Wavelet Analysis of Human Brain CT Images with Ischemic Stroke Region

Abstract

An application of wavelet transform for analysis of human brain CT images is described in this article. Wavelet transform, as a timefrequency signal analysis tool, has some advantages over other time or frequency image analysis methods. Particularly wavelet analysis is useful for simultaneous description and evaluation of time and frequency properties of the signals. The main focus of this article is evaluation of connection between appearance of ischemic brain stroke region in CT images and changes in wavelet spectrum. It was shown that ischemic brain stroke region is related with local low frequency wavelet coefficients and the main changes are observed at the contour of the region. It is recommended two concepts of ischemic stroke region detection, first, by grouping of wavelet coefficients with certain absolute value and detecting the contour of the region, second, by grouping small value coefficients and detecting the internal points of the region. Ill. 7, bibl. 9 (in Lithuanian; summaries in Lithuanian, English and Russian).