Image quality assessment of digitized chest X ray images and feature characterization of "Chest X-ray Age"
Principal Investigator
Name
Tomoki Nakamizo
Degrees
Ph.D.
Institution
Radiation Effects Research Foundation
Position Title
Senior Research Scientist
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1071
Initial CDAS Request Approval
May 30, 2023
Title
Image quality assessment of digitized chest X ray images and feature characterization of "Chest X-ray Age"
Summary
In this project, we will evaluate the quality of chest X-ray images stored at our institute for their suitability in deep learning models like CXR-Age(JACC Img. 2021; 14:2226). For this purpose, We will compare the image quality metrics of our images with those used in the original study to determine their eligibility thresholds.
In addition, we will try to identify and characterize the features contributing to "chest X-ray age" by Raghu et. al. by extracting features from the penultimate layer. We will localize these features on the images and assess their association with disease outcomes through survival analysis.
Once we achieve these objectives, we will proceed to investigate our images further.
In addition, we will try to identify and characterize the features contributing to "chest X-ray age" by Raghu et. al. by extracting features from the penultimate layer. We will localize these features on the images and assess their association with disease outcomes through survival analysis.
Once we achieve these objectives, we will proceed to investigate our images further.
Aims
1. To evaluate the quality of the chest X-ray images stored at our institute by comparing them with those used in previous deep learning studies.
2. To identify and characterize the features contributing to "chest X-ray age" by Raghu et. al.
Collaborators
Zhenqiu Liu, Radiation Effects Research Foundation
Satoru Ono, Radiation Effects Research Foundation