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

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Artificial Intelligence in Healthcare

Background:

• Aging often leads to decreased muscle strength and endurance in the elderly, impacting daily activities and independence.
• Limited mobility and strength hinder the ability to perform routine tasks, affecting overall quality of life.

Purpose of the Study:

• To evaluate the effectiveness of a developed upper limb exoskeleton for weight lifting in elderly individuals.
• To predict the suitability of the exoskeleton using ergonomic analysis and a weighted K-Nearest Neighbors (KNN) machine learning algorithm.

Main Methods:

• Experimental measurements of Maximum Voluntary Isometric Contraction (MVIC) and Mean Power Frequency (MPF) were taken before and after device use.
• Ergonomic analysis and a weighted KNN algorithm were employed to assess muscle strength and predict device effectiveness.
• Elderly subjects performed weight-lifting tasks with and without the exoskeleton to measure muscle response.

Main Results:

• Exoskeleton use reduced %MVIC values significantly during 5kg and 15kg weight lifting.
• Muscle fatigue in Biceps Brachii and flexor carpi radialis increased without the exoskeleton but decreased with its use.
• %MVIC values ranged from 2-6% (no load), 25-40% (5kg), and 30-71% (15kg) with the device.

Conclusions:

• The developed upper limb exoskeleton effectively reduces muscle fatigue and compensates for strength loss in the elderly during weight lifting.
• A weighted KNN algorithm can predict exoskeleton suitability based on Body Mass Index and muscle fatigue levels.
• Findings support the development of user-friendly assistive devices, highlighting the role of ergonomics and AI in enhancing rehabilitation technology.