A structured summary should contain the following subheadings: Rationale/ gaps in existing knowledge, Novelty, Objectives, Methods, and Expected outcome.

Rationale: Stroke is a leading cause of disability globally. Passive movement and motor imagery (MI) induce similar EEG patterns over the motor cortex. Motor

imagery based Brain computer interface systems induces brain plasticity and the modulations in the sensory motor rhythm embedded in the EEG can be decoded

and translated for control of external devices. The identification and decoding of brain signal patterns associated with MI parameters such as speed and direction is

a challenging research problem .In BCI-based neurofeedback, the user’s imagination of hand movement decoded by the BCI is presented to the user as visual

feedback. The erroneous neurofeedback perceived by the human brain in the case of erroneous hand movement direction decoding triggers an endogenously

elicited brain response termed as Error-Related Potential (ErrP) and can be used to correct the error. Objectives: We aim to develop an online EEG-based MI-BCI for

stroke rehabilitation by decoding imagined hand movement kinematics and improve the accuracy of the MI decoding by integrating with ErrP detection. Novelty:

Integrating the online realtime multi class MI decoding module with the ErrP detection module to develop an online self-correcting BCI that actuates robotic hand

support towards the imagined direction/speed Methods The study design will be a feasibility randomised control study with 15 control subjects of stroke

undergoing conventional physiotherapy, 15 subjects of stroke undergoing MI” Intervention and another 15 subjects undergoing “MI with ErrP” Intervention. The

functional performance metrics will be documented pre and post BCI intervention and the efficacy assessed.