Predicting Tumor Invasiveness of Subcentimeter Pulmonary Adenocarcinomas from CT Scan with 3D Convolutional Neural Networks
Principal Investigator
Name
Ming Li
Degrees
Ph.D
Institution
Huadong hospital
Position Title
professor
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-403
Initial CDAS Request Approval
Apr 20, 2018
Title
Predicting Tumor Invasiveness of Subcentimeter Pulmonary Adenocarcinomas from CT Scan with 3D Convolutional Neural Networks
Summary
The system processes a 3D patch of raw CT, and learns a deep representation from a given nodule. A dataset of more than 500 nodules was used in this paper. We train and validate our deep learning system on 4/5 nodules, and test the performance on 1/5 nodules. An independent public dataset is designed to refine our model.
Aims
In this project, we aimed to develop a deep learning system based on 3D convolutional neural networks, which automatically predicts the tumor invasiveness and to build a deep learning model which could help doctors working efficiently and facilitate the precision medicine.
Collaborators
Jiancheng Yang Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai 200240 P.R. China, Diannei Technology, Shanghai 200050 P.R. China
Peijun Wang Department of Radiology, Tongji Hospital, School of Medicine, Tongji University, Shanghai 200065, China