Optimizing BCI Rehabilitation Protocols for Stroke: Exploring Task Design and Training Duration

TL;DR

The paper addresses the challenge of variable outcomes in stroke rehabilitation with brain-computer interfaces by testing an affected-hand movement versus rest MI paradigm, instead of the standard left-versus-right MI, across two EEG datasets. It employs CSP/FBCSP and EEGNet to classify MI and investigates the impact of shorter training sessions, finding that MI:Rest generally outperforms L:R and that smaller data subsets can yield higher accuracies, especially in stroke patients. The study demonstrates distinct ERD/ERS patterns across healthy and patient groups and suggests that task design and training duration critically influence BCI-based neurorehabilitation efficacy. The findings support adaptive, patient-tailored BCI protocols with shorter, multiple training sessions to enhance neuroplasticity and functional recovery after stroke.

Abstract

Stroke is a leading cause of long-term disability and the second most common cause of death worldwide. Although acute treatments have advanced, recovery remains challenging and limited. Brain-computer interfaces (BCIs) have emerged as a promising tool for post-stroke rehabilitation by promoting neuroplasticity. However, clinical outcomes remain variable, and optimal protocols have yet to be established. This study explores strategies to optimize BCI-based rehabilitation by comparing motor imagery of affected hand movement versus rest, instead of the conventional left-versus-right motor imagery. This alternative aims to simplify the task and address the weak contralateral activation commonly observed in stroke patients. Two datasets, one from healthy individuals and one from stroke patients, were used to evaluate the proposed approach. The results showed improved performance using both FBCSP and EEGNet. Additionally, we investigated the impact of session duration and found that shorter training sessions produced better BCI performance than longer sessions.

Figures (4)

• Figure 1: A schematic representation of the MI-based BCI pipeline.
• Figure 2: ERSP spectrum of central region in healthy subjects, LHP and RHP patients.
• Figure 3: Classification performance for healthy subjects, and patients with LHP and RHP, using three different approaches. L:R refers to the classification between left and right-hand motor imagery. MI:Rest is the classification between motor imagery (left MI for LHP, right MI for RHP and healthy subjects) and rest.
• Figure 4: Classification accuracy of the FBCSP algorithm using smaller subsets of the original datasets (three subsets for the BCI Competition IV dataset and two for the stroke dataset).