Predicting Lung Cancer Treatment Using an End to End 3D Serial Imaging CNN Model

Principal Investigator

Name
Amy Li

Degrees
B.S.E.

Institution
Princeton University

Position Title
Student

Email
al25@princeton.edu

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-618

Initial CDAS Request Approval
Dec 11, 2019

Title
Predicting Lung Cancer Treatment Using an End to End 3D Serial Imaging CNN Model

Summary
This project attempts to evaluate 3D computed-tomography (CT) radiomic features for their capability to predict four main cancer endpoints: 1) distant metastasis (DM) 2) local recurrence (LR) 3) progression and 4) overall survival via the use of a time series 3D convoluted neural network in non-small cell lung cancer.

Aims

- identify a list of significant radiomic features that predict lung cancer treatment response
- build a functioning CNN model with a significant accuracy in solving the project problem

Collaborators

Ludovic Tangpi, Advisor, Princeton University