Review on Mammogram Mass Detection by Machine Learning Techniques

Abstract

Mass Detection Feature Extraction Classification observed in detection mass are as follows: • Intensity levels vary greatly across different regions in a mammogram, and features for segmentation are hard to formulate. • Subtle grey level variations across different parts of the image make the segmentation of tumor areas by grey level alone difficult. • Tumors (masses) are not always obvious. • Poor illumination and high noise levels in the image that can vary up to 10-15% of the maximum pixel entity III. METHODOLOGY-EXISTING METHODS TO DETECT MASS IN MAMMOGRAM IMAGES Most of the mass detection algorithm will be performed in four phases. First, image preprocessing of the digitized mammogram can suppress noise and improve the contrast of the image. Second, image segmentation is defined by most of the articles about mass detection as locating the suspicious regions. It is different from the common definition of segmentation in image processing. In the third phase, features are extracted and selected for classifying lesion types or removing false positives. Finally, the detection/ classification of masses will be conducted. In this review preprocessing of mammogram image, detection of ROI and segmentation, feature selection and extraction, classification and evaluation are studied [5] [6]. Fig. 2. Illustrates the phases of mammogram mass detection