Epilepsy electroencephalogram classification method based on supervised feature fusion algorithm

Abstract

The invention discloses an epilepsy electroencephalogram classification method based on a supervised feature fusion algorithm. The method comprises the following steps: firstly, extracting power spectral density and fluctuation indexes of frequency slice wavelet transform from an electroencephalogram fragment as features; next, using a supervised locality preserving canonical correlation analysis algorithm, obtaining the optimal projection direction by maximizing the weight correlation between paired samples in the class and neighbors of the paired samples, and the projection combination of the original features in the optimal projection direction is fusion features; and the fusion features are input into a least square support vector machine for training and testing. The method is verified on a Born data set and a CHB-MIT data set, and a good result is obtained. In addition, the parameter sensitivity of the supervised locality preserving canonical correlation analysis algorithm and the relationship between the dimension of the fusion feature and the classification result are also discussed, and the stability and effectiveness of the method are further verified.

Description

technical field [0001] The invention belongs to the field of biological signal processing, and relates to an epileptic EEG classification method based on a supervised feature fusion algorithm. Background technique [0002] Epilepsy is a sudden, recurrent, involuntary brain disease. According to the World Health Organization, nearly 50 million people worldwide suffer from epilepsy. At present, the clinical diagnosis of epilepsy mainly relies on medical history data and brain examination, such as visual inspection of long-term EEG by experienced doctors. However, this method is time-consuming and the results are subjective. In order to solve this problem, researchers have proposed a large number of methods to identify epileptic EEG signals through signal processing and machine learning methods. These methods can reduce the burden on doctors and improve the accuracy of diagnosis. [0003] Due to the obvious nonlinearity and non-stationarity of epileptic EEG signals, researc...

AI Technical Summary

Problems solved by technology

Although there are already many dimensionality reduction algorithms for a single feature, these algorithms do not take the relationship between features into account well.

It may lead to redundant information between different features, thereby increasing unnecessary feature dimensionality

Images