An Agent-based Approach to Physical Rehabilitation of Patients affected by Neurological Diseases

Neurological disorders such as strokes, dementia, or ABI, among others, represent a major burden on European and worldwide healthcare systems, and can be understood as an unmet clinical need, recognised as a global challenge. This work discusses how agent technology can contribute to increase the autonomy of patients affected by neurological diseases who require physical rehabilitation. This approach can reduce face-to-face rehabilitation and the associated costs. We propose an architecture composed of software agents that play different roles to support the rehabilitation process from the point of view of patients and clinicians. This architecture has been used to deploy a multi-agent system for the physical rehabilitation of patients affected by a stroke. The involved rehabilitation exercises can be automatically assessed, classified and supervised. The conducted performance tests allow to conclude that the system can provide feedback to the patients in real-time.