Wearable airbag technology and machine learned models to mitigate falls after stroke.

BACKGROUND: Falls are a common complication experienced after a stroke and can cause serious detriments to physical health and social mobility, necessitating a dire need for intervention. Among recent advancements, wearable airbag technology has been designed to detect and mitigate fall impact. However, these devices have not been designed nor validated for the stroke population and thus, may inadequately detect falls in individuals with stroke-related motor impairments. To address this gap, we investigated whether population-specific training data and modeling parameters are required to pre-detect falls in a chronic stroke population. METHODS: We collected data from a wearable airbag's inertial measurement units (IMUs) from individuals with (n = 20 stroke) and without (n = 15 control) history of stroke while performing a series of falls (842 falls total) and non-falls (961 non-falls total) in a laboratory setting. A leave-one-subject-out crossvalidation was used to compare the performance of two identical machine learned models (adaptive boosting classifier) trained on cohort-dependent data (control or stroke) to pre-detect falls in the stroke cohort. RESULTS: The average performance of the model trained on stroke data (recall = 0.905, precision = 0.900) had statistically significantly better recall (P = 0.0035) than the model trained on control data (recall = 0.800, precision = 0.944), while precision was not statistically significantly different. Stratifying models trained on specific fall types revealed differences in pre-detecting anterior-posterior (AP) falls (stroke-trained model's F1-score was 35% higher, P = 0.019). Using activities of daily living as non-falls training data (compared to near-falls) significantly increased the AUC (Area under the receiver operating characteristic) for classifying AP falls for both models (P < 0.04). Preliminary analysis suggests that users with more severe stroke impairments benefit further from a stroke-trained model. The optimal lead time (time interval pre-impact to detect falls) differed between control- and stroke-trained models. CONCLUSIONS: These results demonstrate the importance of population sensitivity, non-falls data, and optimal lead time for machine learned pre-impact fall detection specific to stroke. Existing fall mitigation technologies should be challenged to include data of neurologically impaired individuals in model development to adequately detect falls in other high fall risk populations. Trial registration https://clinicaltrials.gov/ct2/show/NCT05076565 ; Unique Identifier: NCT05076565. Retrospectively registered on 13 October 2021.

A smartphone-based online system for fall detection with alert notifications and contextual information of real-life falls.

BACKGROUND: Falls are a leading cause of accidental deaths and injuries worldwide. The risk of falling is especially high for individuals suffering from balance impairments. Retrospective surveys and studies of simulated falling in lab conditions are frequently used and are informative, but prospective information about real-life falls remains sparse. Such data are essential to address fall risks and develop fall detection and alert systems. Here we present the results of a prospective study investigating a proof-of-concept, smartphone-based, online system for fall detection and notification. METHODS: The system uses the smartphone's accelerometer and gyroscope to monitor the participants' motion, and falls are detected using a regularized logistic regression. Data on falls and near-fall events (i.e., stumbles) is stored in a cloud server and fall-related variables are logged onto a web portal developed for data exploration, including the event time and weather, fall probability, and the faller's location and activity before the fall. RESULTS: In total, 23 individuals with an elevated risk of falling carried the phones for 2070 days in which the model classified 14,904,000 events. The system detected 27 of the 37 falls that occurred (sensitivity = 73.0 %) and resulted in one false alarm every 46 days (specificity > 99.9 %, precision = 37.5 %). 42.2 % of the events falsely classified as falls were validated as stumbles. CONCLUSIONS: The system's performance shows the potential of using smartphones for fall detection and notification in real-life. Apart from functioning as a practical fall monitoring instrument, this system may serve as a valuable research tool, enable future studies to scale their ability to capture fall-related data, and help researchers and clinicians to investigate real-falls.

Shoulder moments and angles during single and combined manual material handling tasks.

Shoulder musculoskeletal disorders due to manual material handling tasks are common workplace injuries. Here we investigated the difference in shoulder biomechanics (moments and angles) between a single task of removing a box from a shelf (or depositing a box on a shelf) and the equivalent part of a combined task that consisted of removing, carrying and depositing boxes; that is, a single removing [depositing] task was compared with the removing [depositing] part of a combined task. We found that the peak and cumulative shoulder moments were larger during the single-task paradigm than during the equivalent part of the combined task by 26.3 and 25.5%, respectively. The two paradigms also differed in terms of shoulder angles. It is likely that the main contributors to this overestimation were differences between the single and combined tasks in terms of the lever arm (i.e. horizontal distance), the shoulder angle, and the task duration. Practitioners' Summary: We investigated shoulder moments during single and combined manual material handling tasks. Shoulder moments were found to be smaller during combined tasks. Practitioners should consider that analysing combined tasks using estimations based on single tasks could result in an overestimation of 26.3 and 25.5% in peak and cumulative shoulder moments, respectively.Abbrevaitions: MSDs: musculoskeletal disorders; MMH: manual material handling; LMM: linear mixed model.

Inpatient stroke rehabilitation: prediction of clinical outcomes using a machine-learning approach.

BACKGROUND: In clinical practice, therapists often rely on clinical outcome measures to quantify a patient's impairment and function. Predicting a patient's discharge outcome using baseline clinical information may help clinicians design more targeted treatment strategies and better anticipate the patient's assistive needs and discharge care plan. The objective of this study was to develop predictive models for four standardized clinical outcome measures (Functional Independence Measure, Ten-Meter Walk Test, Six-Minute Walk Test, Berg Balance Scale) during inpatient rehabilitation. METHODS: Fifty stroke survivors admitted to a United States inpatient rehabilitation hospital participated in this study. Predictors chosen for the clinical discharge scores included demographics, stroke characteristics, and scores of clinical tests at admission. We used the Pearson product-moment and Spearman's rank correlation coefficients to calculate correlations among clinical outcome measures and predictors, a cross-validated Lasso regression to develop predictive equations for discharge scores of each clinical outcome measure, and a Random Forest based permutation analysis to compare the relative importance of the predictors. RESULTS: The predictive equations explained 70-77% of the variance in discharge scores and resulted in a normalized error of 13-15% for predicting the outcomes of new patients. The most important predictors were clinical test scores at admission. Additional variables that affected the discharge score of at least one clinical outcome were time from stroke onset to rehabilitation admission, age, sex, body mass index, race, and diagnosis of dysphasia or speech impairment. CONCLUSIONS: The models presented in this study could help clinicians and researchers to predict the discharge scores of clinical outcomes for individuals enrolled in an inpatient stroke rehabilitation program that adheres to U.S. Medicare standards.

Effect of using real motion versus predicted motion as input for digital human modeling of back and shoulder loads during manual material handling.

Digital human modeling (DHM) technology is considered the state of the art in designing and evaluating workstations. Previous studies examined the differences between DHM's posture and motion prediction relative to human experimental data. Yet, the effect the two different inputs on biomechanical loads was not assessed. Therefore, this study evaluates the differences in L4/L5 compression force and shoulder torques during a work process calculated using DHM with motion prediction (Jack by Siemens) and DHM with experimental data. The work process is a sequential removing, carrying, and depositing task performed by nine females and nine males and recorded using a motion capture system. The analysis shows that using experimental data results in larger back compression force during the removing task (average 15.4%), similar force during the depositing task (average 0.68%), and less force during the carrying task (19.875%). Using experimental data resulted in larger shoulder torque during all tasks (average 24.97%).

Validation of Amazon Halo Movement: a smartphone camera-based assessment of movement health.

Movement health is understanding our body's ability to perform movements during activities of daily living such as lifting, reaching, and bending. The benefits of improved movement health have long been recognized and are wide-ranging from improving athletic performance to helping ease of performing simple tasks, but only recently has this concept been put into practice by clinicians and quantitatively studied by researchers. With digital health and movement monitoring becoming more ubiquitous in society, smartphone applications represent a promising avenue for quantifying, monitoring, and improving the movement health of an individual. In this paper, we validate Halo Movement, a movement health assessment which utilizes the front-facing camera of a smartphone and applies computer vision and machine learning algorithms to quantify movement health and its sub-criteria of mobility, stability, and posture through a sequence of five exercises/activities. On a diverse cohort of 150 participants of various ages, body types, and ability levels, we find moderate to strong statistically significant correlations between the Halo Movement assessment overall score, metrics from sensor-based 3D motion capture, and scores from a sequence of 13 standardized functional movement tests. Further, the smartphone assessment is able to differentiate regular healthy individuals from professional movement athletes (e.g., dancers, cheerleaders) and from movement impaired participants, with higher resolution than that of existing functional movement screening tools and thus may be more appropriate than the existing tests for quantifying functional movement in able-bodied individuals. These results support using Halo Movement's overall score as a valid assessment of movement health.

Investigation of 3D dynamic and quasistatic models for spinal moments during combined manual material handling tasks.

Digital human modeling software uses biomechanical models to compute workers' risk of injury during industrial work processes. In many cases, the biomechanics are calculated using quasistatic models, which neglect the body's dynamics and therefore might be erroneous. This study investigated the differential effect of using a dynamic vs. a quasistatic model on spinal loading during combined manual material handling tasks that are prevalent in industry. An experiment was conducted involving nine male and nine female participants performing a total of 3402 cycles of a box-conveying task (removing, carrying and depositing) for different box masses and shelf heights. Using motion capture data, the peak and cumulative moments acting on the L5/S1 joint were calculated using 3D dynamic and quasistatic models. This revealed that neglecting the dynamic movements (i.e., using a quasistatic model) results in an on average underestimation of 19.7% in the peak spinal moment and 3.6% in the cumulative moment that in some cases exceeds the maximal limit for the compression forces acting on the lower back.

Workers' biomechanical loads and kinematics during multiple-task manual material handling.

This study investigated the biomechanical loads and kinematics of workers during multiple-task manual material handling (MMH) jobs, and developed prediction models for the moments acting on a worker's body and their peak joint angles. An experiment was conducted in which 20 subjects performed a total of 3780 repetitions of a box-conveying task. This task included continuous sequential removing, carrying and depositing of boxes weighing 2-12 kg. The subjects' motion was captured using motion-capture technology. The origin/destination height was the most influencing predictor of the spinal and shoulder moments and the peak trunk, shoulder and knee angles. The relationship between the origin/destination heights and the above parameters was nonlinear. The mass of the box, and the subject's height and mass, also influenced the spinal and shoulder moments. A tradeoff between the moments acting on the L5/S1 vertebrae and on the shoulder joint was found. Compared to the models developed in similar studies that focused on manual material handling (albeit under different conditions), the high-order prediction equation for peak spinal moment formulated in the present study was found to explain between 10% and 48% more variability in the moments. This suggests that using a high-order equation in future studies might improve the prediction.

Paramedic equipment bags: How their position during out-of-hospital cardiopulmonary resuscitation (CPR) affect paramedic ergonomics and performance.

This study investigates how the positions of paramedic equipment bags affect paramedic performance and biomechanical loads during out-of-hospital Cardiopulmonary Resuscitation (CPR). An experiment was conducted in which 12 paramedic teams (each including two paramedics) performed in-situ simulations of a cardiac-arrest scenario. CPR quality was evaluated using five standard resuscitation measures (i.e., pre- and post-shock pauses, and compression rate, depth and fraction). The spinal loads while lifting, pulling and pushing the equipment bags were assessed using digital human modeling software (Jack) and prediction equation from previous studies. The results highlight where paramedics are currently choosing to position their equipment. They also demonstrate that the positions of the equipment bags affect CPR quality as well as the paramedics' work efficiency, physiological effort and biomechanical loads. The spinal loads ranged from 1901 to 4030N; furthermore, every occasion on which an equipment bag was lifted resulted in spinal forces higher than 3400N, thus exceeding the maximum threshold stipulated by the National Institute for Occupational Safety and Health. 72% of paramedics' postures were categorized as high or very high risk for musculoskeletal disorders by the Rapid Entire Body Assessment. Guidelines related to bag positioning and equipment handling might improve CPR quality and patient outcomes, and reduce paramedics' risk of injury.

Differences in spinal moments, kinematics and pace during single-task and combined manual material handling jobs.

This study compared the spinal moments (i.e., peak and cumulative moments acting on the L5/S1 joint), kinematics (i.e., peak trunk and knee angles) and work pace of workers, when either removing a box from a shelf or depositing a box on a shelf, under two conditions: as a single task or as part of a combined task. An experiment was conducted, in which the subjects performed the tasks and were recorded using a motion capture system. An automated program was developed to process the motion capture data. The results showed that, when the removing and depositing tasks were performed as part of a combined task (rather than as single tasks), subjects experienced smaller peak and cumulative spinal moments and they performed the tasks faster. The results suggest that investigations into the separate tasks that comprise a combination have a limited ability to predict kinematics and kinetics during the combined job.

Factors determining workers' pace while conducting continuous sequential lifting, carrying, and lowering tasks.

To plan a new manual material handling work process, it is necessary to predict the times required to complete each task. Current time prediction models lack validity when the handled object's mass exceeds 2 kg. In this study, we investigated the effect of workplace design parameters on continuous sequential lifting, carrying, and lowering of boxes weighing from 2 kg to 14 kg. Both laboratory and field experiments were conducted. Results revealed that the box's weight and the lifting and lowering heights influenced the tasks' times. Further, the time to perform a task was influenced by the performance of other tasks in the same work process. New time prediction models were developed using the laboratory experiment data. Our models were found to be more accurate on average than the Maynard Operation Sequence Technique (MOST) and Methods Time Measurement (MTM-1) by 42% and 20%, respectively, for predicting the times of real workers at an actual workplace.