Computer Vision News - May 2021

13 Roxane Licandro Dynamics of developmental patterns especially in the medical field are versatile in space and also over time and require to be addressed in all their aspects without deciding on preserving time point specific dynamics or between time-dependent dynamics. The concept Spatio Temporal Modelling of Dynamic Developmental Patterns proposed in this work provides a novel strategy to address the gaps in current longitudinal modelling approaches and provides a strategy to handle dynamics in space and also over time. The focus of the work lies in encoding and understanding baseline trajectories disentangled from time-dependent or systemic dynamics. Thus, on the one hand the identification of suitable baseline states is essential and on the other hand the development of techniques to analyze the dynamics’ deviations and relations to the baseline is required. In Figure 2 it is shown that e.g. a spatio temporal model of focal bone lesion progression created by the concept proposed, can be used to predict the temporal evolution risk of focal bone lesions for a defined time point in the future. Figure 2: The Spatio Temporal Model of Focal Lesion Progression in Multiple Myeloma is used to predict the temporal risk of bone lesions to emerge at a defined future time point. Image courtesy Medical University of Vienna. [Projec t DACHMM] In this work it is demonstrated that the proposed modelling concept is capable to flexibly model DDPs independent of the imaging modalities, of different populations/ age ranges and applications to answer research questions in the field of computer vision, cancer research, brain development and functional connectivity network analysis. It led to the development of novel data representation forms for DDPs, segmentation strategies, classification procedures and time-dependent prediction approaches, outperforming state of the art methods. A list of corresponding publications is provided here. A big thank you to my supervisors Georg Langs and Martin Kampel for their ongoing support and to all the passionate researchers I met during the journey of my doctorate. You made it a special one!