2021.08.30
Overview
A method for processing breast tomosynthesis images includes: obtaining multiple breast images, performing binarization operation and pooling operation to at least one marked breast image to obtain a disease mask; training a convolutional neural network (CNN) according to the breast images and the disease mask; obtaining multiple cropped images from breast images at different layers and a particular two-dimensional position, and inputting each cropped image into the CNN to obtain cropped heatmaps; inputting the cropped heatmaps into a recurrent neural network (RNN) in an order of the layers to obtain output heatmaps, and combining the heatmaps to obtain a first integrated heatmap, and training the RNN according to the first integrated heatmap and the disease mask.Category
發明
Patented
17/446,457
US 12,008,688 B2
Filing Date
2021.08.30Expired Date
2042.09.16Notification
2023.11.10