Predicting 5-year risk of lung cancer using LDCT images
Principal Investigator
Name
Shu Jiang
Degrees
PhD
Institution
Washington University School of Medicine
Position Title
Associate professor
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1293
Initial CDAS Request Approval
Jul 15, 2024
Title
Predicting 5-year risk of lung cancer using LDCT images
Summary
We aim to develop and internally validate a statistical model using LDCTs from the National Lung Screening Trial (NLST). We hypothesize that LDCT images contain information that is predictive of up to 5-years of lung cancer risk in addition to lung nodules. The algorithm will be assessed with area under the ROC curve (AUC) as well as calibration to national absolute risk for translation. The developed algorithm that can accurately predict future lung cancer risk over the next 5-year could maximize benefit and minimize harms.
Aims
Our specific aims are as follows:
Aim 1: Develop a statistical model using the LDCT slices to predict 5-year risk of lung cancer
Aim 2: Internally validate the model performance with an independent hold out set in NLST
Aim 3: Compare with existing risk prediction models using risk factors only
Collaborators
Dr. Graham Colditz