Single-Case Design for Network-based BCI Rehabilitation: Individual Learning Trajectories and Effects of Motor Imagery Timing in Stroke Patients
Abstract
Motor imagery (MI) brain-computer interfaces (BCIs) are increasingly investigated for promoting motor recovery after stroke. However, it remains unclear how BCI control improves through repeated training in stroke patients and whether the timing of MI decoding systematically influences this trajectory. To address this question, we implemented a single-case experimental design involving five chronic stroke patients enrolling in a longitudinal electroencephalography MI-BCI training involving their affected hand. Across 18 sessions conducted over six weeks, the onset of a fixed-duration MI decoding window was randomized to occur 1, 2, or 3~s after cue onset according to an alternating-treatments single-case design. Notably, the BCI combined spectral power and functional connectivity features to classify MI states. The MI classification timing had no significant effect on BCI accuracy, fatigue, or spasticity. BCI accuracy increased across training in four of five patients and was accompanied by patient-specific longitudinal changes in local spectral activity and functional connectivity. These convergent trajectories were consistent with acquisition of BCI control, but were not accompanied by clinically meaningful improvement in motor function. Varying the onset of the MI decoding window provided no systematic benefit for BCI control. Instead, differences in neurophysiological trajectories across patients indicate that MI decoding parameters need to be personalized and reassessed throughout the training. Overall, these findings inform more individualized and effective BCI settings for neuromotor stroke rehabilitation.
Related
- From Individual Trajectories to Population Densities: Weak-Form Inference of Heterogeneous Growth Laws
- Characterizing Individual and Population-Level Heterogeneity in Infant BMI Trajectories Using Flexible and Functional Modeling Approaches
- Digital twins forecast individual immune trajectories
- Longitudinal trajectories and persistence of frequent emergency department use: A two-year individual-level cohort study
- VARION: A Network Propagation Framework for Individual Patient Somatic Mutation Interpretation in Cancer Molecular Subtyping
BibTeX
@misc{bousfiha2026singlecase,
title = {{Single-Case Design for Network-based BCI Rehabilitation: Individual Learning Trajectories and Effects of Motor Imagery Timing in Stroke Patients}},
author = {Camile Bousfiha and Camille Gabot and Diana Baili and Erin Soller and Marion Couton and Grace Adwane and Michaël Obadia and Chiara Zavanone and Sophie Dupont and Eléonore Bayen and Laurent Bougrain and Paolo Bartolomeo and Fabrizio de Vico Fallani},
howpublished = {HAL},
year = {2026},
month = oct,
doi = {10.64898/2026.10.21.26366015},
url = {https://inria.hal.science/hal-05837840v1},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Single-Case Design for Network-based BCI Rehabilitation: Individual Learning Trajectories and Effects of Motor Imagery Timing in Stroke Patients
AU - Camile Bousfiha
AU - Camille Gabot
AU - Diana Baili
AU - Erin Soller
AU - Marion Couton
AU - Grace Adwane
AU - Michaël Obadia
AU - Chiara Zavanone
AU - Sophie Dupont
AU - Eléonore Bayen
AU - Laurent Bougrain
AU - Paolo Bartolomeo
AU - Fabrizio de Vico Fallani
PY - 2026
DA - 2026/10/21
PB - HAL
DO - 10.64898/2026.10.21.26366015
UR - https://inria.hal.science/hal-05837840v1
AB - Motor imagery (MI) brain-computer interfaces (BCIs) are increasingly investigated for promoting motor recovery after stroke. However, it remains unclear how BCI control improves through repeated training in stroke patients and whether the timing of MI decoding systematically influences this trajectory. To address this question, we implemented a single-case experimental design involving five chronic stroke patients enrolling in a longitudinal electroencephalography MI-BCI training involving their affected hand. Across 18 sessions conducted over six weeks, the onset of a fixed-duration MI decoding window was randomized to occur 1, 2, or 3~s after cue onset according to an alternating-treatments single-case design. Notably, the BCI combined spectral power and functional connectivity features to classify MI states. The MI classification timing had no significant effect on BCI accuracy, fatigue, or spasticity. BCI accuracy increased across training in four of five patients and was accompanied by patient-specific longitudinal changes in local spectral activity and functional connectivity. These convergent trajectories were consistent with acquisition of BCI control, but were not accompanied by clinically meaningful improvement in motor function. Varying the onset of the MI decoding window provided no systematic benefit for BCI control. Instead, differences in neurophysiological trajectories across patients indicate that MI decoding parameters need to be personalized and reassessed throughout the training. Overall, these findings inform more individualized and effective BCI settings for neuromotor stroke rehabilitation.
ER -