Prediction of Target-Drug Therapy by Identifying Gene Mutations in Lung Cancer With Histopathological Stained Image and Deep Learning Techniques

• DOI: 10.3389/fonc.2021.642945
• Publication Date: 2021-04-13T06:33:24Z
• Authors: Kaimei Huang; Zhiyi Mo; Wen Zhu; Bo Liao; Yachao Yang; Fang-Xiang Wu
• Keywords: convolution neural network; lung cancer; 03 medical and health sciences; 0302 clinical medicine; residual network; Oncology; Neoplasms. Tumors. Oncology. Including cancer and carcinogens; targeted therapy; pathological images; RC254-282; 3. Good health

Lung cancer is a kind of with high morbidity and mortality which associated various gene mutations. Individualized targeted-drug therapy has become the optimized treatment lung cancer, especially benefit for patients who are not qualified lobectomy. It crucial to accurately identify mutant genes within tumor region from stained pathological slice. Therefore, we mainly focus on identifying by analyzing images. In this study, have proposed method mutations in histopathological image deep learning predict target-drug therapy, referred as DeepIMLH. The DeepIMLH algorithm first downloaded 180 hematoxylin-eosin staining (H&E) images Cancer Gene Atlas (TCGA). Then convolution Gaussian mixture model (DCGMM) was used perform color normalization. Convolutional neural network (CNN) residual (Res-Net) were mutated H&E imaging achieved good accuracy. demonstrated that our can be choose might applied clinical practice. More studies should conducted though.