Zhaohuan Ding1, Gaoxiang Ouyang, He Chen, Xiaoli Li. 1. State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, People's Republic of China.
Abstract
OBJECTIVE: Transcranial magnetic stimulation (TMS) as a safe, noninvasive brain regulation technology has been gradually applied to clinical treatment. Traditional TMS devices do not adjust output based on real-time brain activity information when regulating the cerebral cortex, but the current activity information from the brain, especially the EEG phase, may affect the stimulation effect. It is necessary to calculate the synchronous EEG phase during TMS. APPROACH: In this study, a set of closed-loop TMS device a fast EEG phase prediction algorithm based on the AR model was designed to meet the demand. EEG data for twenty-seven healthy college students were collected to verify the accuracy of the algorithm. MAIN RESULTS: The calculation results showed that the prediction accuracy of the AR model algorithm is better than that of the conventional algorithm when the model order is lower, and the prediction accuracy will increase with improvements in the signal quality. SIGNIFICANCE: When the experimental environment is good, the EEG data with a high SNR can be recorded, and when the order of the AR model is properly set, the prediction algorithm can make correct judgments most of the time and the stimulation pulse can be output when the EEG phase reaches a set value.
OBJECTIVE: Transcranial magnetic stimulation (TMS) as a safe, noninvasive brain regulation technology has been gradually applied to clinical treatment. Traditional TMS devices do not adjust output based on real-time brain activity information when regulating the cerebral cortex, but the current activity information from the brain, especially the EEG phase, may affect the stimulation effect. It is necessary to calculate the synchronous EEG phase during TMS. APPROACH: In this study, a set of closed-loop TMS device a fast EEG phase prediction algorithm based on the AR model was designed to meet the demand. EEG data for twenty-seven healthy college students were collected to verify the accuracy of the algorithm. MAIN RESULTS: The calculation results showed that the prediction accuracy of the AR model algorithm is better than that of the conventional algorithm when the model order is lower, and the prediction accuracy will increase with improvements in the signal quality. SIGNIFICANCE: When the experimental environment is good, the EEG data with a high SNR can be recorded, and when the order of the AR model is properly set, the prediction algorithm can make correct judgments most of the time and the stimulation pulse can be output when the EEG phase reaches a set value.