Optimized textural features for mass classification in digital mammography using a weighted average gravitational search algorithm

Early detection of breast cancer cells can be predicted through a precise feature extraction technique that produce efficient features. The application Gabor filters, gray level co-occurrence matrices (GLCM) and other textural techniques have proven to achieve promising results but were often characterized by high false-positive rate (FPR) false-negative (FNR) with computational complexities. This study optimized features for mass classification in digital mammography using the weighted average gravitational search algorithm (WA-GSA). GLCM fused WA-GSA overcome weakness techniques. With support vector machine (SVM) used as classifier, proposed was compared commonly applied Experimental show SVM achieved FPR, FNR accuracy 1.60%, 9.68% 95.71% at 271.83 s, respectively. Meanwhile, 3.21%, 12.90% 93.57% 2351.29 respectively, while 4.28%, 18.28% 91.07% 384.54 obtained prevalence algorithm, WA-GSA, tumor detection.