TY - GEN
T1 - Transforming Upper Limb Rehabilitation
T2 - 2nd IEEE International Conference on Computing, Semiconductor, Mechatronics, Intelligent Systems and Communications, COSMIC 2025
AU - Kini, Vasantha T.
AU - Basu, Haimanti
AU - Acharya, Aneesha K.
AU - Meenatchisundaram, S.
N1 - Publisher Copyright:
©2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Over the past decade, Python-based rehabilitation tools have dramatically transformed access to interactive, low-cost therapy by combining powerful technologies such as computer vision, sensor fusion, and intuitive user interfaces. This paper presents the development and evaluation of a novel, holistic real-time hand orientation feedback system which can be used in upper limb rehabilitation. Our implementation uses MediaPipe Holistic, OpenCV, and NumPy to estimate elbow, wrist, index finger, and neck angles from synchronized video feeds, comparing live movements to reference motions. Corrective prompts are generated immediately, enabling responsive guidance and enhanced patient autonomy. Preliminary results show smooth, sinusoidal angle tracking, reliable phase segmentation, and accurate real-time feedback, suggesting high potential for deployment in remote and home-based therapy contexts. Future enhancements include adaptive AI calibration, engaging gamification strategies, and multimodal sensor integration to personalize and optimize rehabilitation further.
AB - Over the past decade, Python-based rehabilitation tools have dramatically transformed access to interactive, low-cost therapy by combining powerful technologies such as computer vision, sensor fusion, and intuitive user interfaces. This paper presents the development and evaluation of a novel, holistic real-time hand orientation feedback system which can be used in upper limb rehabilitation. Our implementation uses MediaPipe Holistic, OpenCV, and NumPy to estimate elbow, wrist, index finger, and neck angles from synchronized video feeds, comparing live movements to reference motions. Corrective prompts are generated immediately, enabling responsive guidance and enhanced patient autonomy. Preliminary results show smooth, sinusoidal angle tracking, reliable phase segmentation, and accurate real-time feedback, suggesting high potential for deployment in remote and home-based therapy contexts. Future enhancements include adaptive AI calibration, engaging gamification strategies, and multimodal sensor integration to personalize and optimize rehabilitation further.
UR - https://www.scopus.com/pages/publications/105034707282
UR - https://www.scopus.com/pages/publications/105034707282#tab=citedBy
U2 - 10.1109/COSMIC67569.2025.11380887
DO - 10.1109/COSMIC67569.2025.11380887
M3 - Conference contribution
AN - SCOPUS:105034707282
T3 - COSMIC 2025 - 2nd IEEE International Conference on Computing, Semiconductor, Mechatronics, Intelligent Systems and Communications, Conference Proceedings
SP - 147
EP - 153
BT - COSMIC 2025 - 2nd IEEE International Conference on Computing, Semiconductor, Mechatronics, Intelligent Systems and Communications, Conference Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 21 November 2025 through 22 November 2025
ER -