KDD 2022 | EEG AI helps diagnose epilepsy

Epilepsy is one of the most common serious neurological diseases , Surgical removal of brain regions involved in seizures is considered the only effective treatment . To locate the seizure area （SOZ） To guide epilepsy surgery , It is necessary to record and analyze the discharge activity of the patient's brain . There are usually two types of electrophysiological monitoring methods ： Cortical EEG （EEG） and Stereoscopic EEG （SEEG）.

lately , Yang Yang's research group from Zhejiang University cooperates with a third class hospital , A collection based on real patient records SEEG Data sets , The data set contains high-frequency multi-channel data of multiple epileptic patients SEEG The signal . Based on this dataset , Researchers combine graph and timing techniques , This paper studies and proposes a method that can automatically predict epileptic waves AI Model . The model can be independent of expert experience , Confirm the time and position of epileptic wave in brain wave by data-driven method , The experimental effect is very remarkable . At present, this work has been recognized by the top conference of data mining KDD 2022 Included .

Address of thesis ：http://yangy.org/works/brainnet/KDD22_BrainNet.pdf

 

The figure below shows BrainNet The overall structure and process of ：

a： An electrode with three channels is implanted into the patient's brain to record SEEG data ;b：BrainNet The overall solution ： Learn the diffusion pattern of brain waves at multiple levels , So as to predict such as a The epileptic band shown in the dotted box in the