{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T23:56:41Z","timestamp":1792799801280,"version":"4.5.0"},"reference-count":19,"publisher":"Walter de Gruyter GmbH","issue":"1","license":[{"start":{"date-parts":[[2025,5,23]],"date-time":"2025-05-23T00:00:00Z","timestamp":1747958400000},"content-version":"unspecified","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,2,27]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec id=\"j_dx-2025-0014_abs_001\">\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>Interpreting skin findings can be challenging for both laypersons and clinicians. Large language models (LLMs) offer accessible decision support, yet their diagnostic capabilities for dermatological images remain underexplored. This study evaluated the diagnostic performance of LLMs based on image interpretation of common dermatological diseases.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"j_dx-2025-0014_abs_002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>A total of 500 dermatological images, encompassing four prevalent skin conditions (psoriasis, vitiligo, erysipelas and rosacea), were used to compare seven multimodal LLMs (GPT-4o, GPT-4o mini, Gemini 1.5 Pro, Gemini 1.5 Flash, Claude 3.5 Sonnet, Llama3.2 90B and 11B). A standardized prompt was used to generate one top diagnosis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"j_dx-2025-0014_abs_003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The highest overall accuracy was achieved by GPT-4o (67.8\u202f%), followed by GPT-4o mini (63.8\u202f%) and Llama3.2 11B (61.4\u202f%). Accuracy varied considerably across conditions, with psoriasis with the highest mean LLM accuracy of 59.2\u202f% and erysipelas demonstrating the lowest accuracy (33.4\u202f%). 11.0\u202f% of all images were misdiagnosed by\u00a0all LLMs, whereas 11.6\u202f% were correctly diagnosed by all\u00a0models. Correct diagnoses by all LLMs were linked to clear, disease-specific features, such as sharply demarcated erythematous plaques in psoriasis. Llama3.2 90B was the only LLM to decline diagnosing images, particularly those involving intimate areas of the\u00a0body.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"j_dx-2025-0014_abs_004\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>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.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1515\/dx-2025-0014","type":"journal-article","created":{"date-parts":[[2025,5,22]],"date-time":"2025-05-22T15:58:19Z","timestamp":1747929499000},"page":"75-81","source":"Crossref","is-referenced-by-count":15,"title":["Large language models for dermatological image interpretation\u00a0\u2013\u00a0a comparative study"],"prefix":"10.1515","volume":"13","author":[{"given":"Lasse","family":"Cirkel","sequence":"first","affiliation":[{"name":"Institute of Artificial Intelligence, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"},{"name":"Institute for Digital Medicine, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fabian","family":"Lechner","sequence":"additional","affiliation":[{"name":"Institute of Artificial Intelligence, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"},{"name":"Institute for Digital Medicine, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lukas Alexander","family":"Henk","sequence":"additional","affiliation":[{"name":"Institute for Digital Medicine, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"},{"name":"Department of Dermatology and Allergology , University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin","family":"Krusche","sequence":"additional","affiliation":[{"name":"Division of Rheumatology and Systemic Inflammatory Diseases, III. Department of Medicine, University Medical Center Hamburg-Eppendorf , Hamburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin C.","family":"Hirsch","sequence":"additional","affiliation":[{"name":"Institute of Artificial Intelligence, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Hertl","sequence":"additional","affiliation":[{"name":"Department of Dermatology and Allergology , University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastian","family":"Kuhn","sequence":"additional","affiliation":[{"name":"Institute for Digital Medicine, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9695-0657","authenticated-orcid":false,"given":"Johannes","family":"Knitza","sequence":"additional","affiliation":[{"name":"Institute for Digital Medicine, University Hospital Gie\u00dfen-Marburg, Philipps University , Marburg , Germany"},{"name":"Universit\u00e9 Grenoble Alpes, AGEIS , Grenoble , France"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"374","published-online":{"date-parts":[[2025,5,23]]},"reference":[{"key":"2026030615414913548_j_dx-2025-0014_ref_001","doi-asserted-by":"crossref","unstructured":"Alowais, SA, Alghamdi, SS, Alsuhebany, N, Alqahtani, T, Alshaya, AI, Almohareb, SN, et al.. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Med Educ 2023;23:689. https:\/\/doi.org\/10.1186\/s12909-023-04698-z.","DOI":"10.1186\/s12909-023-04698-z"},{"key":"2026030615414913548_j_dx-2025-0014_ref_002","doi-asserted-by":"crossref","unstructured":"Aung, YYM, Wong, DCS, Ting, DSW. The promise of artificial intelligence: a review of the opportunities and challenges of artificial intelligence in healthcare. Br Med Bull 2021;139:4\u201315. https:\/\/doi.org\/10.1093\/bmb\/ldab016.","DOI":"10.1093\/bmb\/ldab016"},{"key":"2026030615414913548_j_dx-2025-0014_ref_003","doi-asserted-by":"crossref","unstructured":"Clusmann, J, Kolbinger, FR, Muti, HS, Carrero, ZI, Eckardt, JN, Laleh, NG, et al.. The future landscape of large language models in medicine. Commun Med 2023;3:141. https:\/\/doi.org\/10.1038\/s43856-023-00370-1.","DOI":"10.1038\/s43856-023-00370-1"},{"key":"2026030615414913548_j_dx-2025-0014_ref_004","unstructured":"Saab, K, Tu, T, Weng, WH, Tanno, R, Stutz, D, Wulczyn, E, et al.. Capabilities of gemini models in medicine. Available from: https:\/\/arxiv.org\/abs\/2404.18416."},{"key":"2026030615414913548_j_dx-2025-0014_ref_005","doi-asserted-by":"crossref","unstructured":"Strotzer, QD, Nieberle, F, Kupke, LS, Napodano, G, Muertz, AK, Meiler, S, et al.. Toward foundation models in radiology? Quantitative assessment of GPT-4V\u2019s multimodal and multianatomic region capabilities. Radiology 2024;313:e240955. https:\/\/doi.org\/10.1148\/radiol.240955.","DOI":"10.1148\/radiol.240955"},{"key":"2026030615414913548_j_dx-2025-0014_ref_006","doi-asserted-by":"crossref","unstructured":"Gui, H, Rezaei, SJ, Schlessinger, D, Weed, J, Lester, J, Wongvibulsin, S, et al.. Dermatologists\u2019 perspectives and usage of Large Language Models in practice: an exploratory survey. J\u00a0Invest Dermatol 2024;144:2298\u2013301. https:\/\/doi.org\/10.1016\/j.jid.2024.03.028.","DOI":"10.1016\/j.jid.2024.03.028"},{"key":"2026030615414913548_j_dx-2025-0014_ref_007","doi-asserted-by":"crossref","unstructured":"Brancaccio, G, Balato, A, Malvehy, J, Puig, S, Argenziano, G, Kittler, H. Artificial intelligence in skin cancer diagnosis: a reality check. J\u00a0Invest Dermatol 2024;144:492\u20139. https:\/\/doi.org\/10.1016\/j.jid.2023.10.004.","DOI":"10.1016\/j.jid.2023.10.004"},{"key":"2026030615414913548_j_dx-2025-0014_ref_008","doi-asserted-by":"crossref","unstructured":"Escal\u00e9-Besa, A, Y\u00e9lamos, O, Vidal-Alaball, J, Fuster-Casanovas, A, Mir\u00f3 Catalina, Q, B\u00f6rve, A, et al.. Exploring the potential of artificial intelligence in improving skin lesion diagnosis in primary care. Sci Rep 2023;13:4293. https:\/\/doi.org\/10.1038\/s41598-023-31340-1.","DOI":"10.1038\/s41598-023-31340-1"},{"key":"2026030615414913548_j_dx-2025-0014_ref_009","doi-asserted-by":"crossref","unstructured":"Sanchez, K, Kamal, K, Manjaly, P, Ly, S, Mostaghimi, A. Clinical application of artificial intelligence for non-melanoma skin cancer. Curr Treat Options Oncol 2023;24:373\u20139. https:\/\/doi.org\/10.1007\/s11864-023-01065-4.","DOI":"10.1007\/s11864-023-01065-4"},{"key":"2026030615414913548_j_dx-2025-0014_ref_010","doi-asserted-by":"crossref","unstructured":"Liu, X, Duan, C, Kim, M, Zhang, L, Jee, E, Maharjan, B, et al.. Claude 3 opus and ChatGPT with GPT-4 in dermoscopic image analysis for melanoma diagnosis: comparative performance analysis. JMIR Med Inform 2024;12:e59273. https:\/\/doi.org\/10.2196\/59273.","DOI":"10.2196\/59273"},{"key":"2026030615414913548_j_dx-2025-0014_ref_011","unstructured":"da Silva, SF: Atlas dermatol\u00f3gico. https:\/\/atlasdermatologico.com.br\/index.jsf [Accessed 20 Dec 2024]."},{"key":"2026030615414913548_j_dx-2025-0014_ref_012","unstructured":"A cooperation between the dept of clinical social medicine (Univ of Heidelberg) and the dept of dermatology (Univ of Erlangen). DermIS - Dermatology Information System. https:\/\/www.dermis.net\/dermisroot\/en\/home\/index.htm [Accessed 20 Dec 2024]."},{"key":"2026030615414913548_j_dx-2025-0014_ref_013","unstructured":"Clark, JA and contributors. Pillow documentation. https:\/\/pillow.readthedocs.io\/en\/stable\/index.html [Accessed 20 Dec 2024]."},{"key":"2026030615414913548_j_dx-2025-0014_ref_014","doi-asserted-by":"crossref","unstructured":"Zhou, J, He, X, Sun, L, Xu, J, Chen, X, Chu, Y, et al.. Pre-trained multimodal large language model enhances dermatological diagnosis using SkinGPT-4. Nat Commun 2024;15:5649. https:\/\/doi.org\/10.1038\/s41467-024-50043-3.","DOI":"10.1038\/s41467-024-50043-3"},{"key":"2026030615414913548_j_dx-2025-0014_ref_015","doi-asserted-by":"crossref","unstructured":"Bhandari, A. Revolutionizing radiology with artificial intelligence. Cureus 2024;16:e72646. https:\/\/doi.org\/10.7759\/cureus.72646.","DOI":"10.7759\/cureus.72646"},{"key":"2026030615414913548_j_dx-2025-0014_ref_016","doi-asserted-by":"crossref","unstructured":"Navigli, R, Conia, S, Ross, B. Biases in Large Language Models: origins, inventory, and discussion. J\u00a0Data Inf Qual 2023;15. https:\/\/doi.org\/10.1145\/3597307.","DOI":"10.1145\/3597307"},{"key":"2026030615414913548_j_dx-2025-0014_ref_017","doi-asserted-by":"crossref","unstructured":"Gilbert, S. The EU passes the AI Act and its implications for digital medicine are unclear. NPJ Digit Med 2024;7:135. https:\/\/doi.org\/10.1038\/s41746-024-01116-6.","DOI":"10.1038\/s41746-024-01116-6"},{"key":"2026030615414913548_j_dx-2025-0014_ref_018","doi-asserted-by":"crossref","unstructured":"Vasey, B, Nagendran, M, Campbell, B, Clifton, DA, Collins, GS, Denaxas, S, et al.. Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. Nat Med 2022;28:924\u201333. https:\/\/doi.org\/10.1038\/s41591-022-01772-9.","DOI":"10.1136\/bmj-2022-070904"},{"key":"2026030615414913548_j_dx-2025-0014_ref_019","doi-asserted-by":"crossref","unstructured":"Lau, K, Halligan, J, Fontana, G, Guo, C, O\u2019Driscoll, FK, Prime, M, et al.. Evolution of the clinical simulation approach to assess digital health technologies. Future Healthc J 2023;10:173\u20135. https:\/\/doi.org\/10.7861\/fhj.2022-0145.","DOI":"10.7861\/fhj.2022-0145"}],"container-title":["Diagnosis"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/dx-2025-0014\/xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/dx-2025-0014\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T15:42:38Z","timestamp":1772811758000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/dx-2025-0014\/html"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,23]]},"references-count":19,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,2,25]]},"published-print":{"date-parts":[[2026,2,27]]}},"alternative-id":["10.1515\/dx-2025-0014"],"URL":"https:\/\/doi.org\/10.1515\/dx-2025-0014","relation":{},"ISSN":["2194-802X"],"issn-type":[{"value":"2194-802X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,5,23]]}}}