| abstract |
The invention discloses a method and system for classifying multiple stained pathological images based on deep learning. The pathological image classification model uses a self-attention mechanism to fuse medical information from multiple stained pathological images, and classify multiple stained pathological images based on the fused information. The method utilizes artificial intelligence and deep learning technology, effectively integrates the features of various stained pathological images, realizes the accurate classification of various stained pathological images, helps to improve the efficiency and accuracy of pathological diagnosis, and provides clinicians with clinical features. policy support. |