Evaluate Medical Students' History Gathering and Physical Examination Skills Using Natural Language Processing Methods (Poster 51)

• In Event: AERA Poster Session Nine
  In Poster Session: Division I Poster Session

Abstract

Medical school students are required to write patient notes to demonstrate their ability to gather information, perform physical examinations, and summarize their findings. The purpose of this study is to employ Natural Language Processing (NLP) methods to evaluate patient notes written by medical school students on history gathering and physical examination skills. We collected 104 PN from third-year medical students during Internal Medicine clerkship in our medical school with chief complaints on chest pain. The proposed NLP algorithm achieve around 80% accuracy compared with human annotators' performance. The intrinsic and extrinsic evaluations validate the performance of the NLP methods and point to the potential use of it in aiding medical short answer assessment.

Authors

• Ling Wang, Michigan State University
  Presenting Author

• Heather Laird-Fick, Michigan State University
  Non-Presenting Author

• Mohammad Saad Salam, Michigan State University
  Non-Presenting Author

• Maham Khan, Michigan State University
  Non-Presenting Author

• Adolfo Martinez Salazar
  Non-Presenting Author

• Yifei Li, Michigan State University
  Non-Presenting Author

• Zongqiang Liao, Michigan State University
  Non-Presenting Author

• Churlsun Han
  Non-Presenting Author

• Sumugdha Rayamajhi, Michigan State University
  Non-Presenting Author

• Parisa Kordjamshidi
  Non-Presenting Author