Clinicopathologic features and machine learning tools to inform clinical outcome in lung cancer
Principal Investigator
Name
Arkadiusz Gertych
Degrees
Ph.D
Institution
Cedars-Sinai Medical Center
Position Title
Assistant Professor
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-456
Initial CDAS Request Approval
Dec 6, 2018
Title
Clinicopathologic features and machine learning tools to inform clinical outcome in lung cancer
Summary
Optimization of cancer treatment requires reliable assessment of multiple parameters derived from the tumor, the tumor microenvironment, and host response. Image analysis can increase the understanding of biological complexity and tumor heterogeneity by extracting and quantitating data from pathology slides that are exceedingly difficult and time consuming to obtain manually by pathologists. When combined with machine learning tools, data obtained by image analysis can potentially result in the identification of new prognostic biomarkers, treatment targets, and/or predictors of treatment response.
Aims
The project will use NLST digitized images of histologic slides and neural network models to identify and quantitate a variety of cytologic and histologic features of lung cancers, tumor microenvironment, and host reaction variables. Results will be correlated with clinical outcome. It is anticipated that features identified by applying machine learning tools and image analysis will outperform those currently used in clinical practice to help improve the diagnostic stratification of lung cancers and the prediction of clinical outcome prediction, and potentially help guide selection of treatment options.
Collaborators
Ann Walts MD