High-fidelity Cardiovascular Modeling from Super-Fast Magnetic Resonance Imaging

ended

Biophysical computational models of the cardiovascular system need to be adapted to each particular patient from clinical data. The state-of-the-art imaging method for assessing cardiovascular diseases is Magnetic Resonance Imaging (MRI), which is hence the preferred source of data for the personalization of the models. However, MRI is still not able to reliably image the kinematics of thin structures like cardiac valves and the arterial wall. Moreover, MRI measurements of the 3D kinematics of the heart is a challenging task. These restrictions hamper the clinical translation of patient-specific modeling. Therefore, a new paradigm for model personalization is urgently needed. The ambition of CardioZoom is to propose novel methods for biophysical parameter estimation in computational models of the heart, large vessels and valves using MRI data acquired in very short scan times. The approach will be based on the deep integration imaging and biophysical principles, relaxing the constraints of standard cardiovascular imaging implying long MRI scans. Extensive validations using experimental (phantom) data will be performed and tests on volunteer and patients data are planned. The findings of CardioZoom will allow obtaining clinically feasible, detailed characterizations of the cardiovascular system.