Fall detection from a manual wheelchair: preliminary findings based on accelerometers using machine learning techniques.

Automated fall detection devices for individuals who use wheelchairs to minimize the consequences of falls are lacking. This study aimed to develop and train a fall detection algorithm to differentiate falls from wheelchair mobility activities using machine learning techniques. Thirty, healthy, ambulatory, young adults simulated falls from a wheelchair and performed other wheelchair-related mobility activities in a laboratory. Neural Network classifiers were used to train the algorithm developed based on data retrieved from accelerometers mounted at the participant's wrist, chest, and head. Results indicate excellent accuracy to differentiate between falls and wheelchair mobility activities. The sensors mounted at the wrist, chest, and head presented with an accuracy of 100%, 96.9%, and 94.8%, respectively, using data from 258 falls and 220 wheelchair mobility activities. This pilot study indicates that a fall detection algorithm developed in a laboratory setting based on fall accelerometer patterns can accurately differentiate wheelchair-related falls and wheelchair mobility activities. This algorithm should be integrated into a wrist-worn devices and tested among individuals who use a wheelchair in the community.

Acute physiological responses to steady-state arm cycling ergometry with and without blood flow restriction.

PURPOSE: To compare heart rate (HR), oxygen consumption (VO2), blood lactate (BL), and ratings of perceived exertion (RPE) during arm cycling with and without a blood flow restriction (BFR). METHODS: Twelve healthy males (age: 23.9 ± 3.75 years) completed four, randomized, 15-min arm cycling conditions: high-workload (HW: 60% maximal power output), low-workload (LW: 30% maximal power output), low-workload with BFR (LW-BFR), and BFR with no exercise (BFR-only). In the BFR conditions, cuff pressure to the proximal biceps brachii was set to 70% of occlusion pressure. HR, VO2, and RPE were recorded throughout the exercise, and BL was measured before, immediately after, and five minutes post-exercise. Within-subject repeated-measures ANOVA was used to evaluate condition-by-time interactions. RESULTS: HW elicited the greatest responses in HR (91% of peak; 163.3 ± 15.8 bpm), VO2 (71% of peak; 24.0 ± 3.7 ml kg-1 min-1), BL (7.7 ± 2.5 mmol L-1), and RPE (14 ± 1.7) and was significantly different from the other conditions (p < 0.01). The LW and LW-BFR conditions did not differ from each other in HR, VO2, BL, and RPE mean of conditions: ~ 68%, 41%, 3.5 ± 1.6 mmol L-1, 10.4 ± 1.6, respectively; p > 0.05). During the BFR-only condition, HR increased from baseline by ~ 15% (on average) (p < 0.01) without any changes in VO2, BL, and RPE (p > 0.05). CONCLUSIONS: HW arm cycling elicited the largest and most persistent physiological responses compared to LW arm cycling with and without a BFR. As such, practitioners who prescribe arm cycling for their clients should be advised to augment the demands of exercise via increases in exercise intensity (i.e., power output), rather than by adding BFR.

Relationship Between Lower Limb Function and Fall Prevalence in Ambulatory Adults With Spinal Cord Injury: A Systematic Review.

Background: Falls are common, detrimental events among ambulatory individuals with spinal cord injury (SCI). Following SCI, changes to lower limb function are probable and likely to impact an individual's fall risk, yet no comprehensive review has been completed on the topic. Objectives: This study systematically reviewed data on the relationship between lower limb function and fall prevalence in ambulatory individuals with SCI. Methods: A literature search was conducted in PubMed, Web of Science, Scopus, and CINAHL. Two independent reviewers screened abstracts/titles and then full articles. Study details, participants' characteristics, lower limb function assessed, and fall-related data were extracted from the studies. A qualitative analysis of the relationship between lower limb function and fall prevalence was performed. The risk of bias was evaluated using the Newcastle-Ottawa Quality Assessment Scale. Results: The search yielded 1553 articles. Eight prospective, two retrospective, and three cross-sectional studies met the eligibility criteria. These studies ranged from low to high risk of bias. Overall, the qualitative analysis provided little evidence to support the relationship between lower limb function recorded by clinical measures and fall prevalence. Conclusion: This review highlights the inconsistent relationship between lower limb function and falls prevalence in ambulatory adults with SCI. Greater uniformity in methodology and consistent categorization of fallers and nonfallers among researchers is necessary to move the field forward. Investigating additional factors such as behavior traits, assistive device use, and environmental risk factors may be appropriate in understanding fall prevalence in this population.

The Developments and Iterations of a Mobile Technology-Based Fall Risk Health Application.

Falls are a prevalent and serious health concern across clinical populations. A critical step in falls prevention is identifying modifiable risk factors, but due to time constraints and equipment costs, fall risk screening is rarely performed. Mobile technology offers an innovative approach to provide personalized fall risk screening for clinical populations. To inform future development, this manuscript discusses the development and testing of mobile health fall risk applications for three unique clinical populations [older adults, individuals with Multiple Sclerosis (MS), and wheeled-device users]. We focus on key lessons learned and future directions to improve the field of fall risk mHealth. During the development phase, we first identified fall risk factors specific to each population that are measurable with mobile technology. Second, we determined whether inertial measurement units within smartphones can measure postural control within the target population. Last, we developed the interface of each app with a user-centered design approach with usability testing through iterative semi-structured interviews. We then tested our apps in real-world settings. Our cumulative work demonstrates that mobile technology can be leveraged to provide personalized fall risk screening for different clinical populations. Fall risk apps should be designed and tailored for the targeted group to enhance usefulness and feasibility. In addition, fall risk factors measured with mobile technology should include those that are specific to the population, are measurable with mobile technology, and can accurately measure fall risk. Future work should improve fall risk algorithms and implement mobile technology into fall prevention programs.

Sensitivity of Apple Watch fall detection feature among wheelchair users.

A reliable fall detection device is crucial to minimize long-term consequences of falls among wheelchair users. This study examines the sensitivity of Apple Watch to detect intentional falls from a wheelchair. Twenty-five able bodied (age: 21.9 ± 2.5 years) participated in a protocol in which they intentionally fell out of a wheelchair in a laboratory setting. Each participant wore an Apple Watch Series 5 and performed 3 falls in the forward, right and left sideways, and backward directions onto a crash pad totaling 12 falls each. The Apple Watch was manually checked after each fall to determine if the device registered a fall. From 300 fall trials captured, the Apple Watch detected 14 falls showing a sensitivity of 4.7%, a false negative rate of 95.3%. Logistic regression showed that participant's height, impact force, lower limb functioning, and fall direction are parameters that may influence the ability of the Apple Watch to detect falls from a wheelchair. The Apple Watch fall detection feature presented with a very poor sensitivity to detect intentional falls from a wheelchair among able bodied young adults. Due to the high incidence and consequences of falls, a reliable fall detection device specific for wheelchair users is warranted.

The neural underpinnings of motor learning in people with neurodegenerative diseases: A scoping review.

Chronic progressive neurodegenerative diseases (NDD) cause mobility and cognitive impairments that disrupt quality of life. The learning of new motor skills, motor learning, is a critical component of rehabilitation efforts to counteract these chronic progressive impairments. In people with NDD, there are impairments in motor learning which appear to scale with the severity of impairment. Compensatory cortical activity plays a role in counteracting motor learning impairments in NDD. Yet, the functional and structural brain alterations associated with motor learning have not been synthesized in people with NDD. The purpose of this scoping review is to explore the neural alterations of motor learning in NDD. Thirty-five peer-reviewed original articles met the inclusion criteria. Participant demographics, motor learning results, and brain imaging results were extracted. Distinct motor learning associated compensatory processes were identified across NDD populations. Evidence from this review suggests the success of motor learning in NDD populations depends on the neural alterations and their interaction with motor learning networks, as well as the progression of disease.

Frequency and characteristics of falls in people living with and without multiple sclerosis during the COVID-19 pandemic: A cross-sectional online survey.

BACKGROUND: Public health responses to Coronavirus Disease 2019 (COVID-19) including lockdowns may negatively impact physical and mental functioning in clinical populations. People living with multiple sclerosis (MS) may be more susceptible to physical function deterioration while practicing social distancing. Recent reports have suggested that about 50% of people with MS (pwMS) decreased their leisure physical activity during COVID-19, and upwards of 30% reported decreased physical fitness levels. However, the impact of social distancing on adverse health-related outcomes such as falls has not received much scrutiny. Therefore, we explored the frequency and characteristics of falls experienced by people living with and without MS during the COVID-19 pandemic. METHODS: Two-hundred and thirty-nine individuals, including 106 pwMS (median age: 59 years) and 133 people living without MS (median age: 66 years) were recruited for this cross-sectional study. A snowball sampling strategy was used for online recruitment. Participants completed a customized falls questionnaire and the number of falls experienced (if any) during COVID-19 was recorded. Fall-related characteristics such as the timing, locations, activities undertaken before falling and consequences, as well as self-reported physical activity were also recorded. RESULTS: Overall, participants reported 232 falls (1.67 falls/person in pwMS and 0.41 falls/person in non-MS participants). People living with MS (pwMS) had a significantly higher frequency of falls (58.5% vs 21.8%; p< 0.001) and recurrent falls (45.3% vs 9.8%; p< 0.001) compared to non-MS participants. Additionally, pwMS reported a significantly higher proportion of in-home falls (83.9% vs 54.2%; p = 0.004), as well as a higher proportion of overall injuries (44.3% vs 12.5%, p< 0.001), fractures (5.7% vs 0.8%, p = 0.048), and healthcare utilization (9.4% vs 1.6%, p = 0.007) compared to non-MS participants. A similar proportion of pwMS (49.1%) and non-MS participants (52.2%) reported lower physical activity levels during COVID-19. CONCLUSION: This cross-sectional study revealed that pwMS remain at high risk of falls and fall-related outcomes during COVID-19. The high number of falls experienced by pwMS is of clinical concern considering the current strain on the healthcare system. Findings from this study highlight the importance of monitoring falls and the potential for telerehabilitation in persons with MS during COVID-19.

Feasibility and preliminary reliability and validity of remote sitting balance assessments among wheelchair users.

The aim of this study was to investigate the feasibility and preliminary validity and reliability of remote sitting balance assessment. Seven wheelchair users (mean age: 42.7 ± 19.74 years) participated in an in-person and remote sitting balance assessment. The assessments were compared to investigate the concurrent validity of the remote assessment. Reliability of remote assessment was evaluated using intraclass coefficient correlation (ICC) and the Bland-Altman. No significant differences were observed between the mean scores of in-person and remote administrations of the clinical tests (P's > 0.05). High to very high agreement was found between in-person and remote assessments (ICC = 0.88-0.982, P < 0.05). The agreement was confirmed by Bland-Altman graph analysis. Preliminary results indicate remote sitting balance assessment is feasible to perform, valid, and reliable.

The Validity, Reliability, and Sensitivity of a Smartphone-Based Seated Postural Control Assessment in Wheelchair Users: A Pilot Study.

Seated postural control is essential for wheelchair users to maintain proper position while performing activities of daily living. Clinical tests are commonly used to measure seated postural control, yet they are subjective and lack sensitivity. Lab-based measures are highly sensitive but are limited in scope and restricted to research settings. Establishing a valid, reliable, and accessible measurement tool of seated postural control is necessary for remote, objective assessments. Therefore, the purpose of this study was to examine the validity, reliability, and sensitivity of smartphone-based postural control assessments in wheelchair users. Eleven participants (age: 35.4 ± 17.9) completed two experimental visits 1-week apart consisting of three clinical tests: Trunk Control Test (TCT), Function in Sitting Test (FIST), and Tee-shirt Test, as well as, standardized instrumented balance tasks that manipulated vision (eyes open and closed), and trunk movement (functional reach and stability boundary). During these tasks, participants held a smartphone instrumented with a research-grade accelerometer to their chest. Maximum and root mean square (RMS) acceleration in the medial-lateral (ML) and anterior-posterior (AP) axes were derived. Participants were grouped into non-impaired and impaired postural groups based on FIST scores. Spearman rank-order correlations were conducted between the two devices' outcome measurements and between these measures and those of the clinical tests. Receiver operating characteristic (ROC) curves and the area under the curves (AUC) were determined to distinguish participants with and without impaired postural control. The reliability of outcome variables was assessed using inter-class correlations. Strong correlations between outputs derived from the smartphone and research-grade accelerometer were seen across balance tasks (ρ = -0.75-1.00; p ≤ 0.01). Numerous significant moderate correlations between clinical test outcomes and smartphone and research-grade RMS ML accelerometry were seen (ρ = -0.62 to 0.83 (p ≤ 0.044)]. On both devices, the AUC for ROC plots were significant for RMS ML sway during the eyes open task and functional stability boundary (p < 0.05). Reliability of smartphone accelerometry was comparable to the research-grade accelerometer and clinical tests. This pilot study illustrated that smartphone-based accelerometry may be able to provide a valid and reliable assessment of seated postural control and have the ability to distinguish between those with and without impaired postural control.

Next Steps in Wearable Technology and Community Ambulation in Multiple Sclerosis.

PURPOSE OF REVIEW: Walking impairments are highly prevalent in persons with multiple sclerosis (PwMS) and are associated with reduced quality of life. Walking is traditionally quantified with various measures, including patient self-reports, clinical rating scales, performance measures, and advanced lab-based movement analysis techniques. Yet, the majority of these measures do not fully characterize walking (i.e., gait quality) nor adequately reflect walking in the real world (i.e., community ambulation) and have limited timescale (only measure walking at a single point in time). We discuss the potential of wearable sensors to provide sensitive, objective, and easy-to-use assessment of community ambulation in PwMS. RECENT FINDINGS: Wearable technology has the ability to measure all aspects of gait in PwMS yet is under-studied in comparison with other populations (e.g., older adults). Within the studies focusing on PwMS, half that measure pace collected free-living data, while only one study explored gait variability in free-living conditions. No studies explore gait asymmetry or complexity in free-living conditions. Wearable technology has the ability to provide objective, comprehensive, and sensitive measures of gait in PwMS. Future research should investigate this technology's ability to accurately assess free-living measures of gait quality, specifically gait asymmetry and complexity.