Multimodal Information Fusion for Control of Rehabilitation Robots in Motor Dysfunction: A Review

Abstract

This paper reviews recent advances in assistive devices based on multimodal information fusion control, designed for individuals with motor dysfunction. The prevalence of motor dysfunction is increasingly concerning amidst global population aging. Information fusion technology, widely adopted in rehabilitation, enhances the efficacy and specificity of rehabilitation treatments. This paper introduces the concept of multimodal information fusion control into rehabilitation equipment design. It highlights the advantages and disadvantages of data-level, feature-level, and decision-level fusion, along with commonly employed fusion algorithms. By summarizing and analyzing the current state of research, this paper aims to provide a valuable reference for the further development and optimization of assistive devices for motor dysfunction.