Pathological image processing method and model training method and device based on deep learning

Abstract

The invention discloses a pathological image processing method based on deep learning. The pathological image processing method comprises: obtaining a pathological image of a to-be-processed target object; partitioning the pathological image of the target object to be processed to obtain a partitioned image set; performing feature classification extraction processing on the block image set to obtain N first feature sets, N second feature sets and N third feature sets; generating a target fusion feature corresponding to the target object; and calling the pathological image analysis model to analyze and process the target fusion features so as to output a pathological analysis result of the to-be-processed target object. The invention further discloses a model training method. The pathological image can be directly utilized to predict the risk of lesion metastasis. Compared with a biological detection technology, the method is advantageous in that the time for waiting for a detection result can be shortened, the detection efficiency can be improved, errors caused by uncontrollable factors such as experimental errors and operation errors can be avoided, and then a more accurate detection result is provided.

Description

technical field [0001] This application relates to the field of artificial intelligence, in particular to a pathological image processing method, model training method and device based on deep learning. Background technique [0002] Colorectal cancer is a malignant tumor with high morbidity and mortality. In patients with colorectal cancer, as the primary cancer cells infiltrate, they break through the intestinal wall layer by layer, and may infiltrate to the lymph nodes near the intestine, and further metastasize to distant lymph nodes through the lymphatic system. Before resection of colorectal cancer, it is necessary to accurately predict lymph node metastasis, so as to assist doctors in preoperative decision-making and determine whether to dissect the intestinal lymph nodes near the cancer area. [0003] At present, for the task of lymph node metastasis prediction, the technology of colorectal cancer lymph node metastasis prediction can be used. Mainly based on the bio...

AI Technical Summary

Problems solved by technology

[0004] However, there are many uncontrollable factors in the prediction technology of lymph node metastasis of colorectal cancer mentioned above. The kit includes the experiment of extracting and separating exosomes, the experiment of reverse transcription and the experiment of fluorescent quantitative polymerase chain reaction, and this A series of experiments will vary depending on the hospital operator, resulting in large experimental and operational errors in the results

Images