gms | German Medical Science

Objective: Motor learning plays an important role for rehabilitation after intracranial surgery in a variety of cortical and subcortial regions. Already for planning of surgery, identification of essential functional areas is crucial. However, currently available methods are regarded as too conservative. In addition, neuroimaging results typically present activation patterns with unclear relation to the actual motor performance. We therefore investigated EEG correlates of motor learning using archery training of novices as a model. The presentation will show results from the first participant.

Methods: Participants, who had no previous experience with archery, underwent intensive archery training with 100 shots on each of three days. EEG with 64 channels was recorded during the individual shots. Data was preprocessed and inspected for artifacts and noisy channels. Data from 800 to 200ms before the shot were then analyzed using two procedures:

1.

Distribution of theta-activity over the brain and consecutive training days and

2.

Prediction of shot accuracy based on machine learning (ML)analysis.

Results: Theta activity increased significantly over the course of the three training days. Changes were wide-spread with maxima in left frontal areas. The increase of theta paralleled the increase of achieved scores. ML was able to predict scores from EEG electrode data in the theta band with AUC-scores (area-under the curve of receiver-operator characteristic analysis) ranging up to 86.1 on all three days.

Conclusion: Theta may thus represent a neurophysiological marker of motor performance. Increase of theta then shows the effect of learning. Mapping of theta may thus provide valuable information for planning of surgery and motor rehabilitation.