Development and validation of a report generation model for the low-dose chest CT scans

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
NLST (Learn more about this study)

Project ID
NLST-1081

Initial CDAS Request Approval
Jun 2, 2023

Title
Development and validation of a report generation model for the low-dose chest CT scans

Summary
Recent advancement of deep learning in the field of language models showed potential for the automatic image captioning and reporting. It is deemed feasible using the image encoder and language models to develop and validate a report generation model for the low-dose chest CT scans. The model-generated reports can be used as preliminary report templates for radiologists and as useful preliminary reports for clinicians. Reporting lung nodule is one of the primary targets of our project. Thus, in this study, we aim to develop and validate an automatic report generation model for the low-dose chest CT scans.

Aims

1. This project aims to develop and validate an automatic report generation model for the low-dose chest CT scans.
2. Target findings include lung nodules and diffuse parenchymal lung disease.
3. The study outcome is the rate of automatic CT reporting, assessed by radiologists.

Collaborators

Jong Hyuk Lee, Seoul National University Hospital
Seungho Lee, Seoul National University Hospital