Transductive Learning Models for Accurate Ambulatory Gait Analysis in Elderly Residents of Assisted Living Facilities

Fall Risk Prediction Using Instrumented Footwear in Institutionalized Older Adults

This study presents a novel framework that utilizes instrumented footwear to predict fall risk in institutionalized older adults by leveraging stride-to-stride gait data. The older adults are categorized into fallers and non-fallers using three distinct criteria: retrospective fall history, prospective fall occurrence, and a combination of both retrospective and prospective data. Three types of data collected from N=95 institutionalized older adults are analyzed: traditional timed mobility tests, gait data collected from a validated electronic walkway, and gait data collected with instrumented footwear developed by our team. The importance of each type of data is assessed using a brute-force search method, through which the optimal features are selected. AdaBoost algorithms are then utilized to develop predictive models based on the selected features. The models are evaluated using leave-one-out cross-validation and 10-fold cross-validation. The results show that models using gait data from the instrumented footwear outperformed those based on traditional tests and walkway data, with area under the receiver operating characteristic curve (AUC) values for predicting prospective falls being 0.47, 0.66, and 0.80, respectively. The sensitivity of the models increases when they are trained using both past and future falls data, rather than relying solely on past or future falls data. This study demonstrates the potential of instrumented footwear for fall risk assessment in elderly individuals. The findings provide valuable insights for fall prevention and care, highlighting the superior predictive capabilities of the developed system compared to traditional methods.

Accurate COP Trajectory Estimation in Healthy and Pathological Gait Using Multimodal Instrumented Insoles and Deep Learning Models

Measuring center-of-pressure (COP) trajectories in out-of-the-lab environments may provide valuable information about changes in gait and balance function related to natural disease progression or treatment in neurological disorders. Traditional equipment to acquire COP trajectories includes stationary force plates, instrumented treadmills, electronic walkways, and insoles featuring high-density force sensing arrays, all of which are expensive and not widely accessible. This study introduces novel deep recurrent neural networks that can accurately estimate dynamic COP trajectories by fusing data from affordable and heterogeneous insole-embedded sensors (namely, an eight-cell array of force sensitive resistors (FSRs) and an inertial measurement unit (IMU)). The method was validated against gold-standard equipment during out-of-the-lab ambulatory tasks that simulated real-world walking. Root-mean-square errors (RMSE) in the mediolateral (ML) and anteroposterior (AP) directions obtained from healthy individuals (ML: 0.51 cm, AP: 1.44 cm) and individuals with neuromuscular conditions (ML: 0.59 cm, AP: 1.53 cm) indicated technical validity. In individuals with neuromuscular conditions, COP-derived metrics showed significant correlations with validated clinical measures of ambulatory function and lower-extremity muscle strength, providing proof-of-concept evidence of the convergent validity of the proposed method for clinical applications.

SR-POSE: A Novel Non-Contact Real-Time Rehabilitation Evaluation Method Using Lightweight Technology

Rehabilitation movement assessment often requires patients to wear expensive and inconvenient sensors or optical markers. To address this issue, we propose a non-contact and real-time approach using a lightweight pose detection algorithm-Sports Rehabilitation-Pose (SR-Pose), and a depth camera for accurate assessment of rehabilitation movement. Our approach utilizes an E-Shufflenet network to extract underlying features of the target, a RLE-Decoder module to directly regress the coordinate values of 16 key points, and a Weight Fusion Unit (WFU) module to output optimal human posture detection results. By combining the detected human pose information with depth information, we accurately calculate the angle between each joint in three-dimensional space. Furthermore, we apply the DTW algorithm to solve the distance measurement and matching problem of video sequences with different lengths in rehabilitation evaluation tasks. Experimental results show that our method can detect human joint nodes with an average detection speed of 14.32ms and an average detection accuracy for pose of 91.2%, demonstrating its computational efficiency and effectiveness for practical application. Our proposed approach provides a low-cost and user-friendly alternative to traditional sensor-based methods, making it a promising solution for rehabilitation movement assessment.

Instrumented insoles for assessment of gait in patients with vestibular schwannoma

Background

Imbalance and gait disturbances are common in patients with vestibular schwannoma (VS) and can result in significant morbidity. Current methods for quantitative gait analysis are cumbersome and difficult to implement. Here, we use custom-engineered instrumented insoles to evaluate the gait of patients diagnosed with VS.

Methods

Twenty patients with VS were recruited from otology, neurosurgery, and radiation oncology clinics at a tertiary referral center. Functional gait assessment (FGA), 2-minute walk test (2MWT), and uneven surface walk test (USWT) were performed. Custom-engineered instrumented insoles, equipped with an 8-cell force sensitive resistor (FSR) and a 9-degree-of-freedom inertial measurement unit (IMU), were used to collect stride-by-stride spatiotemporal gait parameters, from which mean values and coefficients of variation (CV) were determined for each patient.

Results

FGA scores were significantly correlated with gait metrics obtained from the 2MWT and USWT, including stride length, stride velocity, normalized stride length, normalized stride velocity, stride length CV, and stride velocity CV. Tumor diameter was negatively associated with stride time and swing time on the 2MWT; no such association existed between tumor diameter and FGA or DHI.

Conclusions

Instrumented insoles may unveil associations between VS tumor size and gait dysfunction that cannot be captured by standardized clinical assessments and self-reported questionnaires.

Gait monitoring for older adults during guided walking: An integrated assistive robot and wearable sensor approach

An active lifestyle can mitigate physical decline and cognitive impairment in older adults. Regular walking exercises for older individuals result in enhanced balance and reduced risk of falling. In this article, we present a study on gait monitoring for older adults during walking using an integrated system encompassing an assistive robot and wearable sensors. The system fuses data from the robot onboard Red Green Blue plus Depth (RGB-D) sensor with inertial and pressure sensors embedded in shoe insoles, and estimates spatiotemporal gait parameters and dynamic margin of stability in real-time. Data collected with 24 participants at a community center reveal associations between gait parameters, physical performance (evaluated with the Short Physical Performance Battery), and cognitive ability (measured with the Montreal Cognitive Assessment). The results validate the feasibility of using such a portable system in out-of-the-lab conditions and will be helpful for designing future technology-enhanced exercise interventions to improve balance, mobility, and strength and potentially reduce falls in older adults.

Gait parameters when walking with or without rollator on different surface characteristics: a pilot study among healthy individuals

Objectives

Gait parameters can measure risks of falling and mortality and identify early stages of frailty. The use of walking aid changes gait parameters. The aim of this study was to describe differences in gait parameters among healthy adults when walking on different surfaces and under different conditions, with and without a rollator.

Results

Ten healthy participants walked first without and then with a rollator upslope, downslope and on flat surface, on bitumen and gravel respectively. Step length, walking speed and sideway deviation was measured using an inertial measurement unit. Walking up a slope using a rollator generated the longest step length and walking down a slope using a rollator the shortest. Fastest walking speed was used when walking up a slope with rollator and slowest when walking down a slope with rollator. Sideway deviation was highest when walking down a slope and lowest when walking on gravel, both without rollator. Highest walk ratio was found when walk up a slope without rollator and lowest when walking down a slope with rollator. Data from this study provides valuable knowledge regarding gait parameters among healthy individuals, useful for future clinical research relevant for rehabilitation and public health.

Non-age-related gait kinematics and kinetics in the elderly

BACKGROUND

The change of gait kinematics and kinetics along aging were reported to indicate age-related gait patterns. However, few studies focus on non-age-related gait analysis. This study aims to explore the non-age-related gait kinematics and kinetics by comparing gait analysis outcomes among the healthy elderly and young subjects.

METHODS

Gait analysis at self-paced was conducted on 12 healthy young subjects and 8 healthy elderly subjects. Kinematic and kinetic features of ankle, knee and hip joints were analyzed and compared in two groups. The degree of variation between the young and elderly in each kinematic or kinetic feature was calculated from pattern distance and percentage of significant difference. The k-means clustering and Elbow Method were applied to select and validate non-age-related features. The average waveforms with standard deviation were plotted for the comparison of the results.

RESULTS

A total of five kinematic and five kinetic features were analyzed on ankle, knee and hip joints in healthy young and elderly groups. The degrees of variation in ankle moment, knee angle, hip flexion angle, and hip adduction moment were 0.1074, 0.1593, 0.1407, and 0.1593, respectively. The turning point was where the k value equals two. The clustering centers were 0.1417 and 0.3691, and the two critical values closest to the cutoff were 0.1593 and 0.3037. The average waveforms of the kinematic or kinetic features mentioned above were highly overlapped with a minor standard deviation between the healthy young and elderly but showed larger variations between the healthy and abnormal.

CONCLUSIONS

The cluster with a minor degree of variation in kinematic and kinetic features between the young and elderly were identified as non-age-related, including ankle moment, knee angle, hip flexion angle, and hip adduction moment. Non-age-related gait kinematics and kinetics are essential indicators for gait with normal function, which is essential in the evaluation of mobility and functional ability of the elderly, and data fusion of the assistant device.