Skip to Main Content
An official website of the United States government

Machine Learning with Low Dose CT Data

Principal Investigator

Name
Bennett Landman

Degrees
PhD

Institution
Vanderbilt University

Position Title
Professor

Email
bennett.landman@vanderbilt.edu

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-993

Initial CDAS Request Approval
Dec 8, 2022

Title
Machine Learning with Low Dose CT Data

Summary
We seek to develop method to integrate assessment of body habitus with lung cancer risk. We are developing both cross-sectional and longitudinal learning frameworks. We are targeting quantitative anatomical modeling (e.g., semantic segmentation), lung nodule identification / classification, and risk stratification. We are developing supporting image processing techniques to harmonize imaging and improve inter-site reproducibility.

Aims

* Segment anatomy within the chest CT field of view
* Identify potentially cancerous lung nodules
* Assess likelihood of cancer within specific nodules and a time point in a holistic manner.
* Perform image processing to improve inter-site consistency and enhance performance of the above objectives.

Collaborators

Fabien Maldonado, VUMC
Kim Lori Sandler, VUMC
Tom Lasko, VUMC

Related Publications