Paper Title
An Identification Method Of Normal And Edematous Humeral Head From Pd Weighted Mr Images Based On Glcm Texture Features

Bone edema is an important clinical issue that may cause considerable pain. Patients with shoulder pain may have bone edema frequently. Bone edema is detected in PD weighted MR images however presence of bony edema may hamper both visual and automatic segmentation of bone from the PD weighted MR images. The automatical detection of the presence and the distribution of the bone edema may help clinicians to focus on the location and the magnitude of the bone pathology. The objective of this research is to develop a computer aided detection (CAD) system which is capable of recognizing normal and abnormal humeral head images by using texture features derived from gray level co-occurrence matrices (GLCM). In this study GLCM were calculated with using symmetric and non-symmetric matrices which have different displacement and orientation. GLCM matrices were used in Haralick algorithm to extract feature of the region of interest. To verify the performance of the extracted features, we deployed MLP (Multilayer Perceptron), SVM (Support Vector Machine) and KNN (K nearest neighbors) methods and we demonstrated their power in differentiating the normal and abnormal regions. The proposed approach was tested on our own dataset which consists of 79 normal and 91 edematous humeral heads in PD weighted MR images. The resulting classification accuracy produced by the SVM and KNN were 94, 11% and MLP was 93, 52 %. The proposed system is a promising tool for classification of edematous and normal humeral heads from PD weighted MR images. Keywords: GLCM, image texture analysis, pattern classification, PD weighted MRI