Interaction and Interface Design for Primary Ciliary Dyskinesia Annotation Tool

Dana Hrivnáková

Supervisor(s): Ing. Miroslav Laco, PhD.

Slovak Technical University

Abstract: Artificial intelligence (AI) has become a key tool in various fields, including medicine. AI models can effectively analyze cytological medical imagery data for Primary ciliary dyskinesia (PCD) diagnosis. This is multi-class classification problem, where each cilium is assigned a specific defect and dynein type. These models support doctors in their daily work and speed up their routine tasks. The goal of our research is to propose a methodology that optimizes the interaction between domain expert (DE) and AI in the field of histopathology through an innovative interface design. To achieve this, we focused on user experience (UX) and usability in medicine. Based on the state-of-the-art in the field and case study with DE in the field of PCD diagnostics, we designed a user interface to assist histopathologists in diagnosing PCD. To achieve our goals, we propose a modified Domain-Expert-Centered Double Diamond design methodology (DEDDM). Using our proposed DEDDM model, we designed a prototype of medical imagery annotation tool for PCD that meets the needs of experts working under time constraints. The annotation tool is designed to support manual annotation flow, as well as future integration of AI which will make DE's workflow even more efficient. We preliminary evaluated the design prototype, during which we identified 12 usability issues. After familiarizing themselves with the application layout, participants found it intuitive and pleasant to use. However, they initially faced difficulties understanding the interaction concept of creating manual annotations.
Keywords: Human-Computer Interaction
Full text:
Year: 2025