CLASSIFYING BENIGN AND MALIGNANT MASSES USING STATISTICAL MEASURES

Abstract
Breast cancer is the primary and most common disease found in women which causes second highest rate of death after lung cancer. The digital mammogram is the X-ray of breast captured for the analysis, interpretation and diagnosis. According to Breast Imaging Reporting and Data System (BIRADS) benign and malignant can be differentiated using its shape, size and density, which is how radiologist visualize the mammograms. According to BIRADS mass shape characteristics, benign masses tend to have round, oval, lobular in shape and malignant masses are lobular or irregular in shape. Measuring regular and irregular shapes mathematically is found to be a difficult task, since there is no single measure to differentiate various shapes. In this paper, the malignant and benign masses present in mammogram are classified using Hue, Saturation and Value (HSV) weight function based statistical measures. The weight function is robust against noise and captures the degree of gray content of the pixel. The statistical measures use gray weight value instead of gray pixel value to effectively discriminate masses. The 233 mammograms from the Digital Database for Screening Mammography (DDSM) benchmark dataset have been used. The PASW data mining modeler has been used for constructing Neural Network for identifying importance of statistical measures. Based on the obtained important statistical measure, the C5.0 tree has been constructed with 60-40 data split. The experimental results are found to be encouraging. Also, the results will agree to the standard specified by the American College of Radiology-BIRADS Systems.

Authors
B. Surendiran1 and A. Vadivel2
National Institute of Technology Tiruchirappalli

Keywords
Benign and Malignant Mammogram Masses, Feature Extraction, Weight Function, Statistical Measures, Neural Network, C5.0 Decision Tree Classifier
Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 2 , Issue: 2 )
Date of Publication :
November 2011

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.