L Test Subtask Segmentation for Lower-Limb Amputees Using a Random Forest Algorithm.

Functional mobility tests, such as the L test of functional mobility, are recommended to provide clinicians with information regarding the mobility progress of lower-limb amputees. Smartphone inertial sensors have been used to perform subtask segmentation on functional mobility tests, providing further clinically useful measures such as fall risk. However, L test subtask segmentation rule-based algorithms developed for able-bodied individuals have not produced sufficiently acceptable results when tested with lower-limb amputee data. In this paper, a random forest machine learning model was trained to segment subtasks of the L test for application to lower-limb amputees. The model was trained with 105 trials completed by able-bodied participants and 25 trials completed by lower-limb amputee participants and tested using a leave-one-out method with lower-limb amputees. This algorithm successfully classified subtasks within a one-foot strike for most lower-limb amputee participants. The algorithm produced acceptable results to enhance clinician understanding of a person's mobility status (>85% accuracy, >75% sensitivity, >95% specificity).

Balance confidence classification in people with a lower limb amputation using six minute walk test smartphone sensor signals.

The activities-specific balance confidence scale (ABC) assesses balance confidence during common activities. While low balance confidence can result in activity avoidance, excess confidence can increase fall risk. People with lower limb amputations can present with inconsistent gait, adversely affecting their balance confidence. Previous research demonstrated that clinical outcomes in this population (e.g., stride parameters, fall risk) can be determined from smartphone signals collected during walk tests, but this has not been evaluated for balance confidence. Fifty-eight (58) individuals with lower limb amputation completed a six-minute walk test (6MWT) while a smartphone at the posterior pelvis was used for signal collection. Participant ABC scores were categorized as low confidence or high confidence. A random forest classified ABC groups using features from each step, calculated from smartphone signals. The random forest correctly classified the confidence level of 47 of 58 participants (accuracy 81.0%, sensitivity 63.2%, specificity 89.7%). This research demonstrated that smartphone signal data can classify people with lower limb amputations into balance confidence groups after completing a 6MWT. Integration of this model into the TOHRC Walk Test app would provide balance confidence classification, in addition to previously demonstrated clinical outcomes, after completing a single assessment and could inform individualized rehabilitation programs to improve confidence and prevent activity avoidance.

Design and prototype validation of a laterally mounted powered hip joint prothesis.

Prosthetic technology has advanced with the development of powered prostheses to enhance joint function and movement in the absence of native anatomy. However, there are no powered solutions available for hip-level amputees, and most existing hip prostheses are mounted to the front of the prosthetic socket, thereby limiting range of motion. This research introduces a novel laterally mounted powered hip joint (LMPHJ) that augments user movement. The LMPHJ is mounted on the lateral side of the prosthetic socket, positioning the hip joint closer to the anatomical center of rotation while ensuring user safety and stability. The motor and electronics are located in the thigh area, maintaining a low profile while transmitting the required hip moment to the mechanical joint center of rotation. A prototype was designed and manufactured, and static testing was complete by modifying the loading conditions defined in the ISO 15032:2000 standard to failure test levels for a 100 kg person, demonstrating the joint's ability to withstand everyday loading conditions. Functional testing was conducted using a prosthesis simulator that enabled able-bodied participants to successfully walk with the powered prosthesis on level ground. This validates the mechanical design for walking and indicates the LMPHJ is ready for evaluation in the next phase with hip disarticulation amputee participants.

Multivariable Prediction Models for Traumatic Spinal Cord Injury: A Systematic Review.

Background: Traumatic spinal cord injuries (TSCI) greatly affect the lives of patients and their families. Prognostication may improve treatment strategies, health care resource allocation, and counseling. Multivariable clinical prediction models (CPMs) for prognosis are tools that can estimate an absolute risk or probability that an outcome will occur. Objectives: We sought to systematically review the existing literature on CPMs for TSCI and critically examine the predictor selection methods used. Methods: We searched MEDLINE, PubMed, Embase, Scopus, and IEEE for English peer-reviewed studies and relevant references that developed multivariable CPMs to prognosticate patient-centered outcomes in adults with TSCI. Using narrative synthesis, we summarized the characteristics of the included studies and their CPMs, focusing on the predictor selection process. Results: We screened 663 titles and abstracts; of these, 21 full-text studies (2009-2020) consisting of 33 distinct CPMs were included. The data analysis domain was most commonly at a high risk of bias when assessed for methodological quality. Model presentation formats were inconsistently included with published CPMs; only two studies followed established guidelines for transparent reporting of multivariable prediction models. Authors frequently cited previous literature for their initial selection of predictors, and stepwise selection was the most frequent predictor selection method during modelling. Conclusion: Prediction modelling studies for TSCI serve clinicians who counsel patients, researchers aiming to risk-stratify participants for clinical trials, and patients coping with their injury. Poor methodological rigor in data analysis, inconsistent transparent reporting, and a lack of model presentation formats are vital areas for improvement in TSCI CPM research.

A pelvic kinematic approach for calculating hip angles for active hip disarticulation prosthesis control.

BACKGROUND: Control system design for a microprocessor-controlled hip-knee-ankle-foot (HKAF) prosthesis is a challenge since hip disarticulation amputees lack the entire leg and, therefore, only have pelvis movement as user-guided input. This research proposes a method for determining hip joint angles from pelvis movement in a control system for the next generation of powered prostheses. METHOD: Three-dimensional pelvic motion and stance time of 10 transfemoral (TF) prosthetic users were used to identify important features and to develop an algorithm to calculate hip angles from pelvis movement based on correlation and linear regression results. The algorithm was then applied to a separate (independent) TF group to validate algorithm performance. RESULTS: The proposed algorithm calculated viable hip angles during walking by utilizing pelvic rotation, pelvic tilt, and stance time. Small angular differences were found between the algorithm results and motion capture data. The greatest difference was for hip maximum extension angle (2.5 ± 2.0°). CONCLUSIONS: Since differences between algorithm output and motion data were within participant standard deviations, the developed algorithm could be used to determine the desired hip angle from pelvis movements. This study will aid the future development of gait control systems for new active HKAF prostheses.

A data-driven approach to categorize patients with traumatic spinal cord injury: cluster analysis of a multicentre database.

Background: Conducting clinical trials for traumatic spinal cord injury (tSCI) presents challenges due to patient heterogeneity. Identifying clinically similar subgroups using patient demographics and baseline injury characteristics could lead to better patient-centered care and integrated care delivery. Purpose: We sought to (1) apply an unsupervised machine learning approach of cluster analysis to identify subgroups of tSCI patients using patient demographics and injury characteristics at baseline, (2) to find clinical similarity within subgroups using etiological variables and outcome variables, and (3) to create multi-dimensional labels for categorizing patients. Study design: Retrospective analysis using prospectively collected data from a large national multicenter SCI registry. Methods: A method of spectral clustering was used to identify patient subgroups based on the following baseline variables collected since admission until rehabilitation: location of the injury, severity of the injury, Functional Independence Measure (FIM) motor, and demographic data (age, and body mass index). The FIM motor score, the FIM motor score change, and the total length of stay were assessed on the subgroups as outcome variables at discharge to establish the clinical similarity of the patients within derived subgroups. Furthermore, we discussed the relevance of the identified subgroups based on the etiological variables (energy and mechanism of injury) and compared them with the literature. Our study also employed a qualitative approach to systematically describe the identified subgroups, crafting multi-dimensional labels to highlight distinguishing factors and patient-focused insights. Results: Data on 334 tSCI patients from the Rick Hansen Spinal Cord Injury Registry was analyzed. Five significantly different subgroups were identified (p-value ≤0.05) based on baseline variables. Outcome variables at discharge superimposed on these subgroups had statistically different values between them (p-value ≤0.05) and supported the notion of clinical similarity of patients within each subgroup. Conclusion: Utilizing cluster analysis, we identified five clinically similar subgroups of tSCI patients at baseline, yielding statistically significant inter-group differences in clinical outcomes. These subgroups offer a novel, data-driven categorization of tSCI patients which aligns with their demographics and injury characteristics. As it also correlates with traditional tSCI classifications, this categorization could lead to improved personalized patient-centered care.

Automated Implementation of the Edinburgh Visual Gait Score (EVGS) Using OpenPose and Handheld Smartphone Video.

Recent advancements in computing and artificial intelligence (AI) make it possible to quantitatively evaluate human movement using digital video, thereby opening the possibility of more accessible gait analysis. The Edinburgh Visual Gait Score (EVGS) is an effective tool for observational gait analysis, but human scoring of videos can take over 20 min and requires experienced observers. This research developed an algorithmic implementation of the EVGS from handheld smartphone video to enable automatic scoring. Participant walking was video recorded at 60 Hz using a smartphone, and body keypoints were identified using the OpenPose BODY25 pose estimation model. An algorithm was developed to identify foot events and strides, and EVGS parameters were determined at relevant gait events. Stride detection was accurate within two to five frames. The level of agreement between the algorithmic and human reviewer EVGS results was strong for 14 of 17 parameters, and the algorithmic EVGS results were highly correlated (r > 0.80, "r" represents the Pearson correlation coefficient) to the ground truth values for 8 of the 17 parameters. This approach could make gait analysis more accessible and cost-effective, particularly in areas without gait assessment expertise. These findings pave the way for future studies to explore the use of smartphone video and AI algorithms in remote gait analysis.

Development of an Integrated Powered Hip and Microprocessor-Controlled Knee for a Hip-Knee-Ankle-Foot Prosthesis.

Hip-knee-ankle-foot prostheses (HKAF) are full lower-limb devices for people with hip amputations that enable individuals to regain their mobility and move freely within their chosen environment. HKAFs typically have high rejection rates among users, as well as gait asymmetry, increased trunk anterior-posterior lean, and increased pelvic tilt. A novel integrated hip-knee (IHK) unit was designed and evaluated to address the limitations of existing solutions. This IHK combines powered hip and microprocessor-controlled knee joints into one structure, with shared electronics, sensors, and batteries. The unit is also adjustable to user leg length and alignment. ISO-10328:2016 standard mechanical proof load testing demonstrated acceptable structural safety and rigidity. Successful functional testing involved three able-bodied participants walking with the IHK in a hip prosthesis simulator. Hip, knee, and pelvic tilt angles were recorded and stride parameters were analyzed from video recordings. Participants were able to walk independently using the IHK and data showed that participants used different walking strategies. Future development of the thigh unit should include completion of a synergistic gait control system, improved battery-holding mechanism, and amputee user testing.

Design and evaluation of a hip prosthesis simulator: A technical note.

People with a limb loss at the level of the hip or pelvis have the most difficulty returning to walking compared with those with a lower amputation. This is because their prosthesis must replace the hip, knee, and ankle joints. An adjustable hip-disarticulation/hemipelvectomy prosthesis simulator that allows able-bodied individuals to wear and assess a prosthesis can help researchers and manufacturers when designing new prosthetic components (ie, hip joints). SolidWorks computer-aided design software was used to design and simulate an adapter that can connect prosthetic components to an off-the-shelf hip abduction orthosis. The adapter was made of 1020 stainless steel and aluminium 6061-T3 with a yield strength of 276 MPa. To confirm that this adapter is strong and safe for ambulation, mechanical testing was performed using an INSTRON machine. The maximum loads generated in any activity were chosen according to the International Organization for Standardization 15032:2000 standard for hip disarticulation external prostheses. The designed adapter allowed frontal, lateral, or distal mounting of different prosthetic hip joints. Mechanical testing confirmed that the new adapter can withstand forces and moments experienced during ambulation. The hip disarticulation/hemipelvectomy prosthesis simulator is easy to use and adjustable based on each person's height and pelvic width. Furthermore, this simulator would assist rehabilitation practitioners in experiencing the use of hip-level prostheses and give them a better understanding of people using such technologies. The next step in this project is to evaluate able-bodied participant gait while using this hip simulator prosthesis with different hip joints.

Cross slope gait biomechanics for individuals with and without a unilateral transtibial amputation.

BACKGROUND: This research was conducted to better understand compensatory strategies during cross-slope walking for adults with and without a unilateral transtibial amputation. METHODS: Fourteen individuals with unilateral transtibial amputation and 14 individuals with no lower limb amputation participated in this study. Motion and force data were captured while participants walked on a treadmill in a virtual reality environment for level and ± 5° cross slopes. Temporal-spatial parameters, kinematics (ankle, knee, hip, pelvis, trunk), and ground reaction forces were examined. FINDINGS: Compared to level, participants had similar step width but slightly longer steps for top-cross-slope and slightly shorter steps for bottom-cross-slope. Top-cross-slope required a more flexed limb with ankle eversion, and bottom-cross-slope required a more extended limb with ankle inversion. Participants had similar lateral pelvis and trunk motion for all walking conditions, but slightly more anterior trunk lean for top cross-slope with more anterior trunk lean observed for individuals with a lower limb amputation than without lower limb amputation. Participants with a lower limb amputation compensated for limited prosthetic ankle-foot dorsiflexion on the top-cross-slope by increasing prosthetic side hip flexion, reducing intact ankle/knee flexion, and increasing intact push-off force. INTERPRETATION: Gait adaptations during cross-slope walking were primarily in the lower extremities and were largely similar for those with and without a transtibial amputation. The information presented in this paper provides a better understanding of gait strategies adopted during cross-slope walking and can guide researchers and industry in prosthetic development.

Amputee Fall Risk Classification Using Machine Learning and Smartphone Sensor Data from 2-Minute and 6-Minute Walk Tests.

The 6-min walk test (6MWT) is commonly used to assess a person's physical mobility and aerobic capacity. However, richer knowledge can be extracted from movement assessments using artificial intelligence (AI) models, such as fall risk status. The 2-min walk test (2MWT) is an alternate assessment for people with reduced mobility who cannot complete the full 6MWT, including some people with lower limb amputations; therefore, this research investigated automated foot strike (FS) detection and fall risk classification using data from a 2MWT. A long short-term memory (LSTM) model was used for automated foot strike detection using retrospective data (n = 80) collected with the Ottawa Hospital Rehabilitation Centre (TOHRC) Walk Test app during a 6-min walk test (6MWT). To identify FS, an LSTM was trained on the entire six minutes of data, then re-trained on the first two minutes of data. The validation set for both models was ground truth FS labels from the first two minutes of data. FS identification with the 6-min model had 99.2% accuracy, 91.7% sensitivity, 99.4% specificity, and 82.7% precision. The 2-min model achieved 98.0% accuracy, 65.0% sensitivity, 99.1% specificity, and 68.6% precision. To classify fall risk, a random forest model was trained on step-based features calculated using manually labeled FS and automated FS identified from the first two minutes of data. Automated FS from the first two minutes of data correctly classified fall risk for 61 of 80 (76.3%) participants; however, <50% of participants who fell within the past six months were correctly classified. This research evaluated a novel method for automated foot strike identification in lower limb amputee populations that can be applied to both 6MWT and 2MWT data to calculate stride parameters. Features calculated using automated FS from two minutes of data could not sufficiently classify fall risk in lower limb amputees.

Automated step detection with 6-minute walk test smartphone sensors signals for fall risk classification in lower limb amputees.

Predictive models for fall risk classification are valuable for early identification and intervention. However, lower limb amputees are often neglected in fall risk research despite having increased fall risk compared to age-matched able-bodied individuals. A random forest model was previously shown to be effective for fall risk classification of lower limb amputees, however manual labelling of foot strikes was required. In this paper, fall risk classification is evaluated using the random forest model, using a recently developed automated foot strike detection approach. 80 participants (27 fallers, 53 non-fallers) with lower limb amputations completed a six-minute walk test (6MWT) with a smartphone at the posterior pelvis. Smartphone signals were collected with The Ottawa Hospital Rehabilitation Centre (TOHRC) Walk Test app. Automated foot strike detection was completed using a novel Long Short-Term Memory (LSTM) approach. Step-based features were calculated using manually labelled or automated foot strikes. Manually labelled foot strikes correctly classified fall risk for 64 of 80 participants (accuracy 80%, sensitivity 55.6%, specificity 92.5%). Automated foot strikes correctly classified 58 of 80 participants (accuracy 72.5%, sensitivity 55.6%, specificity 81.1%). Both approaches had equivalent fall risk classification results, but automated foot strikes had 6 more false positives. This research demonstrates that automated foot strikes from a 6MWT can be used to calculate step-based features for fall risk classification in lower limb amputees. Automated foot strike detection and fall risk classification could be integrated into a smartphone app to provide clinical assessment immediately after a 6MWT.

Comparison of Decision Tree and Long Short-Term Memory Approaches for Automated Foot Strike Detection in Lower Extremity Amputee Populations.

Foot strike detection is important when evaluating a person's gait characteristics. Accelerometer and gyroscope signals from smartphones have been used to train artificial intelligence (AI) models for automated foot strike detection in able-bodied and elderly populations. However, there is limited research on foot strike detection in lower limb amputees, who have a more variable and asymmetric gait. A novel method for automated foot strike detection in lower limb amputees was developed using raw accelerometer and gyroscope signals collected from a smartphone positioned at the posterior pelvis. Raw signals were used to train a decision tree model and long short-term memory (LSTM) model for automated foot strike detection. These models were developed using retrospective data (n = 72) collected with the TOHRC Walk Test app during a 6-min walk test (6MWT). An Android smartphone was placed on a posterior belt for each participant during the 6MWT to collect accelerometer and gyroscope signals at 50 Hz. The best model for foot strike identification was the LSTM with 100 hidden nodes in the LSTM layer, 50 hidden nodes in the dense layer, and a batch size of 64 (99.0% accuracy, 86.4% sensitivity, 99.4% specificity, and 83.7% precision). This research created a novel method for automated foot strike identification in lower extremity amputee populations that is equivalent to manual labelling and accessible for clinical use. Automated foot strike detection is required for stride analysis and to enable other AI applications, such as fall detection.

Hip disarticulation and hemipelvectomy prostheses: A review of the literature.

BACKGROUND: Although the global population of people with a hip disarticulation (HD) or hemipelvectomy (HP) amputation is small, the degree of disability is high, affecting function and independence. A comprehensive literature review is needed to examine the evidence for prostheses in these amputation levels. METHOD: A scoping literature review was conducted to examine related research documents from 1950 to September 2020, found using Scopus, Web of Science, PubMed, and Google Scholar databases. Studies evaluated (retrospectively or prospectively) HD or HP prostheses and were written in English. Study design and protocol, research instrument, sample size, and outcome measures were reviewed. RESULTS: In the past 70 years, 53 articles that evaluated HD or HP prostheses were published. Most research was conducted in the United States (24 articles) and Japan (nine articles). In 42 articles, authors prospectively evaluated the effects of prostheses in these amputation levels. On average, prospective studies had four (SD = 5) participants. Since 1950, only five prospective studies evaluated HD or HP prostheses with 10 or more participants. Moreover, sufficient information was often unavailable for research replication. CONCLUSION: More evidence is needed regarding the effects of HD or HP prosthetic components (i.e. hip, knee, ankle, socket type, and suspension system) on gait, patient satisfaction, prosthetic use, interface pressure, and energy expenditure. Articles mostly have small sample sizes that reduce confidence in the reliability of their findings and limit generalizability. Future investigations are needed with vigorous methodology and larger sample sizes to provide strong statistical conclusions.

Fall risk classification for people with lower extremity amputations using random forests and smartphone sensor features from a 6-minute walk test.

Fall-risk classification is a challenging but necessary task to enable the recommendation of preventative programs for individuals identified at risk for falling. Existing research has primarily focused on older adults, with no predictive fall-risk models for lower limb amputees, despite their greater likelihood of fall-risk than older adults. In this study, 89 amputees with varying degrees of lower limb amputation were asked if they had fallen in the past 6 months. Those who reported at least one fall were considered a fall risk. Each participant performed a 6 minute walk test (6MWT) with an Android smartphone placed in a holder located on the back of the pelvis. A fall-risk classification method was developed using data from sensors within the smartphone. The Ottawa Hospital Rehabilitation Center Walk Test app captured accelerometer and gyroscope data during the 6MWT. From this data, foot strikes were identified, and 248 features were extracted from the collection of steps. Steps were segmented into turn and straight walking, and four different data sets were created: turn steps, straightaway steps, straightaway and turn steps, and all steps. From these, three feature selection techniques (correlation-based feature selection, relief F, and extra trees classifier ensemble) were used to eliminate redundant or ineffective features. Each feature subset was tested with a random forest classifier and optimized for the best number of trees. The best model used turn data, with three features selected by Correlation-based feature selection (CFS), and used 500 trees in a random forest classifier. The resulting metrics were 81.3% accuracy, 57.2% sensitivity, 94.9% specificity, a Matthews correlation coefficient of 0.587, and an F1 score of 0.83. Since the outcomes are comparable to metrics achieved by existing clinical tests, the classifier may be viable for use in clinical practice.

Development of a Smart Hallway for Marker-Less Human Foot Tracking and Stride Analysis.

OBJECTIVE: In this research, a marker-less 'smart hallway' is proposed where stride parameters are computed as a person walks through an institutional hallway. Stride analysis is a viable tool for identifying mobility changes, classifying abnormal gait, estimating fall risk, monitoring progression of rehabilitation programs, and indicating progression of nervous system related disorders. METHODS: Smart hallway was build using multiple Intel RealSense D415 depth cameras. A novel algorithm was developed to track a human foot using combined point cloud data obtained from the smart hallway. A method was implemented to separate the left and right leg point cloud data, then find the average foot dimensions. Foot tracking was achieved by fitting a box with average foot dimensions to the foot, with the box's base on the foot's bottom plane. A smart hallway with this novel foot tracking algorithm was tested with 22 able-bodied volunteers by comparing marker-less system stride parameters with Vicon motion analysis output. RESULTS: With smart hallway frame rate at approximately 60fps, temporal stride parameter absolute mean differences were less than 30ms. Random noise around the foot's point cloud was observed, especially during foot strike phases. This caused errors in medial-lateral axis dependent parameters such as step width and foot angle. Anterior-posterior dependent (stride length, step length) absolute mean differences were less than 25mm. CONCLUSION: This novel marker-less smart hallway approach delivered promising results for stride analysis with small errors for temporal stride parameters, anterior-posterior stride parameters, and reasonable errors for medial-lateral spatial parameters.

Maintaining stable transtibial amputee gait on level and simulated uneven conditions in a virtual environment.

PURPOSE: This research compares gait strategies to maintain stable gait over a variety of non-level walking conditions for individuals with a transtibial amputation and able-bodied individuals. METHODS: Twelve people with unilateral transtibial amputation and twelve able-bodied individuals walked on a self-paced treadmill in a park-like virtual environment with level and continuous perturbation conditions. Walking stability was quantified by margin-of-stability, step parameters (walking speed, temporal and spatial parameters, and foot clearance), and gait variability (standard deviations for margin-of-stability, step parameters, and root-mean-square of trunk acceleration). RESULTS AND CONCLUSIONS: For non-level conditions, able-bodied and transtibial groups had greater root-mean-square of trunk acceleration and walked with a cautious and variable step strategy by changing speed, step width, foot clearance, margin-of-stability, and increasing step variability. Overall, able-bodied and transtibial amputee participants adopted similar strategies to maintain stable gait over non-level conditions, but the amputee group was more variable than the able-bodied group. These results demonstrated the importance of measuring gait variability, including trunk acceleration and step variability measures, when quantitatively assessing mobility for individuals with a transtibial amputation. Implications for rehabilitation Able-bodied and transtibial amputee groups adapted gait biomechanics for simulated uneven conditions. Adaptations for non-level conditions included increasing step width, margin-of stability, minimum foot clearance, and varying speed. Gait was also more variable for non-level conditions, with greater variability for transtibial amputee participants compared to able-bodied participants. These results highlight the importance of measuring variability when performing comprehensive walking assessment, particularly for active individuals who achieve maximal performance on standard assessments yet report functional limitations in daily living.

Classification of Aggressive Movements Using Smartwatches.

Recognizing aggressive movements is a challenging task in human activity recognition. Wearable smartwatch technology with machine learning may be a viable approach for human aggressive behavior classification. This research identified a viable classification model and feature selector (CM-FS) combination for separating aggressive from non-aggressive movements using smartwatch data and determined if only one smartwatch is sufficient for this task. A ranking method was used to select relevant CM-FS models across accuracy, sensitivity, specificity, precision, F-score, and Matthews correlation coefficient (MCC). The Waikato environment for knowledge analysis (WEKA) was used to run 6 machine learning classifiers (random forest, k-nearest neighbors (kNN), multilayer perceptron neural network (MP), support vector machine, naïve Bayes, decision tree) coupled with three feature selectors (ReliefF, InfoGain, Correlation). Microsoft Band 2 accelerometer and gyroscope data were collected during an activity circuit that included aggressive (punching, shoving, slapping, shaking) and non-aggressive (clapping hands, waving, handshaking, opening/closing a door, typing on a keyboard) tasks. A combination of kNN and ReliefF was the best CM-FS model for separating aggressive actions from non-aggressive actions, with 99.6% accuracy, 98.4% sensitivity, 99.8% specificity, 98.9% precision, 0.987 F-score, and 0.984 MCC. kNN and random forest classifiers, combined with any of the feature selectors, generated the top models. Models with naïve Bayes or support vector machines had poor performance for sensitivity, F-score, and MCC. Wearing the smartwatch on the dominant wrist produced the best single-watch results. The kNN and ReliefF combination demonstrated that this smartwatch-based approach is a viable solution for identifying aggressive behavior. This wrist-based wearable sensor approach could be used by care providers in settings where people suffer from dementia or mental health disorders, where random aggressive behaviors often occur.

Fall-Risk Classification in Amputees Using Smartphone Sensor Based Features in Turns.

Identifying people at risk of falling can prevent life altering injury. Existing research has demonstrated fall-risk classifier effectiveness in older adults from accelerometer-based data. The amputee population should similarly benefit from these classification techniques; however, validation is still required. 83 individuals with varying levels of lower limb amputation performed a six-minute walk test while wearing an Android smartphone on their posterior belt, with TOHRC Walk Test app to capture accelerometer and gyroscope data. A random forest classifier was applied to feature subsets found using three feature selection techniques. The feature subset with the greatest accuracy (78.3%), sensitivity (62.1%), and Matthews Correlation Coefficient (0.51) was selected by Correlation-based Feature Selection. The peak distinction feature was chosen by all feature selectors. Classification outcomes with this lower extremity amputee group were similar to results from elderly faller classification research. The 62.1% sensitivity and 87.0% specificity would make this approach viable in practice, but further research is needed to improve faller classification results.

Augmented Reality Approach for Marker-based Posture Measurement on Smartphones.

Marker tracking for postural and range of motion (ROM) measurements transcends multiple disciplines (e.g., healthcare, ergonomics, engineering). A viable real-time mobile application is currently lacking for measuring limb angles and body posture. To address this need, a novel Android smartphone augmented-reality-based application was developed using the AprilTag2 fiducial marker system. To evaluate the app, two markers were printed on paper and attached to a wall. A Samsung S6 mobile phone was fixed on a tripod, parallel to the wall. The smartphone app tracked and recorded marker orientation and 2D position data in the camera frame, from front and rear cameras, for different smartphone placements. The average error between mobile phone and measured angles was less than 1 degree for all test settings (back camera=0.29°, front camera=0.33°, yaw rotation=0.75°, tilt rotation=0.22°). The average error between mobile phone and measured distance was less than 4 mm for all test settings (back camera=1.8 mm, front camera=2.5 mm, yaw rotation=3 mm, tilt rotation=3.8 mm). Overall, the app obtained valid and reliable angle and distance measurements with smartphone positions and cameras that would be expected in practice. Thus, this app is viable for clinical ROM and posture assessments.