Analysis and classification of myocardial infarction tissue from echocardiography images based on texture analysis

Abstract

Texture analysis is an important characteristic for automatic visual inspection for surface and object identification from medical images and other type of images. This paper presents an application of wavelet extension and Gray level cooccurrence matrix (GLCM) for diagnosis of myocardial infarction tissue from echocardiography images. Many of applications approach have provided good result in different fields of application, but could not implemented at all when texture samples are small dimensions caused by low quality of images. Wavelet extension procedure is used to determine the frequency bands carrying the most information about the texture by decomposition images into multiple frequency bands and to form an image approximation with higher resolution. Thus, wavelet extension procedure offers the ability to robust feature extraction in images. The gray level co-occurrence matrices are computed for each sub-band. The feature vector of testing image and other feature vector as normal image classified by Mahalanobis distance to decide whether the test image is infarction or not.