Smart Sensors and Deep Learning for Recognizing Rehabilitation Exercises

Parumanchala Bhaskar, Siddi Anitha, Siddannagari Susmitha, Katthigalla Sushma Chandrika, Netla Harshita, Thapeta Anuradha

2025

Abstract

Rehabilitation can be quite a sensitive subject, needing to monitor it closely, especially for those recovering from drug addiction and individuals who are undergoing physical therapy. The importance of rehabilitation in basic terms is that structured rehabilitation exercises are the most essential and play a significant role in physical and mental wellbeing by recovering strength, mobility and stability. In this paper, we propose a method to identify and assess rehabilitation exercises based on the usage of smart sensors and Graph Neural Network (GNN). They capture both spatial and temporal relationship in the movement data, thus improving the accuracy of classifying exercises. Step Two: Utilizing wearable smart sensors to collect information about the patient's movements and physiological parameters, and a GNN model to analyse the raw data and provide feedback for teaching patients about their health status. This leads to high-confidence tracking, few errors, no need for manual marking, and also allows for remote patient monitoring, which improves the overall efficiency of rehabilitation. This not only stays relevant to drug-addicted individuals in control of their lives, but it also keeps them engaged and on track to do this effectively with excellent rehabilitation programs during their recoveries and patients with any kind of therapy-requiring care.

Paper Citation

in Harvard Style

Bhaskar P., Anitha S., Susmitha S., Chandrika K., Harshita N. and Anuradha T. (2025). Smart Sensors and Deep Learning for Recognizing Rehabilitation Exercises. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 111-115. DOI: 10.5220/0013878300004919

in Bibtex Style

@conference{icrdicct`2525,
author={Parumanchala Bhaskar and Siddi Anitha and Siddannagari Susmitha and Katthigalla Chandrika and Netla Harshita and Thapeta Anuradha},
title={Smart Sensors and Deep Learning for Recognizing Rehabilitation Exercises},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={111-115},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013878300004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Smart Sensors and Deep Learning for Recognizing Rehabilitation Exercises
SN - 978-989-758-777-1
AU - Bhaskar P.
AU - Anitha S.
AU - Susmitha S.
AU - Chandrika K.
AU - Harshita N.
AU - Anuradha T.
PY - 2025
SP - 111
EP - 115
DO - 10.5220/0013878300004919
PB - SciTePress