IEMI: Intelligent Extraction of Mental Imagery during Stroke Rehabilitation

Purpose and goal

IEMI has developed and implemented Artificial Intelligence (AI) techniques for real-time extraction of the continuous intensity of Mental Imagery (MI) of hand movements, from brain activity. Extending these achievements to a clinical setting, IEMI has fulfilled its goal and created a prototype system of a clinical decision support in stroke rehabilitation, that provides real-time feedback of MI to therapist and patient. An intervention protocol to clinically explore the decision support has been developed, implemented and tested in a small group of stroke patients.

Expected results and effects

IEMI specifically show that 1) the MI of opening and closing the hand (as compared to rest) can be extracted in real time from EEG brain activity, and 2) this measure can be used as continuous feedback through computer-simulated hand movements to stroke patients. These results are promising for motor rehabilitation of stroke patients with severe movement deficits and we hypothesize that the developed decision support will promote motor recovery by stimulating functional brain reorganization after stroke. Future studies will investigate these aspects of the decision support.

Planned approach and implementation

IEMI has been carried out in close collaboration between the academy of Innovation, Design, and Engineering at Mälardalen University and the department of Rehabilitation Medicine at Danderyd University Hospital. Through early collaboration, the decision support has gained its clinical relevance and user-perspective from the start. Technical aspects of the prototype have been iteratively evaluated and followed by implementation in a clinical setting. All aspects of the project have been conducted and approved according to the regulations of the ethical review authority.