During the training process, the training model for distinguishing and classifying malignant breast cancer and benign disease is shown in Figure 1. It contains a two-stage neural network classification unit, input vector processor, input size controller, training control center, learning rate controller, addition operation unit, target output class, weight update unit, and training weights unit.