The pathologic diagnosis of tumors involves many aspects of analysis and interpretation of plenties biomakers, however, the accuracy of pathologic diagnosis is limited by a shortage of differences in pathologists' diagnostic skill, and the availability of ancillary studies. We have developed an artificial intelligence-assisted (AI-assisted) diagnostic tool, which can reduce tedious workload for pathologists, improve their efficiency and accuracy, provide new information of disease prognosis and therapy response.We established a deep learning-based AI-assisted model, using cell detection and region segmentation algorithm and multi-instance deep learning. The AI model analysis algorithm for this study is using digital pathology slide scanner for machine learning and split the WSI image into more than 100000 patches. The AI algorithm in this study is capable of obtaining the WSI digital images through modules such as foreground segmentation, image block prediction, tumor-related interstitial region estimation, and result integration. In this study, we evaluated the consistency of interpreting plenties biomarkers among pathologists.The AI-assisted model can help different levels of pathologists interpret biomarkers of breast cancer, which achieved excellent consistency and repeatability. The model based on deep learning can accurately predict the prognosis of breast cancer, and the prediction performance was good. With the increasing application in breast pathology, AI not only assists in pathology diagnosis to improve accuracy and reduce pathologists' workload, but also provides new information in predicting prognosis and therapy response.
