Predicting the performance of assistive device for elderly people using weighted KNN machine learning algorithm

Abstract

Background

Elderly people as age increases often struggle with weight lifting in their daily lives due to decreased muscle strength and endurance. This limits their ability to perform routine tasks, which affects their independence and quality of life.

Objective

The aim of this study is to evaluate and predict the effectiveness of the developed upper limb Exo-skeleton for weight lifting, using ergonomic analysis and a weighted K-Nearest Neighbors (KNN) machine learning algorithm.

Methods

Experiments were conducted to measure Maximum Voluntary Isometric Contraction (MVIC) and Mean Power Frequency (MPF) values to assess muscle strength before and after wearing the device on elderly subjects.

Results

The results of the %MVIC value of muscles when lifting no load after wearing the assistive device lies between 2% to 6%, whereas while adding 5 kg load on hand, MVIC lies between 25% to 40%, while adding 15 kg load, the MVIC value is slightly increased to 30% to 71%. The results indicated that the muscle fatigue in the Biceps Brachii (BB) and flexor carpi radialis (FCR) are increased during weight lifting without the Exo-skeleton, whereas the usage of the device significantly reduces the muscle fatigue.

Conclusion

The results demonstrated that the exoskeleton significantly reduces MVIC range when lifting 5 kg and 15 kg weight, indicating decreased muscle fatigue in the biceps and radialis muscles when using the Exo-skeleton. The weighted K nearest neighboring algorithm which predicts the new nerve disordered elderly subject, whether the assistive device is suitable or not based on his Body Mass Index (BMI) and muscle fatigueless. The results suggested that the proposed upper limb assistive device compensates for muscular strength during weight lifting, potentially guiding the development of user-friendly assistive devices for the elderly. The study highlights the significance of ergonomic studies and AI algorithms in enhancing upper limb assistive device design and functionality.