Journal Article

DKFZ-2025-01121

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Dermatologist-like explainable AI enhances melanoma diagnosis accuracy: eye-tracking study.

Chanda, T. (First author)DKFZ* ; Haggenmueller, S.DKFZ* ; Bucher, T.DKFZ* ; Holland-Letz, T.DKFZ* ; Kittler, H. ; Tschandl, P. ; Heppt, M. V. ; Berking, C. ; Utikal, J.DKFZ* ; Schilling, B. ; Buerger, C. ; Navarrete-Dechent, C. ; Goebeler, M. ; Kather, J. N. ; Schneider, C. V. ; Durani, B. ; Durani, H. ; Jansen, M. ; Wacker, J. ; Wacker, J. ; Consortium, R. S. (Collaboration Author) ; Brinker, T. (Last author)DKFZ* ; Booken, N. (Contributor) ; Ahlgrimm-Siess, V. (Contributor) ; Welzel, J. (Contributor) ; Persa, O.-D. (Contributor) ; Dimitriou, F. (Contributor) ; Braun, S. A. (Contributor) ; Maul, L. V. (Contributor) ; Reimer-Taschenbrecker, A. (Contributor) ; Schuh, S. (Contributor) ; Bechara, F. G. (Contributor) ; Feldmeyer, L. (Contributor) ; Mühleisen, B. (Contributor) ; Gössinger, E. (Contributor) ; Braun, S. A. (Contributor) ; Nguyen, V. A. (Contributor) ; Maul, J.-T. (Contributor) ; Hoffmann, F. (Contributor) ; Pföhler, C. (Contributor) ; Thamm, J. (Contributor) ; Ludwig-Peitsch, W. (Contributor) ; Hartmann, D. (Contributor) ; Garzona-Navas, L. (Contributor) ; Sławińska, M. (Contributor) ; Theofilogiannakou, P. (Contributor) ; Vucemilovic, A. S. (Contributor) ; Lluch-Galcerá, J. J. (Contributor) ; Beyens, A. (Contributor) ; Erdil, D. I. (Contributor) ; Afiouni, R. (Contributor) ; Bondare-Ansberga, V. (Contributor) ; Morales-Sánchez, M. A. (Contributor) ; Ferhatosmanoğlu, A. (Contributor) ; Neto, R. R. O. (Contributor) ; Petrovska, L. (Contributor) ; Tsakiri, A. (Contributor) ; Cenk, H. (Contributor) ; Hudson, S. (Contributor) ; Dragolov, M. (Contributor) ; Zafirovik, Z. (Contributor) ; Jocic, I. (Contributor) ; Balcere, A. (Contributor) ; Lengyel, Z. (Contributor) ; Salava, A. (Contributor) ; Hoorens, I. (Contributor) ; Saa, S. R. (Contributor) ; Rácz, E. (Contributor) ; Salerni, G. (Contributor) ; Manuelyan, K. (Contributor) ; Ammar, A. M. (Contributor) ; Erdmann, M. (Contributor) ; Wagner, N. (Contributor) ; Sambale, J. (Contributor) ; Kemenes, S. (Contributor) ; Ronicke, M. (Contributor) ; Sollfrank, L. (Contributor) ; Bosch-Voskens, C. (Contributor) ; Sagonas, I. (Contributor) ; Breakell, T. (Contributor) ; Uebel, C. (Contributor) ; Zieringer, L. (Contributor) ; Hoener, M. (Contributor) ; Rabe, L. (Contributor) ; Sackmann, T. (Contributor) ; Baumert, J. (Contributor) ; Schaarschmidt, M. L. (Contributor) ; Ninosu, N. (Contributor) ; Yilmaz, K. (Contributor) ; Dionysia, D. (Contributor) ; Christ, F. (Contributor) ; Fahimi, S. (Contributor) ; Loos, S. (Contributor) ; Sachweizer, A. (Contributor) ; Gosmann, J. (Contributor) ; Weberschock, T. (Contributor) ; Erdogdu, U. (Contributor) ; Buchinger, A. (Contributor) ; Lunderstedt, J. (Contributor) ; Funk, T. (Contributor) ; Klifo, H. (Contributor) ; Kiefer, S. (Contributor) ; Klifo, D. (Contributor) ; Kalski, M. (Contributor)

2025
Springer Nature [London]

This record in other databases:  

Please use a persistent id in citations: doi:10.1038/s41467-025-59532-5

Abstract: Artificial intelligence (AI) systems substantially improve dermatologists' diagnostic accuracy for melanoma, with explainable AI (XAI) systems further enhancing their confidence and trust in AI-driven decisions. Despite these advancements, there remains a critical need for objective evaluation of how dermatologists engage with both AI and XAI tools. In this study, 76 dermatologists participate in a reader study, diagnosing 16 dermoscopic images of melanomas and nevi using an XAI system that provides detailed, domain-specific explanations, while eye-tracking technology assesses their interactions. Diagnostic performance is compared with that of a standard AI system lacking explanatory features. Here we show that XAI significantly improves dermatologists' diagnostic balanced accuracy by 2.8 percentage points compared to standard AI. Moreover, diagnostic disagreements with AI/XAI systems and complex lesions are associated with elevated cognitive load, as evidenced by increased ocular fixations. These insights have significant implications for the design of AI/XAI tools for visual tasks in dermatology and the broader development of XAI in medical diagnostics.

Keyword(s): Melanoma: diagnosis (MeSH) ; Melanoma: diagnostic imaging (MeSH) ; Humans (MeSH) ; Artificial Intelligence (MeSH) ; Eye-Tracking Technology (MeSH) ; Dermatologists (MeSH) ; Skin Neoplasms: diagnosis (MeSH) ; Skin Neoplasms: diagnostic imaging (MeSH) ; Female (MeSH) ; Dermoscopy: methods (MeSH) ; Male (MeSH) ; Nevus: diagnosis (MeSH) ; Nevus: diagnostic imaging (MeSH) ; Adult (MeSH) ; Middle Aged (MeSH)

Note: #EA:C140#LA:C140#

---

Contributing Institute(s):

1. Digitale Prävention, Diagnostik und Therapiesteuerung (C140)
2. C060 Biostatistik (C060)
3. KKE Dermatoonkologie (A370)

Research Program(s):

1. 313 - Krebsrisikofaktoren und Prävention (POF4-313) (POF4-313)

Appears in the scientific report 2025

---

Database coverage:
; ; ; Article Processing Charges ; BIOSIS Previews ; Biological Abstracts ; Clarivate Analytics Master Journal List ; Current Contents - Agriculture, Biology and Environmental Sciences ; Current Contents - Life Sciences ; Current Contents - Physical, Chemical and Earth Sciences ; DOAJ Seal ; Essential Science Indicators ; Fees ; IF >= 15 ; JCR ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection ; Zoological Record

---