Identification of colorectal cancer using structured and free text clinical data

Authors

Douglas F. Redd, 19986Washington DC VA Medical Center, Washington, DC, USA Biomedical Informatics Center, 43989The George Washington University School of Medicine and Health Sciences, Washington, DC, USA.
Yijun Shao, 19986Washington DC VA Medical Center, Washington, DC, USA Biomedical Informatics Center, 43989The George Washington University School of Medicine and Health Sciences, Washington, DC, USA.
Qing Zeng-Treitler, 19986Washington DC VA Medical Center, Washington, DC, USA Biomedical Informatics Center, 43989The George Washington University School of Medicine and Health Sciences, Washington, DC, USA.
Laura J. Myers, 20015Richard L Roudebush VA Medical Center, Indianapolis, IN, USAIndiana University School of Medicine, Indianapolis, IN, USA Regenstrief Institute Inc, Indianapolis, IN, USA.
Barry C. Barker, 20015Richard L Roudebush VA Medical Center, Indianapolis, IN, USA.
Stuart J. Nelson, Biomedical Informatics Center, 43989The George Washington University School of Medicine and Health Sciences, Washington, DC, USA.
Thomas F. Imperiale, 20015Richard L Roudebush VA Medical Center, Indianapolis, IN, USA Indiana University School of Medicine, Indianapolis, IN, USA Regenstrief Institute Inc, Indianapolis, IN, USA.

Document Type

Journal Article

Publication Date

10-1-2022

Journal

Health informatics journal

Volume

28

Issue

4

DOI

10.1177/14604582221134406

Keywords

Colon cancer; feature utilization; machine learning; model comparison; statistical models

Abstract

Colorectal cancer incidence has continually fallen among those 50 years old and over. However, the incidence has increased in those under 50. Even with the recent screening guidelines recommending that screening begins at age 45, nearly half of all early-onset colorectal cancer will be missed. Methods are needed to identify high-risk individuals in this age group for targeted screening. Colorectal cancer studies, as with other clinical studies, have required labor intensive chart review for the identification of those affected and risk factors. Natural language processing and machine learning can be used to automate the process and enable the screening of large numbers of patients. This study developed and compared four machine learning and statistical models: logistic regression, support vector machine, random forest, and deep neural network, in their performance in classifying colorectal cancer patients. Excellent classification performance is achieved with AUCs over 97%.

APA Citation

Redd, Douglas F.; Shao, Yijun; Zeng-Treitler, Qing; Myers, Laura J.; Barker, Barry C.; Nelson, Stuart J.; and Imperiale, Thomas F., "Identification of colorectal cancer using structured and free text clinical data" (2022). GW Authored Works. Paper 1839.
https://hsrc.himmelfarb.gwu.edu/gwhpubs/1839

Department

Clinical Research and Leadership