that reflect their intra- and interindividual variability when implementing sensorimotor transformations. This opens important questions, such as, At what point should they be readjusted to maintain optimal motor control? Do changes in motor style reveal the onset of a pathological process and can these changes help rehabilitation and recovery? To further investigate the concept of motor style, tests were carried out to quantify posture at rest and motor control in 18 healthy subjects under four conditions: walking at three velocities (comfortable walking, walking at 4 km/h, and race walking) and running at maximum velocity. The results suggest that motor control can be conveniently decomposed into a static component (a stable configuration of the head and column with respect to the gravitational vertical) and dynamic components (head, trunk, and limb movements) in humans, as in quadrupeds, and both at rest and during locomotion. These skeletal configurations provide static markers to quantify the motor style of individuals because they exhibit large variability among subjects. Also, using four measurements (jerk, root mean square, sample entropy, and the two-thirds power law), it was shown that the dynamics were variable at both intra- and interindividual levels during locomotion. Variability increased following a head-to -toe gradient. These findings led us to select dynamic markers that could define, together with static markers, the motor style of a subject. Finally, our results support the view that postural and motor control are subserved by different neuronal networks in frontal, sagittal, and transversal planes.NEW & NOTEWORTHY During human locomotion, motor control can be conveniently decomposed into a static and dynamic components. Variable dynamics were observed at both the intra- and interindividual levels during locomotion. Variability increased following a head-to-toe gradient. Finally, our results support the view that postural and motor control are subserved by different neuronal networks in the frontal, sagittal, and transversal planes.

Medical-Grade ECG Sensor for Long-Term Monitoring

The recent trend in electrocardiogram (ECG) device development is towards wireless body sensors applied for patient monitoring. The ultimate goal is to develop a multi-functional body sensor that will provide synchronized vital bio-signs of the monitored user. In this paper, we present an ECG sensor for long-term monitoring, which measures the surface potential difference between proximal electrodes near the heart, called differential ECG lead or differential lead, in short. The sensor has been certified as a class IIa medical device and is available on the market under the trademark Savvy ECG. An improvement from the user’s perspective—immediate access to the measured data—is also implemented into the design. With appropriate placement of the device on the chest, a very clear distinction of all electrocardiographic waves can be achieved, allowing for ECG recording of high quality, sufficient for medical analysis. Experimental results that elucidate the measurements from a differential lead regarding sensors’ position, the impact of artifacts, and potential diagnostic value, are shown. We demonstrate the sensors’ potential by presenting results from its various areas of application: medicine, sports, veterinary, and some new fields of investigation, like hearth rate variability biofeedback assessment and biometric authentication.

Gait Recognition via Deep Learning of the Center-of-Pressure Trajectory

The fact that every human has a distinctive walking style has prompted a proposal to use gait recognition as an identification criterion. Using end-to-end learning, I investigated whether the center-of-pressure (COP) trajectory is sufficiently unique to identify a person with high certainty. Thirty-six adults walked for 30 min on a treadmill equipped with a force platform that continuously recorded the positions of the COP. The raw two-dimensional signals were sliced into segments of two gait cycles. A set of 20,250 segments from 30 subjects was used to configure and train convolutional neural networks (CNNs). The best CNN classified a separate set containing 2250 segments with an overall accuracy of 99.9%. A second set of 4500 segments from the six remaining subjects was then used for transfer learning. Several small subsamples of this set were selected randomly and used to fine tune the pretrained CNNs. Training with two segments per subject was sufficient to achieve 100% accuracy. The results suggest that every person produces a unique trajectory of underfoot pressures while walking and that CNNs can learn the distinctive features of these trajectories. By applying a pretrained CNN (transfer learning), a couple of strides seem enough to learn and identify new gaits. However, these promising results should be confirmed in a larger sample under realistic conditions.

Validation of a Novel Device for the Knee Monitoring of Orthopaedic Patients

Fast-track surgery is becoming increasingly popular, whereas the monitoring of postoperative rehabilitation remains a matter of considerable debate. The aim of this study was to validate a newly developed wearable system intended to monitor knee function and mobility. A sensor system with a nine-degree-of-freedom (DOF) inertial measurement unit (IMU) was developed. Thirteen healthy volunteers performed five 10-meter walking trials with simultaneous sensor and motion capture data collection. The obtained kinematic waveforms were analysed using root mean square error (RMSE) and correlation coefficient (CC) calculations. The Bland–Altman method was used for the agreement of discrete parameters consisting of peak knee angles between systems. To test the reliability, 10 other subjects with sensors walked a track of 10 metres on two consecutive days. The Pearson CC was excellent for the walking data set between both systems (r = 0.96) and very good (r = 0.95) within the sensor system. The RMSE during walking was 5.17° between systems and 6.82° within sensor measurements. No significant differences were detected between the mean values observed, except for the extension angle during the stance phase (E1). Similar results were obtained for the repeatability test. Intra-class correlation coefficients (ICCs) between systems were excellent for the flexion angle during the swing phase (F1); good for the flexion angle during the stance phase (F2) and the re-extension angle, which was calculated by subtracting the extension angle at swing phase (E2) from F2; and moderate for the extension angle during the stance phase (E1), E2 and the range of motion (ROM). ICCs within the sensor measurements were good for the ROM, F2 and re-extension, and moderate for F1, E1 and E2. The study shows that the novel sensor system can record sagittal knee kinematics during walking in healthy subjects comparable to those of a motion capture system.

Gait Cycle Validation and Segmentation Using Inertial Sensors

In this paper, we develop an algorithm to automatically validate and segment a gait cycle in real time into three gait events, namely midstance, toe-off, and heel-strike, using inertial sensors. We first use the physical models of sensor data obtained from a foot-mounted inertial system to differentiate stationary and moving segments of the sensor data. Next, we develop an optimization routine called sparsity-assisted wavelet denoising (SAWD), which simultaneously combines linear time invariant filters, orthogonal multiresolution representations such as wavelets, and sparsity-based methods, to generate a sparse template of the moving segments of the gyroscope measurements in the sagittal plane for valid gait cycles. Thereafter, to validate any moving segment as a gait cycle, we compute the root-mean-square error between the generated sparse template and the sparse representation of the moving segment of the gyroscope data in the sagittal plane obtained using SAWD. Finally, we find the local minima for the stationary and moving segments of a valid gait cycle to detect the gait events. We compare our proposed method with existing methods, for a fixed threshold, using real data obtained from three groups, namely controls, participants with Parkinson disease, and geriatric participants. Our proposed method demonstrates an average F1 score of 87.78% across all groups for a fixed sampling rate, and an average F1 score of 92.44% across all Parkinson disease participants for a variable sampling rate.

Fugl-Meyer Assessment Scores Are Related With Kinematic Measures in People with Chronic Hemiparesis after Stroke

BACKGROUND

Stroke often results in motor impairment and limited functional capacity. This study aimed to verify the relationship between widely used clinical scales and instrumented measurements to evaluate poststroke individuals with mild, moderate, and severe motor impairment.

METHODS

This cross-sectional study included 34 participants with chronic hemiparesis after stroke. Fugl-Meyer Assessment and Modified Ashworth Scale were used to quantify upper and lower limb motor impairment and the resistance to passive movement (i.e., spasticity), respectively. Upper limb Motor performance (movement time and velocities) and movement quality (range of motion, smoothness and trunk displacement) were analyzed during a reaching forward task using an optoelectronic system (instrumented measurement). Lower limb motor performance (gait and functional mobility parameters) was assessed by using an inertial measurement unit system.

FINDINGS

Fugl-Meyer Assessment correlated with motor performance (upper and lower limbs) and with movement quality (upper limb). Modified Ashworth scale correlated with movement quality (upper limb). Cutoff values of 9.0 cm in trunk anterior displacement and .57 m/s in gait velocity were estimated to differentiate participants with mild/moderate and severe compromise according to the Fugl-Meyer Assessment.

CONCLUSIONS

These results suggest that the Fugl-Meyer Assessment can be used to infer about motor performance and movement quality in chronic poststroke individuals with different levels of impairment.

Machine learning algorithms can classify outdoor terrain types during running using accelerometry data

BACKGROUND

Running is a popular physical activity that benefits health; however, running surface characteristics may influence loading impact and injury risk. Machine learning algorithms could automatically identify running surface from wearable motion sensors to quantify running exposures, and perhaps loading and injury risk for a runner.

RESEARCH QUESTION

(1) How accurately can machine learning algorithms identify surface type from three-dimensional accelerometer sensors? (2) Does the sensor count (single or two-sensor setup) affect model accuracy?

METHODS

Twenty-nine healthy adults (23.3 ± 3.6 years, 1.8 ± 0.1 m, and 63.6 ± 8.5 kg) participated in this study. Participants ran on three different surfaces (concrete, synthetic, woodchip) while fit with two three-dimensional accelerometers (lower-back and right tibia). Summary features (n = 208) were extracted from the accelerometer signals. Feature-based Gradient Boosting (GB) and signal-based deep learning Convolutional Neural Network (CNN) models were developed. Models were trained on 90% of the data and tested on the remaining 10%. The process was repeated five times, with data randomly shuffled between train-test splits, to quantify model performance variability.

RESULTS

All models and configurations achieved greater than 90% average accuracy. The highest performing models were the two-sensor GB and tibia-sensor CNN (average accuracy of 97.0 ± 0.7 and 96.1 ± 2.6%, respectively).

SIGNIFICANCE

Machine learning algorithms trained on running data from a single- or dual-sensor accelerometer setup can accurately distinguish between surfaces types. Automatic identification of surfaces encountered during running activities could help runners and coaches better monitor training load, improve performance, and reduce injury rates.

Sensitivity comparison of inertial to optical motion capture during gait: implications for tracking recovery

Wearable sensors provide a foundation for development of wearable robotic technology to be used in clinical applications. Inertial motion capture (IMC) has emerged as a viable alternative to more cumbersome, non-portable optical methods. Previous work has validated the accuracy of IMC for gait compared to optical motion capture (OMC). However, it is unclear how well IMC can measure the small changes in gait function needed to gauge recovery. In this study, we evaluate the sensitivity of IMC compared to OMC to small changes in gait on a cohort of unimpaired individuals during treadmill walking. Eight individuals walked on a split-belt treadmill in three-minute trials with five randomized conditions: right belt speed decrementing at 0.05 m/s from 1.0 m/s, all with left belt held at 1.0 m/s, simulating recovery of hemiparetic gait. We extracted the root mean square deviation (RMSD) of joint kinematics between limbs and within the limb with modulated gait speed as the main outcome measure. We used linear mixed models to identify differences in sensitivity to changes in gait asymmetry and gait speed. Based on these models, we estimated the minimal detectible interval in gait parameters. We found that IMC was capable of measuring a difference in gait speed of 0.08 m/s, roughly the equivalent of two weeks recovery progress. Statistically we could not conclude a difference of sensitivity between IMC and OMC, although there is a strong trend that IMC is more sensitive to changes in gait. We conclude that IMC is a valid tool to measure progress in gait kinematics over the course of recovery.

How to Select Balance Measures Sensitive to Parkinson's Disease from Body-Worn Inertial Sensors-Separating the Trees from the Forest

This study aimed to determine the most sensitive objective measures of balance dysfunction that differ between people with Parkinson’s Disease (PD) and healthy controls. One-hundred and forty-four people with PD and 79 age-matched healthy controls wore eight inertial sensors while performing tasks to measure five domains of balance: standing posture (Sway), anticipatory postural adjustments (APAs), automatic postural responses (APRs), dynamic posture (Gait) and limits of stability (LOS). To reduce the initial 93 measures, we selected uncorrelated measures that were most sensitive to PD. After applying a threshold on the Standardized Mean Difference between PD and healthy controls, 44 measures remained; and after reducing highly correlated measures, 24 measures remained. The four most sensitive measures were from APAs and Gait domains. The random forest with 10-fold cross-validation on the remaining measures (n = 24) showed an accuracy to separate PD from healthy controls of 82.4%—identical to result for all measures. Measures from the most sensitive domains, APAs and Gait, were significantly correlated with the severity of disease and with patient-related outcomes. This method greatly reduced the objective measures of balance to the most sensitive for PD, while still capturing four of the five domains of balance.

Gait analysis with the Kinect v2: normative study with healthy individuals and comprehensive study of its sensitivity, validity, and reliability in individuals with stroke

Background

Gait is usually assessed by clinical tests, which may have poor accuracy and be biased, or instrumented systems, which potentially solve these limitations at the cost of being time-consuming and expensive. The different versions of the Microsoft Kinect have enabled human motion tracking without using wearable sensors at a low-cost and with acceptable reliability. This study aims: First, to determine the sensitivity of an open-access Kinect v2-based gait analysis system to motor disability and aging; Second, to determine its concurrent validity with standardized clinical tests in individuals with stroke; Third, to quantify its inter and intra-rater reliability, standard error of measurement, minimal detectable change; And, finally, to investigate its ability to identify fall risk after stroke.

Methods

The most widely used spatiotemporal and kinematic gait parameters of 82 individuals post-stroke and 355 healthy subjects were estimated with the Kinect v2-based system. In addition, participants with stroke were assessed with the Dynamic Gait Index, the 1-min Walking Test, and the 10-m Walking Test.

Results

The system successfully characterized the performance of both groups. Significant concurrent validity with correlations of variable strength was detected between all clinical tests and gait measures. Excellent inter and intra-rater reliability was evidenced for almost all measures. Minimal detectable change was variable, with poorer results for kinematic parameters. Almost all gait parameters proved to identify fall risk.

Conclusions

Results suggest that although its limited sensitivity to kinematic parameters, the Kinect v2-based gait analysis could be used as a low-cost alternative to laboratory-grade systems to complement gait assessment in clinical settings.

Estimation of the walking speed of individuals with transfemoral amputation from a single prosthetic shank-mounted IMU

Microprocessor prosthetic knees, able to restore the gait of people with transfemoral amputation, are now often equipped with sensors embedded in the prosthetic shank, which could be used to assess some gait characteristics during real-life activities. In particular, an estimation of the walking speed during the locomotion of those subjects would be a relevant indicator of the performance. However, if methods have already been proposed in the literature to compute this walking speed, none are directly usable in this context and with this population. For these reasons, the current study proposed to estimate the instantaneous walking speed with a shank-embedded Inertial Measurement Units based on a biomechanical model of the prosthetic lower limb. Averaged walking speed estimation has been quantified for nine individuals with transfemoral amputation walking on a treadmill at different speeds and slopes when wearing an instrumented knee ankle prosthesis. Experimental results demonstrated the ability of the model to estimate the walking speed with an accuracy of 9% (normalized root mean squared errors over all the patients), which is consistent with previous reported walking speed estimation errors. In addition, as the walking speed estimation is instantaneous, the proposed method can provide the estimation by the end of the stance phase, which is an originality compared to other methods based on step length estimation. The present method is relevant for the estimation of walking speed during real-life activities of above-knee amputees opening the way to direct activity monitoring from the prosthesis.

Automated Accelerometer-Based Gait Event Detection During Multiple Running Conditions

The identification of the initial contact (IC) and toe off (TO) events are crucial components of running gait analyses. To evaluate running gait in real-world settings, robust gait event detection algorithms that are based on signals from wearable sensors are needed. In this study, algorithms for identifying gait events were developed for accelerometers that were placed on the foot and low back and validated against a gold standard force plate gait event detection method. These algorithms were automated to enable the processing of large quantities of data by accommodating variability in running patterns. An evaluation of the accuracy of the algorithms was done by comparing the magnitude and variability of the difference between the back and foot methods in different running conditions, including different speeds, foot strike patterns, and outdoor running surfaces. The results show the magnitude and variability of the back-foot difference was consistent across running conditions, suggesting that the gait event detection algorithms can be used in a variety of settings. As wearable technology allows for running gait analyses to move outside of the laboratory, the use of automated accelerometer-based gait event detection methods may be helpful in the real-time evaluation of running patterns in real world conditions.

Validation of inertial measurement units with optical tracking system in patients operated with Total hip arthroplasty

Background

Patient reported outcome measurement (PROMs) will not capture in detail the functional joint motion before and after total hip arthroplasty (THA). Therefore, methods more specifically aimed to analyse joint movements may be of interest. An analysis method that addresses these issues should be readily accessible and easy to use especially if applied to large groups of patients, who you want to study both before and after a surgical intervention such as THA. Our aim was to evaluate the accuracy of inertial measurement units (IMU) by comparison with an optical tracking system (OTS) to record pelvic tilt, hip and knee flexion in patients who had undergone THA.

Methods

49 subjects, 25 males 24 females, mean age of 73 years (range 51–80) with THA participated. All patients were measured with a portable IMU system, with sensors attached lateral to the pelvis, the thigh and the lower leg. For validation, a 12-camera motion capture system was used to determine the positions of 15 skin markers (Oqus 4, Qualisys AB, Sweden). Comparison of sagittal pelvic rotations, and hip and knee flexion-extension motions measured with the two systems was performed. The mean values of the IMU’s on the left and right sides were compared with OTS data.

Results

The comparison between the two gait analysis methods showed no significant difference for mean pelvic tilt range (4.9–5.4 degrees) or mean knee flexion range (54.4–55.1 degrees) on either side (p > 0.7). The IMU system did however record slightly less hip flexion on both sides (36.7–37.7 degrees for the OTS compared to 34.0–34.4 degrees for the IMU, p < 0.001).

Conclusions

We found that inertial measurement units can produce valid kinematic data of pelvis- and knee flexion-extension range. Slightly less hip flexion was however recorded with the inertial measurement units which may be due to the difference in the modelling of the pelvis, soft tissue artefacts, and malalignment between the two methods or misplacement of the inertial measurement units.

Trial registration

The study has ethical approval from the ethical committee “Regionala etikprövningsnämnden i Göteborg” (Dnr: 611–15, 2015-08-27) and all study participants have submitted written approval for participation in the study.

Validation of an Inertial Sensor Algorithm to Quantify Head and Trunk Movement in Healthy Young Adults and Individuals with Mild Traumatic Brain Injury

Wearable inertial measurement units (IMUs) may provide useful, objective information to clinicians interested in quantifying head movements as patients’ progress through vestibular rehabilitation. The purpose of this study was to validate an IMU-based algorithm against criterion data (motion capture) to estimate average head and trunk range of motion (ROM) and average peak velocity. Ten participants completed two trials of standing and walking tasks while moving the head with and without moving the trunk. Validity was assessed using a combination of Intra-class Correlation Coefficients (ICC), root mean square error (RMSE), and percent error. Bland-Altman plots were used to assess bias. Excellent agreement was found between the IMU and criterion data for head ROM and peak rotational velocity (average ICC > 0.9). The trunk showed good agreement for most conditions (average ICC > 0.8). Average RMSE for both ROM (head = 2.64°; trunk = 2.48°) and peak rotational velocity (head = 11.76 °/s; trunk = 7.37 °/s) was low. The average percent error was below 5% for head and trunk ROM and peak rotational velocity. No clear pattern of bias was found for any measure across conditions. Findings suggest IMUs may provide a promising solution for estimating head and trunk movement, and a practical solution for tracking progression throughout rehabilitation or home exercise monitoring.

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Objective assessment of an athlete's performance is of importance in elite sports to facilitate detailed analysis. The implementation of automated detection and recognition of sport-specific movements overcomes the limitations associated with manual performance analysis methods. The object of this study was to systematically review the literature on machine and deep learning for sport-specific movement recognition using inertial measurement unit (IMU) and, or computer vision data inputs. A search of multiple databases was undertaken. Included studies must have investigated a sport-specific movement and analysed via machine or deep learning methods for model development. A total of 52 studies met the inclusion and exclusion criteria. Data pre-processing, processing, model development and evaluation methods varied across the studies. Model development for movement recognition were predominantly undertaken using supervised classification approaches. A kernel form of the Support Vector Machine algorithm was used in 53% of IMU and 50% of vision-based studies. Twelve studies used a deep learning method as a form of Convolutional Neural Network algorithm and one study also adopted a Long Short Term Memory architecture in their model. The adaptation of experimental set-up, data pre-processing, and model development methods are best considered in relation to the characteristics of the targeted sports movement(s).

Validation of an accelerometer for measurement of activity in frail older people

BACKGROUND

Specific gait parameters are associated with falls and injury. It is important to identify walking episodes in order to determine the associated gait parameters. Frail older people have a greater risk of falling due to increased probability of inactivity. Therefore, detection and analysis of their physical activities becomes significant. Furthermore, ascertainment of gait parameters and non-sedentary activities for frail older group is difficult in free living environments - an area which hasn't been explored much.

METHODS

Participants were 23 older people residing in independent-living retirement homes. Data was inertial sensor signals, attached to the L5 vertebral area using a belt, from scripted activities (a timed up and go, and sit to stand activities) and unscripted activities of daily living collected in a free-living environment. An algorithm designed to identify walking, standing/sitting and lying is applied to the uSense wearable accelerometer data which has been analysed by processing the raw data with a gait detection algorithm and the results were compared against annotated videos which served as the gold standard. Validity of gait assessment was based on the percentage of agreement between the analysed accelerometer data and the corresponding reference video with 100Hz sampling frequency and 0.01 frames/second.

RESULTS

The median overall agreement between the processed accelerometer data and the annotated video was a match of approximately 92.8% and 95.1% for walking episodes for unscripted and scripted activities respectively.

SIGNIFICANCE

The tri-axial accelerometer with a sampling frequency of 100 Hz provides a valid measure of gait detection in frail older people aged above 75 years. Since a limited number of studies have reported the use of accelerometers for older people in a free-living context, performance evaluation and establishing the validity of body worn sensors for physical activity and gait recognition is the key goal achieved.

Template-DTW based on inertial signals: Preliminary results for step characterization

In this paper, we present a method for the creation of a library of inertial signals based on Dynamic Time Warping (DTW) for step characterization, with preliminary results in control subjects and patients with neurological diseases. Subjects performed a protocol including a 10 m straight walking, then turn back and walking for additional 10 m. The library is constructed with inertial signals (acceleration and angular velocities recorded in three directions) aligned with the DTW. Templates in the library are obtained for a specific cohort and for the different walking phases of the protocol. They are compared to the signal of a single subject by calculating a Pearson correlation coefficient. The method has been tested on a database of 864 exercises, obtained from 71 healthy controls, 24 patients with Parkinson disease and 48 patients with Radiation Induced Leukoencephalopathy (RIL). Pearson correlation classification reports a precision of about 85% for step detection. For exercise characterization the sensitivity is about 57%, 56% and 82% for Parkinson, RIL and control subjects respectively.

Validation of a Wearable IMU System for Gait Analysis: Protocol and Application to a New System

Miniaturized wearable Inertial Measurement Units (IMU) offer new opportunities for the functional assessment of motor functions for medicine, sport, and ergonomics. Sparse reliability validation studies have been conducted without a common specific approach and protocol. A set of guidelines to design validation protocol for these systems is proposed hereafter. They are based on the comparison between video analysis and the gold standard optoelectronic motion capture system for Gait Analysis (GA). A setup of the protocol has been applied to a wearable device implementing an inertial measurement unit and a dedicated harmonic oscillator kinematic model of the center of mass. In total, 10 healthy volunteers took part in the study, and four trials of walking at a self-selected speed and step length have been simultaneously recorded by the two systems, analyzed, and compared blindly (40 datasets). The model detects the steps and the foot which supports body weight. The stride time and the cadence have a mean absolute percentage error of 5.7% and 4.9%, respectively. The mean absolute percentage error in the measurement of step’s length and step’s speed is 5.6% and 13.5%, respectively. Results confirm that the proposed methodology is complete and effective. It is demonstrated that the developed wearable system allows for a reliable assessment of human gait spatio-temporal parameters. Therefore, the goal of this paper is threefold. The first goal is to present and define structured Protocol Design Guidelines, where the related setup is implemented for the validation of wearable IMU systems particularly dedicated to GA and gait monitoring. The second goal is to apply these Protocol Design Guidelines to a case study in order to verify their feasibility, user-friendliness, and efficacy. The third goal is the validation of our biomechanical kinematic model with the gold standard reference.

Good agreement between smart device and inertial sensor-based gait parameters during a 6-min walk

BACKGROUND

Traditional laboratory-based kinetic and kinematic gait analyses are expensive, time-intensive, and impractical for clinical settings. Inertial sensors have gained popularity in gait analysis research and more recently smart devices have been employed to provide quantification of gait. However, no study to date has investigated the agreement between smart device and inertial sensor-based gait parameters during prolonged walking.

RESEARCH QUESTION

Compare spatiotemporal gait metrics measured with a smart device versus previously validated inertial sensors.

METHODS

Twenty neurologically healthy young adults (7 women; age: 25.0 ± 3.7 years; BMI: 23.4 ± 2.9 kg/m²) performed a 6-min walk test (6MWT) wearing inertial sensors and smart devices to record stride duration, stride length, cadence, and gait speed. Pearson correlations were used to assess associations between spatiotemporal measures from the two devices and agreement between the two methods was assessed with Bland-Altman plots and limits of agreement.

RESULTS

All spatiotemporal gait metrics (stride duration, cadence, stride length and gait speed) showed strong (r>0.9) associations and good agreement between the two devices.

SIGNIFICANCE

Smart devices are capable of accurately reflecting many of the spatiotemporal gait metrics of inertial sensors. As the smart devices also accurately reflected individual leg output, future studies may apply this analytical strategy to clinical populations, to identify hallmarks of disability status and disease progression in a more ecologically valid environment.

Recognition of a Person Wearing Sport Shoes or High Heels through Gait Using Two Types of Sensors

Biometrics is currently an area that is both very interesting as well as rapidly growing. Among various types of biometrics the human gait recognition seems to be one of the most intriguing. However, one of the greatest problems within this field of biometrics is the change in gait caused by footwear. A change of shoes results in a significant lowering of accuracy in recognition of people. The following work presents a method which uses data gathered by two sensors: force plates and Microsoft Kinect v2 to reduce this problem. Microsoft Kinect is utilized to measure the body height of a person which allows the reduction of the set of recognized people only to those whose height is similar to that which has been measured. The entire process is preceded by identifying the type of footwear which the person is wearing. The research was conducted on data obtained from 99 people (more than 3400 strides) and the proposed method allowed us to reach a Correct Classification Rate (CCR) greater than 88% which, in comparison to earlier methods reaching CCR’s of <80%, is a significant improvement. The work presents advantages as well as limitations of the proposed method.

On the security of consumer wearable devices in the Internet of Things

Miniaturization of computer hardware and the demand for network capable devices has resulted in the emergence of a new class of technology called wearable computing. Wearable devices have many purposes like lifestyle support, health monitoring, fitness monitoring, entertainment, industrial uses, and gaming. Wearable devices are hurriedly being marketed in an attempt to capture an emerging market. Owing to this, some devices do not adequately address the need for security. To enable virtualization and connectivity wearable devices sense and transmit data, therefore it is essential that the device, its data and the user are protected. In this paper the use of novel Integrated Circuit Metric (ICMetric) technology for the provision of security in wearable devices has been suggested. ICMetric technology uses the features of a device to generate an identification which is then used for the provision of cryptographic services. This paper explores how a device ICMetric can be generated by using the accelerometer and gyroscope sensor. Since wearable devices often operate in a group setting the work also focuses on generating a group identification which is then used to deliver services like authentication, confidentiality, secure admission and symmetric key generation. Experiment and simulation results prove that the scheme offers high levels of security without compromising on resource demands.

Robust Stride Segmentation of Inertial Signals Based on Local Cyclicity Estimation

A novel approach for stride segmentation, gait sequence extraction, and gait event detection for inertial signals is presented. The approach operates by combining different local cyclicity estimators and sensor channels, and can additionally employ a priori knowledge on the fiducial points of gait events. The approach is universal as it can work on signals acquired by different inertial measurement unit (IMU) sensor types, is template-free, and operates unsupervised. A thorough evaluation was performed with two datasets: our own collected FRIgait dataset available for open use, containing long-term inertial measurements collected from 57 subjects using smartphones within the span of more than one year, and an FAU eGait dataset containing inertial data from shoe-mounted sensors collected from three cohorts of subjects: healthy, geriatric, and Parkinson’s disease patients. The evaluation was performed in controlled and uncontrolled conditions. When compared to the ground truth of the labelled FRIgait and eGait datasets, the results of our evaluation revealed the high robustness, efficiency (F-measure of about 98%), and accuracy (mean absolute error MAE in about the range of one sample) of the proposed approach. Based on these results, we conclude that the proposed approach shows great potential for its applicability in procedures and algorithms for movement analysis.

A Review on Accelerometry-Based Gait Analysis and Emerging Clinical Applications

Gait analysis continues to be an important technique for many clinical applications to diagnose and monitor certain diseases. Many mental and physical abnormalities cause measurable differences in a person's gait. Gait analysis has applications in sport, computer games, physical rehabilitation, clinical assessment, surveillance, human recognition, modeling, and many other fields. There are established methods using various sensors for gait analysis, of which accelerometers are one of the most often employed. Accelerometer sensors are generally more user friendly and less invasive. In this paper, we review research regarding accelerometer sensors used for gait analysis with particular focus on clinical applications. We provide a brief introduction to accelerometer theory followed by other popular sensing technologies. Commonly used gait phases and parameters are enumerated. The details of selecting the papers for review are provided. We also review several gait analysis software. Then we provide an extensive report of accelerometry-based gait analysis systems and applications, with additional emphasis on trunk accelerometry. We conclude this review with future research directions.

Flexible Piezoelectric Sensor-Based Gait Recognition

Most motion recognition research has required tight-fitting suits for precise sensing. However, tight-suit systems have difficulty adapting to real applications, because people normally wear loose clothes. In this paper, we propose a gait recognition system with flexible piezoelectric sensors in loose clothing. The gait recognition system does not directly sense lower-body angles. It does, however, detect the transition between standing and walking. Specifically, we use the signals from the flexible sensors attached to the knee and hip parts on loose pants. We detect the periodic motion component using the discrete time Fourier series from the signal during walking. We adapt the gait detection method to a real-time patient motion and posture monitoring system. In the monitoring system, the gait recognition operates well. Finally, we test the gait recognition system with 10 subjects, for which the proposed system successfully detects walking with a success rate over 93 %.

Camera pose estimation to improve accuracy and reliability of joint angles assessed with attitude and heading reference systems

Joint kinematics can be assessed using orientation estimates from Attitude and Heading Reference Systems (AHRS). However, magnetically-perturbed environments affect the accuracy of the estimated orientations. This study investigates, both in controlled and human mobility conditions, a trial calibration technic based on a 2D photograph with a pose estimation algorithm to correct initial difference in AHRS Inertial reference frames and improve joint angle accuracy. In controlled conditions, two AHRS were solidly affixed onto a wooden stick and a series of static and dynamic trials were performed in varying environments. Mean accuracy of relative orientation between the two AHRS was improved from 24.4° to 2.9° using the proposed correction method. In human conditions, AHRS were placed on the shank and the foot of a participant who performed repeated trials of straight walking and walking while turning, varying the level of magnetic perturbation in the starting environment and the walking speed. Mean joint orientation accuracy went from 6.7° to 2.8° using the correction algorithm. The impact of starting environment was also greatly reduced, up to a point where one could consider it as non-significant from a clinical point of view (maximum mean difference went from 8° to 0.6°). The results obtained demonstrate that the proposed method improves significantly the mean accuracy of AHRS joint orientation estimations in magnetically-perturbed environments and can be implemented in post processing of AHRS data collected during biomechanical evaluation of motion.

IMU-Based Gait Recognition Using Convolutional Neural Networks and Multi-Sensor Fusion

The wide spread usage of wearable sensors such as in smart watches has provided continuous access to valuable user generated data such as human motion that could be used to identify an individual based on his/her motion patterns such as, gait. Several methods have been suggested to extract various heuristic and high-level features from gait motion data to identify discriminative gait signatures and distinguish the target individual from others. However, the manual and hand crafted feature extraction is error prone and subjective. Furthermore, the motion data collected from inertial sensors have complex structure and the detachment between manual feature extraction module and the predictive learning models might limit the generalization capabilities. In this paper, we propose a novel approach for human gait identification using time-frequency (TF) expansion of human gait cycles in order to capture joint 2 dimensional (2D) spectral and temporal patterns of gait cycles. Then, we design a deep convolutional neural network (DCNN) learning to extract discriminative features from the 2D expanded gait cycles and jointly optimize the identification model and the spectro-temporal features in a discriminative fashion. We collect raw motion data from five inertial sensors placed at the chest, lower-back, right hand wrist, right knee, and right ankle of each human subject synchronously in order to investigate the impact of sensor location on the gait identification performance. We then present two methods for early (input level) and late (decision score level) multi-sensor fusion to improve the gait identification generalization performance. We specifically propose the minimum error score fusion (MESF) method that discriminatively learns the linear fusion weights of individual DCNN scores at the decision level by minimizing the error rate on the training data in an iterative manner. 10 subjects participated in this study and hence, the problem is a 10-class identification task. Based on our experimental results, 91% subject identification accuracy was achieved using the best individual IMU and 2DTF-DCNN. We then investigated our proposed early and late sensor fusion approaches, which improved the gait identification accuracy of the system to 93.36% and 97.06%, respectively.

Mechanical lifting energy consumption in work activities designed by means of the "revised NIOSH lifting equation"

The aims of the present work were: to calculate lifting energy consumption (LEC) in work activities designed to have a growing lifting index (LI) by means of revised NIOSH lifting equation; to evaluate the relationship between LEC and forces at the L₅-S₁ joint. The kinematic and kinetic data of 20 workers were recorded during the execution of lifting tasks in three conditions. We computed kinetic, potential and mechanical energy and the corresponding LEC by considering three different centers of mass of: 1) the load (CoM_L); 2) the multi-segment upper body model and load together (CoM_Upp+L); 3) the whole body and load together (CoM_Tot). We also estimated compression and shear forces. Results shows that LEC calculated for CoM_Upp+L and CoM_Tot grew significantly with the LI and that all the lifting condition pairs are discriminated. The correlation analysis highlighted a relationship between LEC and forces that determine injuries at the L₅-S₁ joint.

Myoelectric manifestation of muscle fatigue in repetitive work detected by means of miniaturized sEMG sensors

Upper limb work-related musculoskeletal disorders have a 12-month prevalence ranging from 12 to 41% worldwide and can be partly caused by handling low loads at high frequency. The association between the myoelectric manifestation of elbow flexor muscle fatigue and occupational physical demand has never been investigated. It was hypothesized that an elbow flexor muscle fatigue index could be a valid risk indicator in handling low loads at high frequency. This study aims to measure the myoelectric manifestation of muscle fatigue of the three elbow flexor muscles during the execution of the work tasks in different risk conditions. Fifteen right-handed healthy adults were screened using a movement analysis laboratory consisting of optoelectronic, dynamometer and surface electromyographic systems. The main result indicates that the fatigue index calculated from the brachioradialis is sensitive to the interaction among risk classes, session and gender, and above all it is sensitive to the risk classes.

Development and validity of methods for the estimation of temporal gait parameters from heel-attached inertial sensors in younger and older adults

The heel is likely a suitable location to which inertial sensors are attached for the detection of gait events. However, there are few studies to detect gait events and determine temporal gait parameters using sensors attached to the heels. We developed two methods to determine temporal gait parameters: detecting heel-contact using acceleration and detecting toe-off using angular velocity data (acceleration-angular velocity method; A-V method), and detecting both heel-contact and toe-off using angular velocity data (angular velocity-angular velocity method; V-V method). The aim of this study was to examine the concurrent validity of the A-V and V-V methods against the standard method, and to compare their accuracy. Temporal gait parameters were measured in 10 younger and 10 older adults. The intra-class correlation coefficients were excellent in both methods compared with the standard method (0.80 to 1.00). The root mean square errors of stance and swing time in the A-V method were smaller than the V-V method in older adults, although there were no significant discrepancies in the other comparisons. Our study suggests that inertial sensors attached to the heels, using the A-V method in particular, provide a valid measurement of temporal gait parameters.

A ROBUST AUTOMATIC GAIT MONITORING APPROACH USING A SINGLE IMU FOR HOME-BASED APPLICATIONS

A new approach of human activity monitoring with a single Inertial Measurement Unit (IMU) capable of gait recognition and assessment is proposed for home-based applications. The method estimates the foot motion using a single IMU, then automatically segments the motion into steps, and extracts multiple kinematics templates. It classifies each segment by extracting Mahalanobis distance-based features from multiple sections of the motion templates and then training a Support Vector Machine. The proposed wearable system can distinguish between nine classes of activities with a classification accuracy of 99.6%. It can also discriminate between normal and abnormal gait patterns with an accuracy of 98.7%. In addition to a high recognition rate, the proposed approach provides a Gait Similarity Score (GSS) of the performed gait to its desired/normal pattern. The experimental results indicate the capability of GSS measure for assessing the quality of motion in “pre-”, “initial”, “mid” and “terminal” stages of swing phase.

Inertial measurement systems for segments and joints kinematics assessment: towards an understanding of the variations in sensors accuracy

BACKGROUND

Joints kinematics assessment based on inertial measurement systems, which include attitude and heading reference system (AHRS), are quickly gaining in popularity for research and clinical applications. The variety of the tasks and contexts they are used in require a deep understanding of the AHRS accuracy for optimal data interpretation. However, published accuracy studies on AHRS are mostly limited to a single task measured on a limited number of segments and participants. This study assessed AHRS sensors kinematics accuracy at multiple segments and joints through a variety of tasks not only to characterize the system's accuracy in these specific conditions, but also to extrapolate the accuracy results to a broader range of conditions using the characteristics of the movements (i.e. velocity and type of motion). Twenty asymptomatic adults ([Formula: see text] = 49.9) performed multiple 5 m timed up and go. Participants' head, upper trunk, pelvis, thigh, shank and foot were simultaneously tracked using AHRS and an optical motion capture system (gold standard). Each trial was segmented into basic tasks (sit-to-stand, walk, turn).

RESULTS

At segment level, results revealed a mean root-mean-squared-difference [Formula: see text] varying between 1.1° and 5.5° according to the segment tracked and the task performed, with a good to excellent agreement between the systems. Relative sensor kinematics accuracy (i.e. joint) varied between 1.6° and 13.6° over the same tasks. On a global scheme, analysis of the effect of velocity on sensor kinematics accuracy showed that AHRS are better adapted to motions performed between 50°/s and 75°/s (roughly thigh and shank while walking).

CONCLUSION

Results confirmed that pairing of modules to obtain joint kinematics affects the accuracy compared to segment kinematics. Overall, AHRS are a suitable solution for clinical evaluation of biomechanics under the multi-segment tasks performed although the variation in accuracy should be taken into consideration when judging the clinical meaningfulness of the observed changes.