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

Area of Science:

• Rehabilitation Engineering
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Motor dysfunction is a growing concern, particularly in aging populations.
• Assistive devices are crucial for improving the quality of life for individuals with motor impairments.
• Information fusion technology offers enhanced efficacy and specificity in rehabilitation treatments.

Purpose of the Study:

• To review recent advancements in assistive devices utilizing multimodal information fusion control.
• To introduce the application of multimodal information fusion control in rehabilitation equipment design.
• To provide a comprehensive analysis of current research for future development.

Main Methods:

• Review of literature on multimodal information fusion control in assistive devices.
• Analysis of data-level, feature-level, and decision-level fusion techniques.
• Summary and evaluation of commonly employed fusion algorithms.

Main Results:

• Multimodal information fusion enhances the performance and adaptability of assistive devices.
• Different fusion levels (data, feature, decision) offer distinct advantages and disadvantages.
• Various fusion algorithms are applicable, each with specific strengths.

Conclusions:

• Multimodal information fusion control is a promising approach for developing advanced assistive devices.
• Optimizing fusion strategies is key to maximizing the effectiveness of rehabilitation equipment.
• This review serves as a valuable reference for researchers and developers in the field.