Federated Validation of Lung Nodule Detection Algorithms.

Principal Investigator

Name
Brian Ayers

Degrees
M.D., M.B.A.

Institution
Gesund AI

Position Title
Clinical Advisor

Email
brian@gesund.ai

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-869

Initial CDAS Request Approval
Jan 18, 2022

Title
Federated Validation of Lung Nodule Detection Algorithms.

Summary
There have been numerous AI lung nodule detection algorithms developed over the last 5 years. Many demonstrate promising results in their index publications. However, there is not a standardized approach to the validation process for clinical grade AI models, and many models are being approved by the FDA without multi-institutional studies. Gesund AI is developing a platform that allows for the robust and secure testing of AI models on multiple datasets. The NLST dataset will be used as part of the validation process for lung nodule detection algorithms that were not previously trained on this data. The NLST data will remain on the Gesund platform and not available for individual access or download by any other entities.

Aims

1. Use the NLST data for validation of previously developed lung nodule detection models

Collaborators

N/A