Share Email Print

Proceedings Paper

Lesion location algorithm of epilepsy signals based on image features
Author(s): Hui Yi; Chunsheng Wang; Wei Wang; Zijian Liu
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Aiming at the problem that the traditional one - dimensional linear processing method can’t effectively extract the characteristic information of the weak signal of the brain, this paper presents a lesion location algorithm of electroencephalogram (EEG) based on image features. Image features is introduced into epileptic signal processing, which can extract feature information and locate lesions in EEG weak signals. First of all, the theoretical basis of onedimensional signal reconstruction into two-dimensional signal is described in this paper. Then, in order to make the image have the intuitionistic distinguishing feature, the Hadamard transform is introduced, which amplifies the original degree of chaos and provides the basis for qualitative and quantitative analysis. Finally, Gabor filter and Fourier spectrum analysis are used to process the Hadamard transformed image. The processed image retains some larger details such as the edge and contrast, which makes the processed image result more intuitive. The lesion lead images can be distinguished by the intuitive differentiation. At the same time, the mean value, contrast and energy entropy of the Fourier spectrum are calculated, and the results are clear. The effectiveness of the proposed method is verified by real epileptic EEG data.

Paper Details

Date Published: 6 May 2019
PDF: 12 pages
Proc. SPIE 11069, Tenth International Conference on Graphics and Image Processing (ICGIP 2018), 110690U (6 May 2019); doi: 10.1117/12.2524206
Show Author Affiliations
Hui Yi, Central South Univ. (China)
Chunsheng Wang, Central South Univ. (China)
Wei Wang, Guangxi Univ. (China)
Zijian Liu, Central South Univ. (China)

Published in SPIE Proceedings Vol. 11069:
Tenth International Conference on Graphics and Image Processing (ICGIP 2018)
Chunming Li; Hui Yu; Zhigeng Pan; Yifei Pu, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?