Classification of Breast Cancer Based on Improved PSPNet

the effect of traditional PSPNet neural network on the semantic segmentation of mammographic images is not obvious enough, so the network is improved, and the practical application effect of two kinds of network on mammography detection is compared. the multi-layer image features extracted from the pyramid image features of PSPNet network were fused, which not only retained the original network to improve the effect of the receptive field, but also retained the detail features of the image through the fusion of multi-layer information. Results the detection performance of the experimental network was verified by using the mammography images of DDSM Database. Through the project verification, the improved network not only improves the semantic segmentation accuracy of the images, but also has obvious effect on the classification and diagnosis of breast cancer images. the improved PSPNet neural network has obvious semantic segmentation effect because it integrates the multi-level details of the images.