Nodule risk assessment
Principal Investigator
Name
Scott Morris
Degrees
P.S.M., M.B.A., Ph.D.
Institution
Maverix Medical
Position Title
Chief Science Officer
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1252
Initial CDAS Request Approval
May 16, 2024
Title
Nodule risk assessment
Summary
We are building a model that combines clinical, radiological, and blood-based biomarkers to assess cancer risk in lung nodules. Current lung cancer risk models such as Mayo Solitary Pulmonary Nodule risk or Brock University risk are designed to estimate risk based on the interpretation of a single CT scan. It is known that lung cancers are more likely to grow relatively quickly and benign conditions are more likely to be stable in size. We will integrate multiple CT scan results and examine features known to predict cancer risk (e.g., nodule size, spiculation, upper lobe, solid/semi-solid/etc, smoking history, and age) along with changes in CT results over time (e.g., rate of growth) to predict risk. This clinical risk score will then be combined with blood-based biomarker data from another cohort to develop a better risk score for lung nodules.
Aims
We aim to build a pulmonary risk calculator that adjusts risk based on changes in CT scan results over time. Once this aim is complete, we will use this calculator along with blood-based biomarker results to classify a training set of separate cohorts of patients. Iterative adjustments will be made to build an accurate classifier using the training set. The classifier will then be applied to a separate validation set containing clinical, radiological and blood biomarker data to determine its accuracy.
Collaborators
None