Development of a personalized lung cancer risk prediction model

Principal Investigator

Name
Pinchas Akiva

Degrees
PhD

Institution
Medial-Research

Position Title
Bioinformatics Senior Scientist

Email
pini@medial-research.com

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-86

Initial CDAS Request Approval
Aug 26, 2014

Title
Development of a personalized lung cancer risk prediction model

Summary
Reducing the number of individuals needed to screen in order to prevent lung cancer mortality is a challenge. The purpose of this study is to develop and validate a lung cancer risk prediction model for identifying individuals at high risk of having lung cancer. The model should aid optimizing the selection criteria of individuals to be screened for lung cancer, helping to increase the positive predictive value of lung cancer screening.

In the development process, we will utilize machine learning based algorithmic tools (such as decision tree algorithms) that analyze the characteristics of lung cancer patients, and produce a model that enables defining the probability of a specific individual to harbor lung cancer. The model will take into account smoking behavior as well as other data types such as demographics and disease history.

To increase the validity of the results, the study datasets will include the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO) and the National Lung Screening Trial (NLST) datasets.

Aims

Develop a model that generates an individual risk score to enable identifying individuals at high risk for having lung cancer.