Early detection of masses in digitized mammograms using texture features and neuro-fuzzy model

A neuro-fuzzy model for fast detection of candidate circumscribed masses in digitized mammograms is presented. The breast tissue is scanned using variable window size, for each sub-image co-occurrence matrices in different orientations (/spl theta/=0/spl deg/, 45/spl deg/, 90/spl deg/ and 135/spl deg/) are calculated and texture features are estimated for each co-occurrence matrix, then the features are used to train neuro-fuzzy models. The classification results reach 100% for abnormal cases and 80% for normal ones.