Improving Stroke Rehabilitation: A Fusion of Personalized 3D User Interface Guidance and Brain-Computer Interface

Stroke or physical injuries resulting from accidents are leading causes of motor impairment. To restore motor functionality, active patient participation and involvement of neural processes are key factors. Brain-computer interface (BCI) rehabilitation systems pair the imagination of the movements with visual feedback in real-time to encourage and assess the patient’s willingness. Visual feedback through a virtual avatar can encourage the patient to correctly imagine the movements of the upper/lower extremities. The simultaneous activation of sensory and motor functions is a key factor for brain plasticity, which enables effective recovery from neurological impairments. Virtual reality (VR) can enhance brain plasticity and therapeutic outcomes by providing feedback through the visualization of motor imagery (MI) on a screen. In a 3D environment, subjects perceive an avatar’s hands and feet as their own, creating a sense of embodiment. The avatar executes the movements corresponding to the subject’s MI, which the system detects. This project aims to develop an innovative user interface (UI) for a 3D rehabilitation environment, named BCI-3D ReHaB, that focuses on movement visualization and personalization for individuals undergoing rehabilitation. The UI will integrate features that can be modified by the users to enhance their rehabilitation outcomes, such as gender, environment, and spatial location. This will allow the users to adjust the interface to their preferences and goals, enhancing their involvement and so the effectiveness of the prescribed therapy. The motivation for this system came from a commercial device (g.tec recoveriX )'s software solution for the same problem. However, this software was restricted to be used only with g.tec recoveriX with a limited set of configurations. Therefore, similar but more flexible software is required to enable any BCI device to benefit from the system that will be developed. The design’s objective is to enhance the rehabilitation process’s quality and efficiency by enabling users to interact with a 3D environment, observing themselves perform specific movements. Rehabilitation sessions are customizable, and real-time visual feedback will stimulate correct movement imagination and user adherence. The designed UI aims to facilitate user manipulation, allowing end-users to select the appropriate environment and specify the body parts requiring training. After fulfilling the UI specifications, the signal collection capability will be augmented with marker locations and recordings to monitor when users initiate movement. This enhancement is crucial for accurately capturing the timing and nature of physical activities. The final step of the project involves developing a communication module between the UI and the Electroencephalogram (EEG) acquisition system. This module aims to ensure the recording and alignment of the EEG stream with markers indicating the initiation of movement. Initial validation is performed through simulation with a test group, and further tests are planned to focus on movement imagination rather than execution. Indeed, the stored data from these sessions could be post-processed to train a classifier for MI distinction, enabling the system to differentiate between intentions to move specific body parts. This analysis could lead to a future applications involving active feedback, where an avatar moves upon correctly recognizing movement intention, enhancing the interactivity and effectiveness of rehabilitation sessions.