Sentiment analysis of medical notes for lung cancer patients at the Department of Veterans Affairs

Conference Year

January 2022

Abstract

Natural language processing of medical records offers tremendous potential to improve the patient experience. Sentiment analysis of clinical notes has been performed with mixed results, often highlighting the issue that dictionary ratings are not domain specific. Here, for the first time, we re-calibrate the labMT sentiment dictionary on 3.5M clinical notes describing 10,000 patients diagnosed with lung cancer. The found sentiment score of notes was evaluated against platelet counts and treatment arms. We found that the oncology specific labMT dictionary produces a promising signal in notes that can be detected based on a comparative analysis to the aforementioned parameters.

Primary Faculty Mentor Name

Chris Danforth

Secondary Mentor Name

Peter Dodds

Faculty/Staff Collaborators

UVM: Peter Dodds, Chris Danforth, Robert Gramling, VA (non-UVM): Jennifer La, Mary Brophy, Nhan Do, Nathanael Fillmore

Student Collaborators

Josh Minot

Status

Graduate

Student College

College of Engineering and Mathematical Sciences

Program/Major

Complex Systems

Primary Research Category

Health Sciences

Secondary Research Category

Engineering & Physical Sciences

Abstract only.

Share

COinS
 

Sentiment analysis of medical notes for lung cancer patients at the Department of Veterans Affairs

Natural language processing of medical records offers tremendous potential to improve the patient experience. Sentiment analysis of clinical notes has been performed with mixed results, often highlighting the issue that dictionary ratings are not domain specific. Here, for the first time, we re-calibrate the labMT sentiment dictionary on 3.5M clinical notes describing 10,000 patients diagnosed with lung cancer. The found sentiment score of notes was evaluated against platelet counts and treatment arms. We found that the oncology specific labMT dictionary produces a promising signal in notes that can be detected based on a comparative analysis to the aforementioned parameters.