The invention provides an
epileptic seizure prediction method, device, equipment and program product, and the method comprises the steps: obtaining a normal electroencephalogram
signal baseline of a
dura mater of an epileptic in an epileptic non-seizure period, and a personalized feature
library, the personalized feature
library comprises features of the cerebral
dura mater electroencephalogram signals of the epileptic within at least one time period before
epileptic seizure; acquiring an electroencephalogram
signal of the current
dura mater of the
epilepsy patient; according to the normal electroencephalogram
signal base line, the personalized feature
library and the current electroencephalogram signal of the dura mater, predicting whether the
epilepsy is about to seizure
epilepsy or not; under the condition that the epilepsy is about to seize, a first early warning signal is sent out, and the first early warning signal is used for reminding an epileptic that the epilepsy is about to According to the prediction method, the electroencephalogram characteristics of the patient can be reflected more accurately, the
false alarm rate and the missing report rate caused by individual differences are reduced, the accuracy of
epileptic seizure prediction is improved, and the robustness of the prediction method is enhanced.