Accessibility settings

Published on in Vol 14 (2026)

This is a member publication of McMaster University

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/81644, first published .
Doctor and nurse collaborating on patient care, using computer and clipboard.

Understanding Transformer-Based Classifications of Medical Text Using a Large Language Model for the Attribution of Feature Importance: Proof-of-Concept Algorithm Development and Validation Study

Understanding Transformer-Based Classifications of Medical Text Using a Large Language Model for the Attribution of Feature Importance: Proof-of-Concept Algorithm Development and Validation Study

Fangwen Zhou   1 , MSc ;   Ashirbani Saha   2 , PhD ;   Muhammad Afzal   3 , PhD ;   Rick Parrish   1 ;   R Brian Haynes   1 , MD, PhD ;   Alfonso Iorio   1, 4 , MD, PhD ;   Cynthia Lokker   1 , PhD

1 Health Information Research Unit, Department of Health Research Methods, Evidence, and Impact, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada

2 Department of Oncology, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada

3 Department of Computer Science, Birmingham City University, Birmingham, United Kingdom

4 Department of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada

Corresponding Author:

  • Cynthia Lokker, PhD
  • Health Information Research Unit, Department of Health Research Methods, Evidence, and Impact
  • Faculty of Health Sciences, McMaster University
  • 1280 Main Street West
  • Hamilton, ON L8S 4L8
  • Canada
  • Phone: 1 905-525-9140 ext 22208
  • Email: lokkerc@mcmaster.ca