Deep Learning Pulmonary Nodule Classification

Principal Investigator

Name
Yitzi Pfeffer

Degrees
B.Sc

Institution
IMedis-AI LTD.

Position Title
CTO

Email
yitzi@imedis.ai

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-832

Initial CDAS Request Approval
Sep 14, 2021

Title
Deep Learning Pulmonary Nodule Classification

Summary
The management of indeterminate pulmonary nodules (IPNs) is extremely challenging, resulting in unneeded invasive procedures and delays in diagnosis and treatment. Strategies to decrease the rate of unnecessary invasive procedures have yet to be developed.
A lung cancer prediction AI-based model outputting a malignancy score for each detected IPN, can help solve this challenge and improve the management of IPNs.

Aims

1. We propose to develop a lung cancer prediction Convolutional Neural Network model that will be trained using CT images of IPNs from NLST.
2. We will show that the algorithm is robust across many datasets.
3. We will compare the results to existing models and show superiority.

Collaborators

NA