A Method for Precisely Determining Blink Intervals in EEG or Oculometry Using Outlier Detection

Abstract

The invention relates to a method for utilizing abnormal value detection to accurately determine a winking interval of an electroencephalogram or electro-oculogram. According to the method, firstly, collected original signals are pretreated to remove noise interference with a 50-Hz power frequency, baseline drifts and high-frequency noise in the collected signals; secondly, an appropriate threshold value of the pretreated signals is determined by means of an abnormal value detection method, then all logical maximum values and the positions of the logical maximum values in the pretreated signals are found, the adjacent local maximum values are used for subtraction, the absolute value of a difference between every two adjacent local maximum values is compared with the threshold value determined by the abnormal value detection method, and finally, the starting point and ending point of a winking signal are accurately determined. The starting point and ending point of the winking signal inthe electroencephalogram or electro-oculogram signals can be precisely determined, and therefore when winking interference is removed, the interval can return to zero, so that loss of the original signals can be minimized to the greatest extent.

Description

technical field [0001] The invention relates to a method for accurately determining the blink interval in electroencephalogram or eye electricity by using outlier detection. Background technique [0002] Electroencephalogram (EEG) and electrooculogram (EOG) are bioelectrical signals of the human body, which can be used as control signals in some human-computer interaction systems, which can provide convenience for patients with movement disorders. However, as a normal physiological response, unconscious blinking has a large amplitude and cannot be avoided. The blinking signal will become an interference signal and cause the system to misoperate. Therefore, it is necessary to remove the blinking signal in the EEG or eye electricity. In the literature studies that have been published so far, there are mainly the following methods for removing the eye blink signal from the EEG or electrooculogram signals. (1) Use the Independent Component Correlation Algorithm (ICA) method to ...

AI Technical Summary

Problems solved by technology

The above method is to decompose the original signal, find the channel with the greatest correlation with the blink signal and set it to zero directly, which will cause some meaningful signal loss in the original signal

(2) Use the original signal to derive the derivative within the sliding window to determine the position of the blink, but the derivative value is rough as a blink judgment, and the size of the sliding window will also affect the result

(3) Use the wavelet packet to decompose the original signal in the frequency domain, and combine the statistical theory to determine the threshold, and use the threshold judgment criterion and reconstruction strategy to remove the blink signal for the low-frequency components decomposed by the wavelet packet. However, due to the limited number of samples, based on statistics The threshold value determined by the information is not effective enough to accurately determine the blink area

Images