Keywords
Augmented Reality, Multiple Sclerosis

Technologies
Unity 6, Meta Quest 3, Shimmer 3 GSR+, C#, Python

People with Multiple Sclerosis (MS) often experience combined motor, cognitive, and fatigue-related impairments that limit engagement with conventional rehabilitation. Dual-task training, practising motor and cognitive activities simultaneously, is a promising approach, but existing tools rarely adapt to how the user is feeling in real time.

This project develops a augmented-reality rehabilitation prototype using the Meta Quest 3 headset. Users complete kitchen-based tasks such as sorting ingredients, stirring pots, and answering verbal prompts, while the system monitors heart rate and skin conductance through a wearable Shimmer 3 GSR+ sensor. A machine-learning model estimates user stress from these signals and automatically adjusts task difficulty within a closed feedback loop.

The prototype supports three training modes across three difficulty levels, gamification elements, and an in-headset progress dashboard. Feasibility testing with healthy participants evaluates system stability, responsiveness, and usability to determine whether real-time physiological adaptation in mixed-reality rehabilitation is practically viable.