Skip to Main Content
An official website of the United States government
Login Change: CDAS is now using the NIH Researcher Auth Service (RAS) to manage our existing login options. When logging into CDAS, you will be taken to RAS, where you can select your preferred login provider to access CDAS.

Diagnosing COVID-19 from deep learning trained on CT scans

Principal Investigator

Name
Jean-Remi King

Degrees
Ph.D.

Institution
CNRS

Position Title
Tenured Researcher (CRCN)

Email
jeanremi.king@gmail.com

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-650

Initial CDAS Request Approval
Mar 17, 2020

Title
Diagnosing COVID-19 from deep learning trained on CT scans

Summary
Diagnosing COVID19 infection from other mild respiratory diseases is a major priority to limit the current pandemic. Yet, the number of kit tests availble is dramatically low, and MDs are currently relying on CT scans as a substitute. The present projects aims to build a deep learning architecture trained to model CT scans. This deep learning architecture will then transform the incoming CT scans from infected and non-infected patients currently acquired in the clinics, in order to facilitate their discriminability.

Aims

- develop open source preprocessing pipeline of CT scans
- implement U-net deep learning architecture to model NLST data
- provide user-friendly interface to project new CT scan onto the model, and allow supervised or unsupervised analysis

Collaborators

Nathan Pfeiffer Smadja