Ventricular blood flow analysis using topological methods

Thanks to the advances of data acquisition techniques, we can acquire ventricular blood flow data with very high quality. This extremely complex spatiotemporal data calls for novel visualization and analysis tools. In particular, the new tools need to assist domain experts in quick identification of critical patterns. In this paper, we present a method using topo-logical data analysis tools with simulated ventricular blood flow, and automatically detect interesting topological features within the flow. We show that this completely unsupervised framework detects and extracts eddies formed from vortex shedding during late diastole, which normally requires highly specialized algorithms to capture.