Prof. Dr. Lars Linsen, Institut für Informatik

Clinical practice often focuses on the investigation of one single disease, while the health status of a human is much more complex and may depend on many factors. Recently, cohort studies have been introduced to investigate, in longitudinal studies, the health status of an entire population (the cohort) by capturing health record data, whole-body medical imaging data, personal data including socio-economical circumstances, and even genetic sequencing data. Given this large amount of heterogeneous data, there is a lack of proper tools for its multi-variate analysis. In this project, we propose novel interactive visual analysis methods for testing hypotheses, supporting the generation of new hypotheses, and investigating changes over time. The goal is to allow for the detection of risk or biomarkers and even genetic associations in a multi-variate setting.In the second funding period, the research conducted in the first funding period shall be enhanced in various aspects. We will put a particular focus on the time aspect in multi-dimensional heterogeneous data from longitudinal studies, the analysis of influencing factors, analyzing multi-dimensional heterogeneous data with missing entries, and analyzing sparse high-dimensional data from genome-wide association studies.Moreover, we would like to validate the effectiveness of the proposed analysis methods by performing comparative visual analyses of the multi-dimensional heterogeneous data from different cohort studies.

Advances in 4D phase-contrast magnetic resonance imaging (PC-MRI) allow for fast in-vivo measurements of unsteady blood flow in animals and humans. Still, the low spatio-temporal resolution, the low signal-to-noise ratio, and the uncertainty due to acceleration inhibit a proper analysis of the hemodynamic parameters which hamper their application as diagnostic markers in the clinical routine. Data-driven computational fluid dynamics (CFD) allow for simulated blood flow at high spatio-temporal resolution. Tuning the many simulation parameters to obtain a simulated blood flow that best matches the imaging data in an iterative procedure is a tedious task though and prone to detecting sub-optimal solutions that do not match the measured blood flow well enough. We propose to transform the task to a data analysis task by generating a simulation ensemble with many feasible parameter settings and analyzing the ensemble in comparison to the measured data. This approach requires multiple contributions: First, a fast reliable PC-MRI sequence will be developed to produce high-quality imaging data with low bias in-vitro and in-vivo in an appropriate measurement time. Second, an efficient data-driven CFD approach that allows for moving walls will be developed. Third, an effective analysis methodology for comparing measured and simulated data will be developed. To incorporate expert knowledge in the analysis process, we propose a user-centric approach that allows for interactive visual comparisons at a global level as well as in selected spatio-temporal regions of interest. Putting the three contributions together in a data assimilation process delivers a software tool to generate subject-specific blood flow fields from measured PC-MRI data at high spatio-temporal resolution. By compensating the limited spatio-temporal domains of MRI datasets it enables an accurate quantitative analysis of hemodynamic parameters. Using this software tool, we will provide a comprehensive description of the hemodynamic changes in the development and progression of atherosclerosis in a murine model.