Domain Expert in the Loop of Digitized Histopathology Education and Artificial Intelligence

Erika Váczlavová

Supervisor(s): Miroslav Laco

Slovak Technical University

Abstract: In this paper, we propose a way to use a graphical user interface to present digitized multi-modal data in the field of medicine for specific domain experts. Our data consisted of digitized histopathology specimens, subject to expert examination. As the digitization of histopathology for educational purposes is only in its beginning stages, we explore how to present the data to experts in a way to encourages them to build up their confidence in digitized workflow. As part of this research, we are working on streamlining the workflow by designing assistance tools based on artificial intelligence (AI). While presenting the results of AI to specific domain experts in medicine, it is important to choose the right explainability of the results of black-box algorithms, and how to present the outputs in the user interface. We found out that the implementation of functionalities driven by artificial intelligence depends on the level of expertise of the domain expert. The differences are observed in a case study with cooperation from medical students and doctors, who got access to digitized multi-modal data with AI-powered functionalities in iteratively designed prototypes of the specialized system for education in the field of histopathology. We present outcomes from the aforementioned case study to serve as a base for the future development of specialized interfaces in the field of digitized histopathology.
Keywords: Human-Computer Interaction
Full text:
Year: 2024