Multi-channel electroencephalogram automatic epilepsy detection device based on one-dimensional CNN-LSTM (convolutional neural network-long short term memory)

Abstract

The invention discloses a multi-channel electroencephalogram automatic epilepsy detection device based on one-dimensional CNN-LSTM (convolutional neural network-long short term memory). The multi-channel electroencephalogram automatic epilepsy detection device based on the one-dimensional CNN-LSTM comprises a computer. The computer is programmed to perform the following step: acquiring collected data, wherein the collected data is collected by placing data acquisition electrodes according to a 10-20 system electrode method meeting the international standard. The multi-channel electroencephalogram automatic epilepsy detection device has the advantages that different from conventional epileptic seizure detection, the device does not need manual feature design for classification; instead, multi-channel original signals are input into a training network directly, features of the signals are automatically learned through one-dimensional CNN and LSTM neural networks, and finally, classification is performed; due to the multi-channel signals, the effect is better than that of a method only using single-channel signals, so that the stability and universality are achieved; the device has agood effect in practical clinical data in addition to excellent performance in a database.

Description

Technical field [0001] The invention relates to the field of epilepsy detection, in particular to a multi-channel EEG epilepsy automatic detection device based on one-dimensional CNN-LSTM. Background technique [0002] Epilepsy is a major neurological disease caused by abnormal electrical activity in the brain. Due to the different starting location and transmission mode of abnormal discharge, the clinical manifestations of epileptic seizures are complex and diverse, which can be manifested as paroxysmal motor, sensory, autonomic, consciousness and mental disorders. These onset symptoms bring great inconvenience to the patient's life. [0003] EEG signals contain a large amount of physiological and disease information. Clinically, EEG signals are often used to analyze and diagnose certain brain diseases. The traditional automatic epilepsy detection is to extract certain features from the EEG to classify whether the epilepsy has seizures, which can greatly reduce the workload of d...

AI Technical Summary

Problems solved by technology

[0009] 1. The extraction of traditional features requires manual design of features, which brings some problems:

[0010] (1) Artificially designed features rely on a large amount of prior knowledge, which is difficult

[0011] (2) Additional feature extraction and selection increase the complexity of the algorithm;

[0013] (4) When using fixed artificially designed features, it is dif

Images