A cancer detection device utilizing multi-tiered neural networks for improved classification

We introduce a multi-tiered neural network architecture that accurately classifies malignant breast tissue from benign breast tissue. The methodology implemented six different backpropagation neural network (BNN) architectures on 180 malignant and 180 benign breast tissue impedance data files sampled at 47 frequencies from 1 hertz (Hz) to 32 megahertz (MHz). The data were collected utilizing a NovaScan cancer detection prototype device in an approved IRB study at Aurora Medical Center, Milwaukee. The BNN analysis consists of a multi-tiered consensus approach autonomously selecting 4 of 6 neural networks to determine a malignant or benign classification. The BNN analysis was then compared to the histology results with consistent sensitivity of 100 percent and a specificity of 100 percent. This implementation successfully relied solely on statistical variation between the histologically confirmed benign and malignant impedance data and intricate neural network analysis. This approach could be a valuable tool to augment current medical practice assessment of the health of breast and other tissue.