Medical Image Retrieval Based on Gray Cluster Co-occurrence Matrix and Edge Strength Levels

Abstract

In Content Based Image Retrieval (CBIR), the objective is to query image databases in order to retrieve images with specific desired meaningful features. Important types of features include texture features and shape features. A common approach is to divide retrieval process into two stages; the first one is based on high-level features followed by the second that is based on low-level features. Within the previous mentioned general approach, methods vary in techniques they use. For each stage to extract features, and for matching feature sets based on these variations, existing methods perform differently in terms of precision, recall, and distance variance. In this research, we find that this variation in performance is an opportunity for us to propose a new method. We focus primarily on medical images, and follow the above approach but make the following two basic contributions: a) introduce the gray cluster co-occurrence matrix as texture feature extraction and use it as high-level features, and b) introduce edge strength levels as shape feature extraction and use it as low-level features. The system is evaluated using images retrieval performance the evaluation precision and recall rate, and distance variance. Our proposed system suggests the precision rate was 94.90% and recall rate was 89.72%. The distance variance achieved lowest rate (0.0022) in images retrieval compared to each of partial systems individually and related works. Our method has better performance in retrieving the results than other related works and each of partial system individually.