Research progresses of pathological type prediction of early lung adenocarcinoma based on CT images
Lung adenocarcinoma is the most common type of non-small cell lung cancer.Most patients with lung adenocarcinoma have unobvious clinical manifestations in the early stage of cancer,they are already at an advanced stage when they are discovered,and the prognosis of advanced lung adenocarcinoma patients is extremely unsatisfactory.Therefore,early detection,early diagnosis,and early treatment are the most effective measures to improve the survival rate of patients.Early lung adenocarcinoma can be divided into minimally invasive adenocarcinoma and invasive adenocarcinoma according to the degree of infiltration of cancer cells into surrounding tissues.Meanwhile,glandular prodromal lesions also need long-term observation and follow-up to prevent further development and deterioration.The five-year disease-free survival rate of early lung adenocarcinoma varies greatly with different pathological types.Correctly predicting the type of early lung adenocarcinoma can help radiologists make better treatment plans and further improve the prognosis of patients.Early lung adenocarcinoma is closely related to ground-glass nodule(GGN).CT has become the most important imaging method to observe GGN with the advantages of non-invasive and high resolution.Existing studies on the prediction of pathological types of early lung adenocarcinoma mainly focus on artificial intelligence technology:traditional radiomics builds classification models based on high-throughput extraction and screening of GGN quantitative features;the deep learning method automatically extracts the deep features of GGN and learns the implicit relationship between GGN and category to complete the category prediction task.At present,researchers at home and abroad have published a lot of literature on the prediction of pathological tissue types of early lung adenocarcinoma based on GGN CT images,radiomics and deep learning models.This article mainly reviews the research progress of GGN CT imaging features,radiomics features,and deep learning methods in predicting the histological types of early lung adenocarcinoma,in order to provide a valuable reference for related researchers.
万方数据
维普
