Assistive system for people with Apraxia using a Markov decision process.

Area of Science:

• Assistive technology
• Neurorehabilitation
• Human-computer interaction

Background:

• Stroke survivors often experience Apraxia or Action Disorganization Syndrome (AADS), impacting their ability to perform activities of daily living (ADLs).
• Existing rehabilitation methods may not provide sufficient real-time support for relearning complex motor tasks.
• There is a need for innovative assistive systems to bridge the gap in functional recovery.

Purpose of the Study:

• To introduce CogWatch, an assistive system for stroke survivors with AADS.
• To detail the real-time planning approach using a Markov Decision Process (MDP).
• To evaluate the system's effectiveness in improving task performance through user simulation.

Main Methods:

• Development of the CogWatch assistive system.
• Implementation of a real-time planning module based on Markov Decision Process (MDP).
• Validation of the system's efficacy using user simulation studies.

Main Results:

• The user simulation demonstrated CogWatch's ability to improve task performance for individuals with AADS.
• The MDP-based planning approach showed potential for adaptive and effective assistance.
• Simulated results indicate a positive impact on relearning activities of daily living.

Conclusions:

• CogWatch presents a promising assistive technology for neurorehabilitation.
• The MDP approach offers a robust framework for real-time adaptive planning in assistive systems.
• Further research and real-world testing are needed to address remaining challenges and explore future enhancements.