Inverse distillation eeg epilepsy prediction model training, prediction method and device

Abstract

The application provides a reverse distillation stereoelectroencephalogram epilepsy prediction model training, prediction method and equipment, which can be applied to the field of artificial intelligence technology. The reverse distillation stereoelectroencephalogram epilepsy prediction model training method comprises the following steps: inputting a normal stereoelectroencephalogram picture segment of a target object into an expert network to obtain a plurality of reference features; inputting the normal stereoelectroencephalogram picture segment and an abnormal stereoelectroencephalogram picture segment into a teacher network to obtain a plurality of teacher network normal features and a plurality of teacher network abnormal features; inputting the plurality of teacher network normal features and the plurality of teacher network abnormal features into a student network to obtain a plurality of student network normal features and a plurality of student network abnormal features; updating the teacher network based on the differences between the plurality of reference features and the plurality of teacher network normal features, and the differences between the plurality of reference features and the plurality of teacher network abnormal features for a predetermined region, to obtain an optimized teacher network.

Images