cInIn this study, an approach is proposed to develop a computer-aided classification system for mass and breast tissue (fat-connective-gland) characterization from digital mammograms using IDL programming language byto apply feature extraction for 9 features. The proposed system consists of two steps. The first step is the feature extraction, by usingexamining first order statistics using 3 features (mean-energy-standard deviation) and the classification accuracy of breast tissues and tumors is for Tumorwas: tumor - 96.8%, gland - 57.9%, and fat - 98.9, Whilewhile the connective tissue showed a classification accuracy of 98.5%. The overall classification accuracy of the breast area was 94.0%.
The second step is feature extraction by usingapplying second order statistics. and theThe classification accuracy of breast tissue and tumor showed a classification accuracy for of: tumor - 88.9%, gland - 98.9%, fat - 86.3%, and connective tissue - 91.9%. The overall classification accuracy of the breast area was 91.5%.
The overall accuracy was:
Overall accuracy 92.75%
tumor 92Tumor 92.85%
Glandular tissue 78.4%
fatt 92Fat 92.6%
Connective tissue 90.7%