Deep Learning for Analysis of Pulmonary Nodules at Low-Dose CT
Principal Investigator
Name
Yajun Wu
Degrees
M.D.
Institution
Shenzhen Yorktal Digital Medical Imaging Technologies Company Ltd， China
Position Title
Manager
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1005
Initial CDAS Request Approval
Jan 10, 2023
Title
Deep Learning for Analysis of Pulmonary Nodules at Low-Dose CT
Summary
In order to validate a deep learning model which is trained using LIDC-IDRI dataset for pulmonry nodules malignancy/benign classification and malignancy risk estimation, we need a stand alone dataset to test the model's inference metrics, such as Area Under ROC, Sensitivity, Specifity, F1 Score, and etc. Besides, another deep learning model is aimed to automatically classify pulmonary nodules as Solid, Part-solid, Non-solid. We believe NLST dataset can also provide more stand alone data to validate this model.
Aims
1. Pulmonry nodules malignancy/benign classification and malignancy risk estimation using deep learning;
2. Pulmonry nodules classification as Solid, Part-solid, Non-solid types using deep learning.
Collaborators
This project will be done within our instituion only, without other collaborators.