Deep learning-based head and neck squamous cell carcinoma lymph node metastatic carcinoma diagnosis auxiliary identification system

Abstract

A deep learning-based head and neck squamous cell carcinoma lymph node metastatic carcinoma diagnosis auxiliary identification system comprises an image scanning import subsystem, an image identification and evaluation subsystem, a diagnosis auxiliary model autonomous correction subsystem and a result output module. The image identification and evaluation subsystem comprises a tissue region identification model and a metastatic cancer region discrimination model; the tissue region identification model is used for removing a background region of an image by using an image segmentation algorithm to obtain a tissue region; and the metastatic cancer area discrimination model is used for performing segmentation diagnosis on a tissue area by using a secondary diagnosis model constructed based on a deep learning algorithm, and performing labeled diagnosis on the overall potential cancer focus area of a lymph node in a result output module. According to the invention, a secondary diagnosis model is constructed through the deep learning algorithm to identify a lymph node full slice, the possible cancer focus area is screened out, the film reading time of pathologists is reduced, the diagnosis efficiency is improved, and the diagnosis model also has autonomous correction and updating capabilities.

Description

Technical field [0001] The present invention relates to a computer analytical technology of a medical image, and more particularly to depth learning-based head-neck squamous cancer lymph node metastasis cancer diagnostic auxiliary identification systems. Background technique [0002] Head-necked squamous cell is the seventh largest malignant tumor of the world, which brings great harm to the health and life of patients; regional neck lymph node metastasis is an important adverse pre-factor of head-necked squamous cell carcinoma; lymph nodes precise diagnosis is head neck scale The composition of the cancer stage system provides a basis for clinical decision making and prompt prognosis. [0003] At present, pathological examination is a gold standard for diagnosing head-necked squamous cancer lymph nodes transfer, mainly through pathophists observing the morphological characteristics of the tissue specimens under microscope. However, artificial pathological evaluation lymph node i...

AI Technical Summary

Problems solved by technology

However, the identification of macrophages, germinal centers, stroma and other regions is still tricky in the existing artificial intelligence diagnosis research of lymph node metastases

At present, the artificial intelligence diagnosis of lymph node metastases in head and neck squamous cell carcinoma is still blank. It is very urgent and meaningful to develop an auxiliary identification system for pathological diagnosis of lymph node metastases in head and neck squamous cell carcinoma.

Images