Advancing Post‐Stroke Rehabilitation: Emerging and Current Neuromodulation Approaches and Integration of Artificial Intelligence‐Driven Closed‐Loop Systems

AI is almost completely worthless until the underlying research for 100% recovery is there! 

 Advancing Post‐Stroke Rehabilitation: Emerging and Current Neuromodulation Approaches and Integration of Artificial Intelligence‐Driven Closed‐Loop Systems

Sensory Neuroscience

- PERSPECTIVEOPEN ACCESS

Advancing Post‐Stroke Rehabilitation: Emerging and Current Neuromodulation Approaches and Integration of Artificial Intelligence‐Driven Closed‐Loop Systems

Tiago Cunha Reis1 | Ana Machado2

1Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal | 2Champalimaud Neuroscience Programme, Champalimaud Centre for the Unknown,

Lisbon, Portugal

Correspondence: Tiago Cunha Reis (reis.tiago@edu.ulisboa.pt)

Received: 6 January 2025 | Revised: 7 February 2025 | Accepted: 12 February 2025

Keywords: artificial intelligence | brain–computer interfaces | chronic stroke | deep brain stimulation | neuromodulation | vagus nerve stimulation

ABSTRACT

Chronic stroke represents a significant global health burden, requiring innovative rehabilitation strategies that extend beyond conventional therapies. Neuromodulation, including transcutaneous vagus nerve stimulation, deep brain stimulation, and brain–computer interfaces, has emerged as a transformative approach, leveraging neuroplasticity to enhance motor and cognitive recovery. Integrating artificial intelligence (AI) within these modalities enables adaptive, patient‐specific interventions through real‐time feedback, predictive modeling, and advanced signal processing. This perspective article provides a comparative analysis of these neuromodulation techniques, examines clinical evidence, while also identifying AI‐centric research priorities to address current challenges.

1 | Introduction

The increasing survival rates of acute stroke patients have underscored the need for advanced rehabilitation strategies for the chronic phase of stroke [1, 2]. Traditional physical rehabilitation, while effective in promoting neuroplasticity, often yields diminishing returns after the first year post‐stroke. With a significant portion of global healthcare expenditure allocated to stroke management and an aging population driving increased incidence, innovative interventions that extend or enhance recovery trajectories are essential.

Neuromodulation, applying electromagnetic energy to modulate neural activity, has emerged as a promising adjunct to conventional therapies [3]. Both invasive and non‐invasive

modalities have demonstrated the capacity to induce synaptogenesis and functional reorganization, critical processes in post‐stroke recovery [4, 5]. Recent breakthroughs in technologies

like vagus nerve stimulation (VNS), deep brain stimulation(DBS), and brain–computer interfaces (BCIs) offer novel pathways to harness these mechanisms (Figure 1). Integrating AI into these platforms holds the potential to create adaptive, patient‐specific therapeutic systems, transforming the neuromodulation landscape