Construction and evaluation of an intelligent diagnosis model for lung cancer based on multi-round Chest X-ray images
Principal Investigator
Name
Qianyun Jin
Degrees
M.D.
Institution
Tianjin medical University Cancer Institute and Hospital
Position Title
Student
Email
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCOI-1627
Initial CDAS Request Approval
Jul 22, 2024
Title
Construction and evaluation of an intelligent diagnosis model for lung cancer based on multi-round Chest X-ray images
Summary
Chest X-ray (CXR) is the main method of lung cancer screening in grass-roots community of China. Traditional X-ray interpretation is time-consuming, laborious and subjective. Previous studies have shown that the decision effect of intelligent diagnostic model of single-round chest radiography is better than that of single-round doctors. A single round of chest X-rays does not rule out the risk of lung cancer, and 19% -26% of lung cancers visible on chest x-rays are missed at the first reading ; Multiple rounds of chest X-ray screening are needed to find potential high-risk population of lung cancer so as to further improve the screening effect of lung cancer. This study will further explore the construction of multiple rounds of chest X-ray intelligent diagnosis model and its effect judgment. Furthermore, although environmental risk factors, such as smoking, contribute most to lung cancer, genetic variants can explain 12-21% of the heritability of lung cancer. It has been shown that polygenic risk scores (PRS) derived from genome-wide association studies (GWAS) can be effectively used to distinguish high-risk populations of lung cancer for accurate prevention in the population. There are no studies to explore the association between genetic and multi-round chest radiograph progression, and whether combining multi-round chest radiograph progression judged by intelligent diagnostic model with genetic can further improve the effect of lung cancer screening.
Aims
1) Based on the model of single-round chest radiography for intelligent diagnosis of lung nodules, the model of multi-round chest radiography for intelligent diagnosis of lung cancer will be constructed and verified.
2) Based on the PLCO database, the effect of genetic risk on radiographic progression judged by the intelligent diagnostic model, and the relationship between smoking and genetic, radiographic progression will be evaluated.
Collaborators
None