Large language models for dermatological image interpretation – a comparative study
LLM performance varied significantly, emphasizing the need for cautious usage. Notably, a free, locally hostable model correctly identified the top diagnosis for approximately two-thirds of all images, demonstrating the potential for safer, locally deployed LLMs. Advancements in model accuracy and the integration of clinical metadata could further enhance accessible and reliable clinical decision support systems.
Authors
- Sebastian Kühn (ORCID: https://orcid.org/0000-0002-8031-2973)
- Martin Krusche (ORCID: https://orcid.org/0000-0002-0582-7790)
- Michael Hertl (ORCID: https://orcid.org/0000-0002-5494-2461)
- Martin C. Hirsch (ORCID: https://orcid.org/0000-0003-3268-4921)
- Johannes Knitza (ORCID: https://orcid.org/0000-0001-9695-0657)
- Fabian Lechner (ORCID: https://orcid.org/0009-0003-9200-477X)
- Lasse Cirkel
- Lukas Alexander Henk
Institutions
- Universität Hamburg (DE)
- Universitätsklinikum Gießen und Marburg (DE)
- University Medical Center Hamburg-Eppendorf (DE)
- Université Grenoble Alpes (FR)
Publication Details
- Journal
- Diagnosis
- Published
- 2025-05-22
- DOI
- https://doi.org/10.1515/dx-2025-0014
- Citations
- 4
- Primary Topic
- Cutaneous Melanoma Detection and Management
- Type
- article
- Field-Weighted Citation Impact
- 4.18