The Determination of the Number of Suspicious Clustered Micro Calcifications on ROI of Mammogram Images

摘要

Micro calcifications (MCCs) appear as a small cluster of white spots on mammographic images. Numerous researches have been conducted on this abnormality. However, most of the methods focus on MCCs detection without further processing of the original mammogram image. The purpose of this paper is to detect and determine the number of suspicious MCCs on the mammogram image. In the MCCs detection, the system allows the manipulation of mammogram image by using digital image processing techniques. An automated segmentation cluster of suspicious MCCs is done based on the region of interest (ROI). For MCCs detection and determination, this paper proposes the use of Contrast-Limited Adaptive Histogram Equalization (CLAHE), Morphological Tophat filtering, Sobel edge detection and Morphological operation. The number of MCCs from the ROI mammogram image is determined by using the process of morphological structuring. As a result, the approach has been successfully tested on a number of samples and returns an accurate detection of MCCs on the ROIs of the mammogram image.