APPLICATION OF BACKPROPAGATION NEURAL NETWORKS TO DIAGNOSIS OF BREASTAND OVARIAN-CANCER

Neural network programs have been developed in an attempt to improve t he diagnosis of breast and ovarian cancer using a group of laboratory tests and the age of the patient. The laboratory tests employed in thi s study include albumin, cholesterol, HDL-cholesterol, triglyceride, a polipoproteins Al and B, NMR linewidth (the Fossel Index) and a tumor marker (i.e., CA 15-3 or CA 125). The breast cancer study involved 104 patients (45 malignant and 59 benign subjects). The ovarian cancer st udy involved 98 individuals (35 malignant, 36 benign and 27 control su bjects). Methods are outlined for identification of the most influenti al input parameters and optimization of network structure and training . Network characteristics were contrasted with the test results of the appropriate serum tumor marker assay. For the breast cancer study, th e best neural network program, using six input parameters, had a sensi tivity of only 55.6% and a specificity of 72.9%. The tumor marker CA 1 5-3 alone gave results of 61.3% and 64.4%, respectively. For the ovari an cancer study, the best neural network program, using six input para meters, had a sensitivity of 80.6% and a specificity of 85.5%. The tum or marker CA 125 alone gave results of 77.8% and 82.3%, respectively. These methods provide an objective approach to neural network optimiza tion and parameter selection applicable to other data bases of clinica l and laboratory data.