Dermatologist-like explainable AI enhances trust and confidence in diagnosing melanoma
Tirtha Chanda, Katja Hauser, Sarah Hobelsberger, Tabea-Clara Bucher, Carina Nogueira Garcia, Christoph Wies, Harald Kittler, Philipp Tschandl, Cristián Navarrete‐Dechent, Sebastián Podlipnik, Emmanouil Chousakos, Iva Crnaric, Jovana Majstorović, Linda Alhajwan, Tanya Foreman, Sandra Peternel, Sergei Sarap, İrem Özdemir, Raymond L. Barnhill, Mar Llamas‐Velasco, Gabriela Poch, Sören Korsing, Wiebke Sondermann, Frank Friedrich Gellrich, Markus V. Heppt, Michael Erdmann, Sebastian Haferkamp, Konstantin Drexler, Matthias Goebeler, Bastian Schilling, Jochen Utikal, Kamran Ghoreschi, Stefan Fröhling, Eva Krieghoff‐Henning, Alexander Salava, Alexander Thiem, Alexandris Dimitrios, Amr Mohammad Ammar, Ana Sanader Vučemilović, Andrea Miyuki Yoshimura, Andzelka Ilieva, Anja Gesierich, Antonia Reimer‐Taschenbrecker, Antonios G.A. Kolios, Arturs Kaļva, Arzu Ferhatosmanoğlu, Aude Beyens, Claudia Pföhler, Dilara Ilhan Erdil, Dobrila Jovanovic, Emöke Rácz, Falk G. Bechara, Federico Vaccaro, Florentia Dimitriou, Günel Rasulova, Hülya Cenk, Irem Yanatma, Isabel Kolm, Isabelle Hoorens, Iskra Petrovska Sheshova, Ivana Jocic, Jana Knuever, Janik Fleißner, Janis Thamm, Johan Dahlberg, Juan José Lluch‐Galcerá, Juan Sebastián Andreani Figueroa, Julia Holzgruber, Julia Welzel, Katerina Damevska, Kristine Elisabeth Mayer, Lara Valeska Maul, Laura Garzona-Navas, Laura Isabell Bley, Laurenz Schmitt, Lena Reipen, Lidia Shafik, Lidija Petrovska, Linda Golle, Luise Jopen, Magda Gogilidze, Maria Rosa Burg, Martha Alejandra Morales‐Sánchez, Martyna Sławińska, Miriam Mengoni, Miroslav Dragolov, N. Iglesias-Pena, Nina Booken, Nkechi Anne Enechukwu, Oana‐Diana Persa, Olumayowa Abimbola Oninla, Panagiota Theofilogiannakou, Paula Kage, Roque Rafael Oliveira Neto, Rosario Peralta, Rym Afiouni, Sandra Schuh, Saskia Schnabl-Scheu, Seçil Vural, Sharon Hudson
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.