Frontiers in bioengineering and biotechnology · 2025
Development of an upper limb muscle strength rehabilitation assessment system using particle swarm optimisation.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Zhou Chuangan, Wang Siqi, Wu Meiyi, Lai Wei, Yao Junyu, Gou Xingyue, Ye Hui, Yi Jun, Cao Dong
Аннотация
PURPOSE: This study develops a particle swarm optimization (PSO)-based assessment system for evaluating upper extremity and shoulder joint muscle strength with potential application to stroke rehabilitation. This study validates the system on healthy adult volunteers using surface electromyography and joint motion data.
METHODS: The system comprises a multimodal data acquisition module and a computational analysis pipeline. sEMG signals were collected non-invasively from the anterior, medial, and posterior deltoid muscles using bipolar electrode arrays. These signals are subjected to noise reduction and feature extraction. Simultaneously, triaxial kinematic data of the glenohumeral joint were obtained via an MPU6050 inertial measurement unit, processed through quaternion-based orientation estimation. Machine learning models, including Backpropagation Neural Network (BPNN), Support Vector Machines (SVM), and particle swarm optimization algorithms (PSO-BPNN, PSO-SVR), were applied for regression analysis. Model performance was evaluated using R-squared (R 2), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Bias Error (MBE).
RESULTS: The system successfully collected electromyographic and kinematic data. PSO-SVR achieved the best predictive performance (R 2 = 0.8600, RMSE = 0.3122, MAE = 0.2453, MBE = 0.0293), outperforming SVR, PSO-BPNN, and BPNN.
CONCLUSION: The PSO-SVR model demonstrated the highest accuracy, which can better facilitate therapists in conducting muscle strength rehabilitation assessments.
SIGNIFICANCE: This system enhances quantitative assessment of muscle strength in stroke patients, providing a reliable tool for rehabilitation monitoring and personalized therapy adjustments.
METHODS: The system comprises a multimodal data acquisition module and a computational analysis pipeline. sEMG signals were collected non-invasively from the anterior, medial, and posterior deltoid muscles using bipolar electrode arrays. These signals are subjected to noise reduction and feature extraction. Simultaneously, triaxial kinematic data of the glenohumeral joint were obtained via an MPU6050 inertial measurement unit, processed through quaternion-based orientation estimation. Machine learning models, including Backpropagation Neural Network (BPNN), Support Vector Machines (SVM), and particle swarm optimization algorithms (PSO-BPNN, PSO-SVR), were applied for regression analysis. Model performance was evaluated using R-squared (R 2), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Bias Error (MBE).
RESULTS: The system successfully collected electromyographic and kinematic data. PSO-SVR achieved the best predictive performance (R 2 = 0.8600, RMSE = 0.3122, MAE = 0.2453, MBE = 0.0293), outperforming SVR, PSO-BPNN, and BPNN.
CONCLUSION: The PSO-SVR model demonstrated the highest accuracy, which can better facilitate therapists in conducting muscle strength rehabilitation assessments.
SIGNIFICANCE: This system enhances quantitative assessment of muscle strength in stroke patients, providing a reliable tool for rehabilitation monitoring and personalized therapy adjustments.