Lea Herschbach

ORCID: 0009-0005-6378-5073
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Artificial Intelligence in Healthcare and Education
  • Clinical Reasoning and Diagnostic Skills
  • Innovations in Medical Education
  • Machine Learning in Healthcare
  • Social Media in Health Education
  • Patient-Provider Communication in Healthcare
  • Hormonal Regulation and Hypertension
  • Explainable Artificial Intelligence (XAI)

University of Tübingen
2023-2024

Background Large language models such as GPT-4 (Generative Pre-trained Transformer 4) are being increasingly used in medicine and medical education. However, these prone to “hallucinations” (ie, outputs that seem convincing while factually incorrect). It is currently unknown how errors by large relate the different cognitive levels defined Bloom’s taxonomy. Objective This study aims explore performs terms of taxonomy using psychosomatic exam questions. Methods We a data set multiple-choice...

10.2196/52113 article EN cc-by Journal of Medical Internet Research 2024-01-23

Communication is a core competency of medical professionals and utmost importance for patient safety. Although curricula emphasize communication training, traditional formats, such as real or simulated interactions, can present psychological stress are limited in repetition. The recent emergence large language models (LLMs), generative pretrained transformer (GPT), offers an opportunity to overcome these restrictions.

10.2196/53961 article EN cc-by JMIR Medical Education 2024-01-16

Abstract Introduction Large language models (LLMs) such as GPT-4 are increasingly used in medicine and medical education. However, these prone to “hallucinations” – outputs that sound convincing while being factually incorrect. It is currently unknown how errors by LLMs relate the different cognitive levels defined Bloom’s Taxonomy. Methods We a large dataset of psychosomatic multiple-choice questions (MCQ) (N = 307) with real-world results derived from school exams. answered MCQs using two...

10.1101/2023.08.18.23294159 preprint EN medRxiv (Cold Spring Harbor Laboratory) 2023-08-21

<sec> <title>BACKGROUND</title> Communication is a core competency of medical professionals and utmost importance for patient safety. Although curricula emphasize communication training, traditional formats, such as real or simulated interactions, can present psychological stress are limited in repetition. The recent emergence large language models (LLMs), generative pretrained transformer (GPT), offers an opportunity to overcome these restrictions </sec> <title>OBJECTIVE</title> aim this...

10.2196/preprints.53961 preprint EN 2023-10-25

<sec> <title>BACKGROUND</title> Large language models such as GPT-4 (Generative Pre-trained Transformer 4) are being increasingly used in medicine and medical education. However, these prone to “hallucinations” (ie, outputs that seem convincing while factually incorrect). It is currently unknown how errors by large relate the different cognitive levels defined Bloom’s taxonomy. </sec> <title>OBJECTIVE</title> This study aims explore performs terms of taxonomy using psychosomatic exam...

10.2196/preprints.52113 preprint EN 2023-08-23
Coming Soon ...