Machine Learning for Classification of Physical Therapy Exercises from Inertial and Magnetic Sensor Data

Authors

  • Pasupunooti Anusha, Dubba Maniteja, Komara Namith Varun, Thakur Rohan Singh, Devoju Shiva Nagappa Chary Author

DOI:

https://doi.org/10.48047/7d2n8z19

Keywords:

Keywords: Patient rehabilitation, Physical therapy, Machine Learning, Support vector classifier, AdaBoost classifier, Inertial and magnetic sensor data

Abstract

Physical therapy exercises play a crucial role in patient rehabilitation, yet accurate classification remains a challenge. Studies show that sensor-based recognition of human activities can achieve up to 90% accuracy using machine learning, while traditional manual assessments are prone to inconsistencies, resulting in a 15–20% variation in evaluation outcomes. Currently, physical therapists rely heavily on subjective observations, which are time-consuming and susceptible to human error, limiting the precision of patient progress tracking. To address these challenges, we propose a machine learning based approach f

Downloads

Download data is not yet available.

Downloads

Published

09.04.2025

How to Cite

Machine Learning for Classification of Physical Therapy Exercises from Inertial and Magnetic Sensor Data . (2025). International Journal of Information and Electronics Engineering, 15(4), 74-86. https://doi.org/10.48047/7d2n8z19