event-icon
Description

Abstract: This study tested the feasibility of a natural language processing (NLP) approach for detecting mental stressors from clinical notes. A total of 197,346 clinical notes for 3,138 eligible prostate cancer patients were used to iteratively develop a lexicon of mental stressors, customized negations, and NLP algorithms. With negations, 1,917 (61.1%) patients were detected having stressor mention(s). Mental stressors are commonly documented in clinical narratives for patients with prostate cancer and could be detected by NLP.

Learning Objective 1: Extracting mental stressors from clinical notes using NLP

Authors:

Vivienne Zhu (Presenter)
Medical University of South Carolina

Chanita Halbert, Medical University of South Carolina
Melanie Jefferson, Medical University of South Carolina
Richard Wolfe, Medical University of South Carolina
Leslie Lenert, Medical University of South Carolina

Presentation Materials:

Keywords