Stockwell transform based epilepsy electroencephalogram detection device

Abstract

The invention discloses a Stockwell transform based epilepsy electroencephalogram detection device. The device comprises an electroencephalogram amplifier, a data acquisition card and a computer. An electroencephalogram signal preprocessing module, an electroencephalogram signal segmentation module, a Stockwell transform feature extraction module, a classification module and an after-processing module are arranged in the computer. Feature extraction of time-varying unstable epilepsy electroencephalogram signals is realized by means of Stockwell transform, and epileptic seizures in the electroencephalogram signals are detected through the computer. The device is effective in detection of epilepsy electroencephalogram signals and suitable for online electroencephalogram classification detection.

Description

technical field [0001] The invention relates to the field of automatic epilepsy EEG detection technology equipment, in particular to an epilepsy EEG detection device based on Stockwell transformation. Background technique [0002] Epilepsy is a common neurological disorder characterized by hypersynchronous discharge of brain neurons and characterized by paroxysmal, acute, and temporary brain dysfunction. According to incomplete statistical results, the prevalence of epilepsy is 0.5%-2%. Since epilepsy is a chronic disease, recurrent epileptic seizures have brought great harm to patients both physically and mentally. sometimes even life-threatening. The most effective method for diagnosing epilepsy is electroencephalogram (EEG), which is the spontaneous and rhythmic electrical activity of brain neurons recorded from extracranial scalp or intracranial electrodes. At present, EEG detection of epilepsy is mainly done by medical workers relying on experience to visually inspect...

AI Technical Summary

Problems solved by technology

For example, the automatic detection of epilepsy based on DWV (differential windowed variance) has a detection sensitivity of 89.3%, which cannot meet the accuracy requirements of epilepsy EEG detection (see Majumdar KK, Vardhan P. Automatic seizure detection in ECoG by differential operator and windowed variance.IEEE transactions on neural systems and rehabilitation engineering, 19(4):356-365, 2011)

CN1255320 (CN99124210.6) discloses a method and device for automatically determining the threshold value of artificial neural network for identifying epileptiform discharges. This method needs to determine multiple network parameters and a large number of samples for repeated training, the training speed is slow, and the clinical application has limitations.

Images