External validation of a chest radiograph-based prognostic model

Principal Investigator

Name
Hyungjin Kim

Degrees
M.D., Ph.D.

Institution
Seoul National University Hospital

Position Title
Clinical associate professor

Email
khj.snuh@gmail.com

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCOI-1208

Initial CDAS Request Approval
Apr 28, 2023

Title
External validation of a chest radiograph-based prognostic model

Summary
Chest radiographs contain diagnostic and prognostic information for various thoracic and cardiovascular diseases. The current practice is limited to the qualitative analyses of radiographs by radiologists. It is reasonable to extract quantitative data from chest radiographs automatically by using a deep learning model, which may be representative of thoracic and cardiovascular abnormalities. Such information can be used for the prognostication and disease prevention. Our model, which was developed outside the PLCO dataset, is able to capture prognostic imaging signatures from various anatomical systems (e.g., lungs, heart, and vessels). We aims to externally validate our model using the PLCO dataset.

Aims

The study purpose is to externally validate a chest radiograph-based, deep learning prognostic model, which can capture prognsotic imaging signatures. The study outcomes include cardiovascular mortality and all-cause mortality.

Collaborators

Jong Hyuk Lee, Seoul National University Hospital
Soon Ho Yoon, Seoul National University Hospital