Machine Learning- and Deep Learning-Based Myoelectric Control System for Upper Limb Rehabilitation Utilizing EEG and EMG Signals: A Systematic Review

Bioengineering, Год журнала: 2025, Номер 12(2), С. 144 - 144

Опубликована: Фев. 3, 2025

Upper limb disabilities, often caused by conditions such as stroke or neurological disorders, severely limit an individual’s ability to perform essential daily tasks, leading a significant reduction in quality of life. The development effective rehabilitation technologies is crucial restoring motor function and improving patient outcomes. This systematic review examines the application machine learning deep techniques myoelectric-controlled systems for upper rehabilitation, focusing on use electroencephalography electromyography signals. By integrating non-invasive signal acquisition methods with advanced computational models, highlights how these can enhance accuracy efficiency devices. A comprehensive search literature published between January 2015 July 2024 led selection fourteen studies that met inclusion criteria. These showcase various approaches decoding intentions controlling assistive devices, models Long Short-Term Memory Networks, Support Vector Machines, Convolutional Neural Networks showing notable improvements control precision. However, challenges remain terms model robustness, complexity, real-time applicability. aims provide researchers deeper understanding current advancements this field, guiding future research efforts overcome barriers facilitate transition from experimental settings practical, real-world applications.

Язык: Английский

---

Wearable device for personalized EMG feedback-based treatments

Results in Engineering, Год журнала: 2024, Номер 23, С. 102472 - 102472

Опубликована: Июнь 25, 2024

Curative effects of electromyography (EMG) feedback in treatment various conditions and/or recovery after injuries have been earlier reported. However, wider application on EMG is somehow limited due to the overall price such systems and availability outside specialized centers. Development a personalized device for would be great importance home stroke or injuries, achieving better success fitness improving biofeedback-based treatments as urinary incontinence. Despite extensive research signal collection, there lack focus in-situ analysis that considers intensity duration muscle activities. This gap presents motivation our research. In this paper, we present methodology realization wearable, rechargeable battery-powered, small-sized (90 mm × 60 mm) electronic recording two channels (12-bits resolution, sampling frequency up 1.6 kHz) with Bluetooth Low Energy connectivity smartphone. An average current consumption 20.5 mA was experimentally determined, suggesting multiday continuous functionality possible. Advancing state art, propose cross-correlation-based algorithm dynamical computing evaluation activation levels. can determine if follows predefined profile contractions/relaxations (as needed treatment) indicate muscles specific exercise were not engaged proper time intensity. The performed simulation showed proposed approach exhibited shorter processing compared Morlet Wavelet Transform Dynamic Time Warping. Finally, experimental work five human volunteers demonstrated reliability acquisition processing. Therefore, main contribution cost-effective, small-sized, customizable system an efficient collection

Язык: Английский

---

A survey on hand gesture recognition based on surface electromyography: Fundamentals, methods, applications, challenges and future trends

Applied Soft Computing, Год журнала: 2024, Номер 166, С. 112235 - 112235

Опубликована: Сен. 11, 2024

Язык: Английский

---

Automated Grasp Recognition using sEMG: Recent Advances, Challenges and Future Developments

IEEE Transactions on Instrumentation and Measurement, Год журнала: 2024, Номер 74, С. 1 - 17

Опубликована: Ноя. 13, 2024

Язык: Английский

---