Research On Cardiac Electrophysiology Fast Simulation And Interactive Visualization

Virtual heart is used to reproduce the cardiac structures and functionalities bymeans of mathematical modeling and computer simulation methods. Its purposeis to reveal the operating mechanism of the heart and to study the problems inthe physiological efects of the drug phenomenon and cardiac reactions. Cardiacelectrophysiological simulation is the main research content of the virtual heart,which could be used to analyze physiology, pathology and pharmacological aspectsof cardiac electrophysiological phenomena. Visualization is the basis of virtual re-production of the heart on computer and the main approach to verifying the resultsof virtual heart simulation and modeling. So the research of cardiac electrophysiol-ogy simulation and visualization is of great importance for the entire virtual heartresearch area.This dissertation starts with cardiac electrophysiological simulation and visual-ization problems and focuses on the study of electrophysiological simulation, cardiacanatomy data and electrophysiological simulation data visualization. In this disser-tation, we try to establish the linkage between cardiac anatomy data visualizationand functional data visualization and the relationship between electrophysiologicalsimulation and visualization with graphics processing unit,(GPU). This dissertationare organized as follows:GPU-based fast simulation of cardiac electrophysiology The main prob-lem of cardiac electrophysiological simulation is the difculty of fnding an efcientsimulation implementation. Based on the electrophysiological characteristics we pro-pose a fast GPU-based electrophysiological simulation method. The parallelism ofcardiac electrophysiology simulation is analyzed, the feasibility of the GPU imple-mentation of electrophysiological simulation is discussed and the fast GPU-basedelectrophysiology simulation is implemented. Optimization strategies are discussedbased on the characteristics of electrophysiological simulation process.GPU-based visualization of cardiac anatomy Cardiac anatomy is com-plicated and the internal tissues are similar, which make various organs of cardiacnot efectively distinguished. To solve this problem, this dissertation proposes a texture-slicing based visualization method of heart tissue segmentation and thus re-veals the specifc heart organization. For the problem of contour information of eachorgan difcult to be efectively revealed in heart layered structure caused by mutu-al occlusion of tissues, this dissertation proposes an improved Context-Preservingmodel. Experimental results show that the model integrated in GPU based ray cast-ing method has higher visualization quality than mainstream anatomy visualizationmethods.Interactive visualization of the fne cardiac structure For the problemof fne tissues in cardiac anatomy structure difcult to be efectively visualized,this dissertation proposes a statistical information based interactive visualizationmethod. In the premise of ensuring efective displaying of whole cardiac anatomy,this method allows users to interactively obtain the visualization of the fne cardiacstructure. On the basis of human cognition to diferent tones of colors, combinedwith light transfer function, this dissertation proposes an enhanced illuminationvisualization method based on perception, which can efectively enhance selectedheart tissues, while suppressing other organizations.Cardiac electrophysiology simulation data visual analysis In order toefectively demonstrate the time-varying property of electrophysiological simulationdata, this dissertation proposes an online visualization method for electrophysiologi-cal simulation data. Under the conditions of assuring visualizing quality, this methodefectively shows the heart tissue voltage distribution changing over time. For theproblem of cardiac electrophysiology simulation data of internal tissue difcult tobe efectively demonstrated, this paper proposes a multi-volume data visualizationapproach accurately showing spatial information of electrophysiological simulationdata by providing anatomical structures context for those organs with electrophysio-logical phenomena. In order to solve the problem of inefcient online visualization ofthree-dimensional simulation of electrophysiology, the dissertation concludes with avisualization method of electrophysiological simulation data sequence to efectivelyimprove the three-dimensional rendering efciency of electrophysiological simulationdata.Experimental results show that the proposed fast GPU-based simulation methodefectively reduced electrophysiological simulation time and improved the operationefciency. Cardiac anatomy and electrophysiological simulation data visualization methods, not only can achieve real-time rendering, but also can efectively enhancecontour information of the cardiac tissues. The selected tissues and the tissues withelectrophysiological phenomena can be efectively highlighted.