AANN-based Classification Technique of Benign and Malignant Tumors
Zhong Ming-xia · Computer Era · 2010
The early detection of breast lesions remains the primary method for preventing breast cancer despite the continuous advance in the diagnosis and treatment technology of breast cancer.The presence of mass in breast tissues is the important feature of breast cancer.The classification technique of benign and malignant tumors is researched by using AANN and MLP techniques.The experiment result shows that the higher sensitivity of CAD systems (TP) and lower false positive rate (FP) are achieved in the training and testing on DDSM database,91% testing classification rate on 100% training classification rate and the maximum area under ROC curve up to 0.948.