Clinical practice still heavily relies on the manual and mental effort of dedicated specialists for diagnosis and prognosis, despite the many advances of the field of visual computing. For example, tumor and healthy tissue segmentations are very often conducted manually by radiologists, who manually delineate these tissues slice-by-slice on medical imaging data. Yet, such human-in-the-loop approaches entail uncertainty, brought in by inter-observer variability (e.g., two radiologists might not explore and analyze the data in the same way, and might produce different results).
In this project, we want to investigate visual analytics strategies that track potential inter-observer variability in the exploration and analysis of medical imaging data, and that provide a means for understanding how inter-observer variability can affect the outcome of, for example, a diagnostic procedure.
• Interest and knowledge in biomedical visualization and in visual storytelling.
• Some basics of graph visualization.
• Good programming skills.
• Creativity and enthusiasm.
To be discussed (depending on the background of the student).
 Amabili, Lorenzo, et al. "Improving Provenance Data Interaction for Visual Storytelling in Medical Imaging Data Exploration." EuroVis (Short Papers). 2018.
 Bors, Christian, et al. "A provenance task abstraction framework." IEEE computer graphics and applications 39.6 (2019): 46-60.
 Heer, Jeffrey, et al. "Graphical histories for visualization: Supporting analysis, communication, and evaluation." IEEE transactions on visualization and computer graphics 14.6 (2008): 1189-1196.