Epilepsy detection method based on generative adversarial network

Abstract

The invention discloses an epilepsy detection method based on a generative adversarial network. According to the method, EEG signals during an epilepsy seizure are generated based on the generative adversarial network, experimental data samples are expanded, the expanded data samples are input into a CNN for training, then the real data samples are used for testing, and the aim of improving accuracy of epilepsy detection is achieved. The epilepsy detection method includes the specific steps which can be described as: step one, data are cut, downsampled and filtered, and processed data fragments are divided into a disease seizure class and a normality class; step two, the generative adversarial network is trained; step three, the trained generative adversarial network is used for generatingepilepsy seizure EGG data; and step four, a training set and a testing set are divided, the CNN is constructed for training and testing, and detection results are analyzed.

Description

technical field [0001] The invention relates to EEG signal processing and an epilepsy detection method, in particular to an epilepsy detection method based on a generative confrontation network. Background technique [0002] Brain-computer interface technology is an emerging technology with good development prospects. With the development of brain-computer interface technology, it plays an increasingly important role in many fields such as human-computer interaction and medical health. An important application of brain-computer interface technology is epilepsy detection, focusing on the processing and analysis of EEG signals. Although significant progress has been made in epilepsy detection methods based on EEG signals at home and abroad in recent years, due to the The influence of unbalanced, dynamic, unstable and other factors poses great challenges to the research of epilepsy detection methods. [0003] Currently, epilepsy detection methods are mainly divided into tradi...

AI Technical Summary

Problems solved by technology

In terms of epilepsy detection, there are also many new developments based on convolutional neural networks. However, when doing epilepsy detection experiments, using convolutional neural networks for training requires a large number of data samples, and most public data sets come from a small number of patients. , there are few samples in the data set and the sample size is small, which will have a certain degree of impact on the accuracy of the epilepsy detection results.

[0005] At present, when doing epilepsy detection experiments, we need to use a large number of data sets, generally speaking, we use public data sets, but most of the public data sets come from a small number of patients, the size of the data set is small, and the sample size is small. The CNN network needs a large number of data samples for training, which will cause large errors in the experimental results.

Images