| Autor: |
Alves JR; Department of Education, Federal Institute of Education, Science and Technology of Mato Grosso, Sorriso 78895-150, Brazil., Berg LA; Department of Computer Science, Federal Univesity of Juiz de Fora, Juiz de Fora 36036-900, Brazil.; Department of Computer Science, University of Oxford, Oxford OX3 7LD, UK., Gaio ED; Department of Computer Science, Federal Univesity of Juiz de Fora, Juiz de Fora 36036-900, Brazil., Rocha BM; Department of Computer Science, Federal Univesity of Juiz de Fora, Juiz de Fora 36036-900, Brazil., de Queiroz RAB; Departament of Computing, Federal Univesity of Ouro Preto, Ouro Preto 35400-000, Brazil., Dos Santos RW; Department of Computer Science, Federal Univesity of Juiz de Fora, Juiz de Fora 36036-900, Brazil. |
| Abstrakt: |
This paper presents a novel hybrid approach for the computational modeling of cardiac perfusion, combining a discrete model of the coronary arterial tree with a continuous porous-media flow model of the myocardium. The constructive constrained optimization (CCO) algorithm captures the detailed topology and geometry of the coronary arterial tree network, while Poiseuille's law governs blood flow within this network. Contrast agent dynamics, crucial for cardiac MRI perfusion assessment, are modeled using reaction-advection-diffusion equations within the porous-media framework. The model incorporates fibrosis-contrast agent interactions and considers contrast agent recirculation to simulate myocardial infarction and Gadolinium-based late-enhancement MRI findings. Numerical experiments simulate various scenarios, including normal perfusion, endocardial ischemia resulting from stenosis, and myocardial infarction. The results demonstrate the model's efficacy in establishing the relationship between blood flow and stenosis in the coronary arterial tree and contrast agent dynamics and perfusion in the myocardial tissue. The hybrid model enables the integration of information from two different exams: computational fractional flow reserve (cFFR) measurements of the heart coronaries obtained from CT scans and heart perfusion and anatomy derived from MRI scans. The cFFR data can be integrated with the discrete arterial tree, while cardiac perfusion MRI data can be incorporated into the continuum part of the model. This integration enhances clinical understanding and treatment strategies for managing cardiovascular disease. |
