AEEG signal image feature reconstruction, representation and extraction method

Abstract

The invention discloses an AEEG signal image feature reconstruction, representation and extraction method. The method includes the following steps of signal segmentation, contour map drawing and image feature extraction, wherein in the signal segmentation step, an AEEG signal is obtained and segmented into multiple signal segments; in the contour map drawing step, the frequencies of different amplitudes in each signal segment are counted, points with the same frequency are connected, and a contour map is drawn; in the image feature extraction step, image features in the contour map are extracted. According to the method, after the contour map is drawn, color distinguishing in the amplitude frequency contour map of the AEEG signal is obvious, distribution conditions of different amplitudes can be visually displayed, accordingly doctors are helped to better define an upper boundary and a lower boundary of the amplitudes, and the image features can be further extracted on the basis of the reconstructed amplitude frequency contour map.

Description

technical field [0001] The patent of the present invention relates to a feature representation and extraction method of amplitude-integrated EEG, especially a feature representation and extraction method of aEEG signal based on image information. Background technique [0002] Amplitude-integrated electroencephalography (aEEG) technology is an effective means of monitoring brain function in newborns (Hellstrom–Westas et al., 2008), and has been widely used as an auxiliary examination method for neonatal brain injury. Doctors provide objective evaluation indicators for neonatal hypoxic-ischemic encephalopathy, asphyxia, epilepsy, intracranial hemorrhage, and hydrocephalus, as well as the screening of encephalopathy treatment targets. Usually the aEEG signal is a strip-shaped track, and the general amplitude is between 10 and 40 μV. Its width is related to the functional state of the neonatal brain, and the trace is usually wider during neonatal deep sleep, while the aEEG sign...

AI Technical Summary

Problems solved by technology

However, the shade boundary of the background image is usually not obvious, and different people have different judgments on the shade of the color, resulting in different interpretation results of the same aEEG signal by different doctors

In the automatic analysis of aEEG, the current work is basically feature extraction based on signal amplitude, and the use of image features has not appeared.

Images