JMIR Publications

ABSTRACT

Background:

Impairments in upper limb (UL) function, often arising from conditions such as stroke and spinal cord injury, significantly contribute to long-term disability by limiting dexterity and diminishing overall quality of life. However, geographical, financial, and scheduling restrictions frequently make it difficult for post-stroke survivors to access traditional rehabilitation services, leaving many of them with unmet therapeutic needs.

Objective:

This study introduces a rehabilitation robotic system (RRS) integrated with a telerehabilitation platform (TRP) that enhances accessibility to UL neurorehabilitation by enabling health professionals to deliver personalized, adaptive UL neurorehabilitation remotely.

Methods:

The RRS leverages the desktop-mounted rehabilitation robot (DMRbotV3), developed in the BioRobotics Lab, to offer a comprehensive range of therapeutic modes including passive, active, and resistive exercises designed to support home-based rehabilitation. Additionally, integrating digital twin technology allows therapists to monitor real-time performance metrics (such as force exertion, joint angles, and torque values) via Azure IoT-based bidirectional communication, thus creating a connected ecosystem between patients, therapists, and rehabilitation devices. Six participants evaluated the TRP by engaging with the DMRbotV3 and HoloLens 2, providing detailed feedback, and completing a system usability assessment.

Results:

The HoloLens 2-based system was safe and user-friendly, with no adverse events reported, and all participants successfully completed the study. Usability assessments indicated high user satisfaction, with participants reporting enhanced engagement and the interactive nature of the mixed-reality (MR) integration. The robotic system demonstrated precise, smooth performance; velocity profiles across the joints ranged from –10 m/s to +10 m/s, ensuring controlled and adaptive movements. Participants actively engaged with the system, exerting forces ranging from 0 N to 25 N across multiple axes, thereby supporting motor recovery and strength development. Constructive participant feedback also highlighted opportunities to enhance system adaptability, personalize exercises further, and improve responsiveness to diverse patient needs.

Conclusions:

The integration of the RRS with TRP and MR technologies offers a promising approach to overcoming traditional barriers in UL neurorehabilitation. The system’s safety, strong usability, precise robotic performance, and patient engagement suggest that it can deliver effective, personalized remote therapy beyond traditional clinical settings.