Classification of medical images using data mining techniques

Abstract

Automated classification of medical images is an increasingly important tool for physicians in their daily activity. This paper proposes data mining classifiers for medical image classification. In this study, we have used J48 decision tree and Random Forest (RF) classifiers for classifying CT scan brain images into three categories namely inflammatory, tumor and stroke. The proposed classification system is based on the effective use of texture information of images. Three different methods implemented are: Haralick (H), Tamura (T) and Wold (W) texture features. All three texture features and the classification methods are compared based on Precision and Recall. The experimental result on pre-diagnosed database of brain images showed Haralick features combined with Random Forest classifier is found to give best results for classification of CT-scan brain images. © 2012 ICST Institute for Computer Science, Social Informatics and Telecommunications Engineering.