BREAST-TUMOR CLASSIFICATION BY NEURAL NETWORKS FED WITH SEQUENTIAL-DEPENDENCE FACTORS TO THE INPUT LAYER

We applied an artificial neural network approach to identify possible tumors into benign and malignant ones in mammograms. A sequential-depe ndence technique, which calculates the degree of redundancy or pattern ing in a sequence, was employed to extract image features from mammogr aphic images. The extracted vectors were then used as input to the net work. Our preliminary results show that the neural network can correct ly classify benign and malignant tumors at an average rate of 85%. Thi s accuracy rate indicates that the neural network approach with the pr oposed feature-extraction technique has potential utility in the compu ter-aided diagnosis of breast cancer.