A novel algorithm for initial lesion detection in ultrasound breast images

Abstract

This paper proposes a novel approach to initial lesion detection in ultrasound breast images.  The objective is to automate the manual process of Region of Interest (ROI) labeling in Computer-aided Diagnosis (CAD). We propose the use of hybrid filtering, multifractals processing and thresholding segmentation, in initial lesion detection and automated ROI labeling.  A total of 360 ultrasound breast images have been used to evaluate performance of the proposed approach.  Histogram equalization is used to pre-process the images followed by hybrid filtering and multifractal analysis. Subsequently thresholding segmentation is applied on the image. Finally the initial lesions are detected using a rule based approach. The accuracy of the automated Region-of-Interest (ROI) labeling is measured by an overlap of 0.4 with the lesion outline compared to the lesions labeled by an expert radiologist. We compare the proposed method to three existing state-of-the-art methods, namely the radial gradient index filtering technique, the local mean technique, and the fractal dimension technique. We conclude that the proposed method is more accurate and performs more effectively when compared to the benchmarks.