Worldwide, millions of people suffer every-year from cerebrovascular diseases. To date, stroke is one of the leading causes of long-term disability (HIV/AIDS), while the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is increasing the factors that can cause ischemic stroke due to thrombotic events (Nannoni et al., 2021). This leads to cognitive and motor impairments, resulting in loss of independence in the daily life of stroke patients, together with an additional psychological impact in mood disorders and depression. Evolving to a chronic condition, stroke requires continuous rehabilitation and therapy, a burden not only at an individual level but also at a broader level, affecting significantly the national health system (Burden of Stroke).

Technology-based approaches, like the use of personalised Virtual-Reality (VR) have been shown to accelerate the recovery process compared to traditional interventions. The utilization of VR is considered a novel and effective low-cost approach to re-train the lost motor and cognitive functions through strictly defined training tasks in a safe simulated environment (Laver et al., 2015). However, most of these novel VR approaches require increased volitional motor control, hence are suitable only for a reduced subset of patients, generally those with better recovery prognostics and better motor ability.

Consequently, the idea of directly training the central nervous system was established, through the use of Brain-Computer Interfaces (BCI’s) and motor-imagery (MI). BCI’s can be described as communication systems able to establish an alternative pathway between the user’s brain activity and a computer system, providing an additional non-muscular channel for communication and control to the external world (Wolpaw et al., 2002).

Prior research has shown that mental practice of action is useful in MI-BCI training, for attaining functional motor recovery through the reorganization of motor networks (Pichiorri et al., 2015). Moreover, MI-BCI training can promote long-lasting improvements of motor function in stroke patients (Ramos-Murguialday et al., 2019). Nonetheless, although the benefits of MI-BCI have been illustrated in recent studies (Vourvopoulos et al, 2019), interventions with patients in longitudinal studies is still limited, lacking long-term evidence to support its clinical relevance. One of the major reasons is due to the reduced ability of stroke patients to accurately control a BCI system, resulting in poor skill acquisition during training (Lotte et al., 2012).

The aim of this project is to develop a novel and more inclusive rehabilitation system with the use of emerging technologies, in order to overcome current limitations of MI-BCI training for rehabilitative applications. This will be achieved by identifying the neural correlates of motor action and skill acquisition during motor imagery through neuroimaging (EEG/fMRI), formulate machine-learning methods for increased accuracy of the BCI system, but also potentiate human training for skill acquisition through the use of neuroadaptive VR.