Eye tracking makes it possible to analyze visual attention during medical decision-making processes and to determine which areas of an image healthcare professionals focus on when making a diagnosis. The platform, developed by an interdisciplinary team, is based on an application previously used in digital neuropathology. As part of a comprehensive overhaul, it was redesigned to be more modular and enhanced with features that make study workflows more flexible and applicable across different domains. The revised platform was then evaluated in a clinical study involving 14 dermatologists to assess the multi-class classification of wounds. The study combined visual data with the analysis of defined image regions and assessed the application’s usability using the standardized System Usability Scale (SUS).
Study Confirms High User-Friendliness
The conclusion: The platform demonstrates very good usability. With an average SUS score of 83.75, participants rated the platform as particularly user-friendly. In addition, the physicians provided positive feedback regarding the platform’s ability to help them reflect on their own diagnostic skills. “The results make it clear that eye tracking is not only a tool for analyzing eye movements, but can also contribute to self-reflection and the further development of diagnostic processes,” explains Vinzent Bücheler.
About the Study
The study combines expertise in medicine, computer science, and health data science: It was developed through an interdisciplinary collaboration among researchers at Neu-Ulm University of Applied Sciences, Würzburg University Hospital, the University of Würzburg, and Ulm University Hospital. The corresponding short paper was authored by Vinzent Bücheler (HNU), Tassilo Dege (University Hospital Würzburg), Christofer Pohl (HNU), Daniel Hieber (University Hospital Ulm), Vanessa Borst (University of Würzburg), Rüdiger Pryss (University of Würzburg), Astrid Schmieder (University Hospital Würzburg), and Prof. Dr. Johannes Schobel (HNU). At IEEE CBMS 2026, lead author Vinzent Bücheler presented the study’s findings to an international audience of experts and exchanged ideas with researchers in the field of medical information systems.
The short paper presented will be published shortly in the proceedings of IEEE CBMS 2026.
