# Neuromodulation and rehabilitation of post-stroke cognitive impairment: challenges and prospects

## TL;DR

This review explores how neuromodulation techniques can help improve cognitive recovery after stroke and highlights future directions for more effective treatments.

## Contribution

The paper proposes a future direction for brain-machine interaction neuromodulation using artificial intelligence for closed-loop cognitive recovery.

## Key findings

- Neuromodulation techniques like TMS and TES show promise in cognitive recovery but lack large-scale validation.
- Photobiomodulation and transcranial ultrasound stimulation face technical and evidence-based challenges.
- AI-driven closed-loop neuromodulation is suggested as a future approach for enhanced cognitive rehabilitation.

## Abstract

It is essential to recognize the significant daily impact that post-stroke cognitive impairment (PSCI) has on patients and their families. Neuromodulation strategies have been increasingly applied in the clinical management of PSCI. This review outlines the mechanisms and brain function detection approaches through which neuromodulation promotes cognitive enhancement in stroke patients. For cognitive recovery, transcranial magnetic stimulation, transcranial electrical stimulation, vagus nerve stimulation, and brain-computer interfaces have shown promising results in clinical and preclinical studies. However, their efficacy remains unproven in large-scale pivotal trials. Preliminary clinical trials have shown that photobiomodulation enhances cognitive performance, but further investigation is required into the issue of skull attenuation of light. Transcranial ultrasound stimulation, a novel technology that overcomes the limitation of requiring deep electrode implantation for focal deep brain stimulation, still lacks scientific evidence. Chemogenetics and optogenetics provide methods for monitoring, disrupting, and regulating neural circuits after a stroke. To enhance the effectiveness of neuromodulation, it is recommended to implement multi-target stimulation, strengthen active participation in rehabilitation, and leverage cognitive-motor interactions to promote holistic recovery after stroke. Finally, we propose that neuromodulation will evolve toward brain-machine interaction neuromodulation, using artificial intelligence to develop a closed-loop strategy encompassing stimulation, detection, optimization, and re-stimulation.