Determination of appropriate imagery task to discriminate ERD of “Pinch” and “Hold” movements in healthy participants and stroke patients

Using video clips of hand movement to facilitate motor imagery and evoke strong ERD, we investigated the appropriate imagery task to discriminate ERD of “Pinch” and “Hold” movements in healthy controls and stroke patients. Participants watched 4 types of video clips that were a combination of 2 hand actions (pinch a small metal ball (“Pinch”) or hold a tennis ball (“Hold”)) and 2 object properties (ball rolls into the hand (“Move” condition) or stay still close to the hand (“Stop” condition)). The mean decrease in the mu band power from the baseline was 47.0±8.8% and 45.9±5.5% in healthy controls and patients, respectively, showing that all video clips evoked stable ERD. However, the relationship in ERD power between these hand actions varied across participants and object properties. We therefore calculated the absolute value of difference in ERD power between two hand actions (adERD) to further investigate the appropriate object property of the video clip stimuli. The mean adERD across participants and object properties was 22.3±6.0% in healthy controls and 33.9±6.6% in patients, showing a good contrast between the ERD powers between 2 hand actions. The adERD was significantly larger in Move condition than Stop condition at electrode C6 in healthy participants for left hand imagery. Similar tendency was also found at electrode C3 in right paralyzed patients. The post-experiment questionnaire demonstrated that the ease in hand motor imagery was greater in Move condition compared to Stop condition in all participant groups. These results suggest that the video clip stimuli with Move object properties are appropriate to induce stable ERD for classification of Pinch and Hold movements.