Dermatologist-like explainable AI enhances trust and confidence in diagnosing melanoma

Brinker TJ, Chanda T, Hauser K, Hobelsberger S, Bucher TC, Garcia CN, Wies C, Kittler H, Tschandl P, Navarrete Dechent C, Podlipnik S, Chousakos E, Crnaric I, Majstorovic J, Alhajwan L, Foreman T, Peternel S, Sarap S, Özdemir İ, Barnhill RL, Llamas-Velasco M, Poch G, Korsing S, Sondermann W, Gellrich FF, Heppt M, Erdmann M, Haferkamp S, Drexler K, Goebeler M, Schilling B, Utikal JS, Ghoreschi K, Fröhling S, Krieghoff-Henning E, Salava A, Thiem A, Dimitrios A, Ammar AM, Vučemilović AS, Yoshimura AM, Ilieva A, Gesierich A, Reimer-Taschenbrecker A, Kolios AG, Kalva A, Ferhatosmanoğlu A, Beyens A, Pföhler C, Erdil DI, Jovanovic D, Racz E, Bechara FG, Vaccaro F, Dimitriou F, Rasulova G, Cenk H, Yanatma I, Kolm I, Hoorens I, Sheshova IP, Jocic I, Knuever J, Fleißner J, Thamm JR, Dahlberg J, Lluch-Galcerá JJ, Figueroa JSA, Holzgruber J, Welzel J, Damevska K, Mayer KE, Maul LV, Garzona-Navas L, Bley LI, Schmitt L, Reipen L, Shafik L, Petrovska L, Golle L, Jopen L, Gogilidze M, Burg MR, Morales-Sánchez MA, Sławińska M, Mengoni M, Dragolov M, Iglesias-Pena N, Booken N, Enechukwu NA, Persa OD, Oninla OA, Theofilogiannakou P, Kage P, Neto RRO, Peralta R, Afiouni R, Schuh S, Schnabl-Scheu S, Vural S, Hudson S, Saa SR, Hartmann S, Damevska S, Finck S, Braun SA, Hartmann T, Welponer T, Sotirovski T, Bondare-Ansberga V, Ahlgrimm-Siess V, Frings VG, Simeonovski V, Zafirovik Z, Maul JT, Lehr S, Wobser M, Debus D, Riad H, Pereira MP, Lengyel Z, Balcere A, Tsakiri A, Braun RP (2024)

Publication Language: English

Publication Type: Journal article

Publication year: 2024

Journal

Book Volume: 15

Article Number: 524

Journal Issue: 1

DOI: 10.1038/s41467-023-43095-4

Abstract

Artificial intelligence (AI) systems have been shown to help dermatologists diagnose melanoma more accurately, however they lack transparency, hindering user acceptance. Explainable AI (XAI) methods can help to increase transparency, yet often lack precise, domain-specific explanations. Moreover, the impact of XAI methods on dermatologists’ decisions has not yet been evaluated. Building upon previous research, we introduce an XAI system that provides precise and domain-specific explanations alongside its differential diagnoses of melanomas and nevi. Through a three-phase study, we assess its impact on dermatologists’ diagnostic accuracy, diagnostic confidence, and trust in the XAI-support. Our results show strong alignment between XAI and dermatologist explanations. We also show that dermatologists’ confidence in their diagnoses, and their trust in the support system significantly increase with XAI compared to conventional AI. This study highlights dermatologists’ willingness to adopt such XAI systems, promoting future use in the clinic.

Authors with CRIS profile

Involved external institutions

Deutsches Krebsforschungszentrum (DKFZ) Germany (DE) Medizinische Universität Wien Austria (AT) University Hospital Split Croatia (HR) Special Hospital for Surgical Diseases Zan Mitrev Clinic / Специјална болница по хируршки болести "Жан Митрев Клиника" / Specijalna bolnica po hirurški bolesti "Zan Mitrev Klinika" Republic of North Macedonia (MK) Sisters of Charity Hospital / Klinički Bolnički Centar Sestre Milosrdnice (KBCSM) Croatia (HR) Dubai London Clinic & Speciality Hospital United Arab Emirates (AE) Sveučilište u Rijeci / University of Rijeka Croatia (HR) Evangelismos Medical Center Greece (GR) Charité - Universitätsmedizin Berlin Germany (DE) Universidad Autónoma de Madrid (UAM) Spain (ES) Universitat de Barcelona (UB) / University of Barcelona Spain (ES) Universitätsklinikum Regensburg Germany (DE) Universitätsklinikum Carl Gustav Carus Dresden Germany (DE) Universitätsklinikum Würzburg Germany (DE) Universitätsklinikum Essen Germany (DE) Universitätsspital Zürich (USZ) Switzerland (CH) Fethiye State Hospital / Fethiye Devlet Hastanesi Turkey (TR) University Hospital Ghent Belgium (BE) Military Medical Academy (Serbia) / Војномедицинска академија / Vojnomedicinska akademija (VMA) Serbia (RS) Universitätsklinikum des Saarlandes (UKS) Germany (DE) Northwestern University United States (USA) (US) Dermatology Aesthetic Clinic Dr.Iskra Republic of North Macedonia (MK) Koç-Universität (KU) Turkey (TR) Rīga Stradiņš University Latvia (LV) Universitätsspital Basel Switzerland (CH) Medical University Gdansk / Gdański Uniwersytet Medyczny Poland (PL) Hospital Clínica Bíblica Costa Rica (CR) Universidad del Desarrollo Chile (CL) Universitätsklinikum Tübingen Germany (DE) Universitätsklinikum Heidelberg Germany (DE) Universidade Federal de Mato Grosso do Sul Brazil (BR) Kepler Universitätsklinikum (KUK) Austria (AT) University Hospital of Umeå / Norrlands universitetssjukhus (NUS) Sweden (SE) Universitätsklinikum Hamburg-Eppendorf (UKE) Germany (DE) Universitätsklinikum Augsburg Germany (DE) Technische Universität München (TUM) Germany (DE) Universitätsklinikum Magdeburg A.ö.R. Germany (DE) Universitätsklinikum Münster Germany (DE) Acibadem City Clinic Tokuda Hospital Bulgaria (BG) Health Sciences Research Institute of the “Germans Trias i Pujol” Foundation (IGTP) Spain (ES) Paracelsus Medizinische Privatuniversität Austria (AT) Obafemi Awolowo University Nigeria (NG) Technische Universität Dresden Germany (DE) Paracelsus Medizinische Privatuniversität, Nürnberg Germany (DE) Riga 1st hospital / Rīgas 1. slimnīca Latvia (LV) Universitätsklinikum Freiburg Germany (DE) Nnamdi Azikiwe University Nigeria (NG) Saints Cyril and Methodius University of Skopje / Универзитет „Св. Кирил и Методиј“ во Скопје Republic of North Macedonia (MK) Klinikum Stuttgart Germany (DE) University of Messina / Università degli Studi di Messina Italy (IT) East Surrey Hospital United Kingdom (GB) Universitätsklinikum Köln Germany (DE) University of Pécs / Pécsi Tudományegyetem Hungary (HU) Universidade Estadual de Campinas (UNICAMP) / University of Campinas Brazil (BR) Clinic Kani Georgia (GE) Pamukkale University / Pamukkale Üniversitesi (PAU) Turkey (TR) Universitätsmedizin Rostock Germany (DE) Helsinki University Central Hospital (HUCH) / Helsingin seudun yliopistollinen keskussairaala (HYKS) Finland (FI) Al-Azhar University Egypt (EG) National and Kapodistrian University of Athens Greece (GR) Gazi University Turkey (TR) Universitätsklinikum Mannheim / University Medical Centre Mannheim (Universitätsmedizin Mannheim) Germany (DE) Universitätsklinikum Halle (Saale) Germany (DE) Université Saint-Joseph (USJ) / جامعة القديس يوسف / Սուրբ Ժոզեֆի համալսարան Lebanon (LB) Institut Curie France (FR) Universidad de Buenos Aires (UBA) / University of Buenos Aires Argentina (AR) Karadeniz Technical University / Karadeniz Teknik Üniversitesi (KTU) Turkey (TR) Ruhr-Universität Bochum (RUB) Germany (DE) University Medical Center Groningen (UMCG) / Universitair Medisch Centrum Groningen Netherlands (NL) Taksim Eğitim ve Araştırma Hastanesi Turkey (TR) Clinical Hospital Shtip Republic of North Macedonia (MK) Newport Beach Surgery Center United States (USA) (US) Pontifical Catholic University of Chile / Pontificia Universidad Católica de Chile Chile (CL)

How to cite

APA:

Brinker, T.J., Chanda, T., Hauser, K., Hobelsberger, S., Bucher, T.C., Garcia, C.N.,... Braun, R.P. (2024). Dermatologist-like explainable AI enhances trust and confidence in diagnosing melanoma. Nature Communications, 15(1). https://doi.org/10.1038/s41467-023-43095-4

MLA:

Brinker, Titus J., et al. "Dermatologist-like explainable AI enhances trust and confidence in diagnosing melanoma." Nature Communications 15.1 (2024).

BibTeX: Download