Session: 05-01-04 Applied CFD

Paper Number: 65886

Start Time: August 11th, 12:50 PM

65886 - A Comprehensive Review of 4D Flow MRI and CFD in Cardiovascular and Congenital Heart Disease 

The population of adults with cardiovascular and congenital heart disease (CHD) in the United States is rising. Many patients in this population have complex anatomical vasculature with corresponding abnormal blood flow that progressively affects the heart and great vessels. Assessment of blood flow patterns using advanced imaging techniques such as four-dimensional (4D) flow magnetic resonance imaging (MRI) and echocardiography-based vector flow mapping have paved a path towards greater understanding of the patterns and implications of flow alterations in complex and changing vasculature. On the other hand, computational fluid dynamics (CFD) continues to evolve as a numerical computational method to solve and analyze complex blood flow patterns using parameters such as wall shear stress and energy loss, and offers promise for predicting potential therapeutic intervention in the future. Both these fields have evolved independently over the last few decades leveraging advances in imaging technologies and computational methods.

The primary focus of this review is to synthesize findings from literature on 4D flow MRI and CFD as related to blood flow patterns in the heart and great vessels in cardiovascular and congenital heart disease. Furthermore, the paper will describe strengths and limitations of individual techniques and future directions for the interplay of 4D flow MRI and CFD to better compute blood flow patterns and predict individual surgical or interventional treatment scenarios in these patients.  

Prior studies have demonstrated the potential of blood flow analysis, using both 4D flow MRI and CFD, to improve our understanding of the natural progression of cardiac decompensation in abnormal hearts; offer quantifiable metrics for disease progression; allow monitoring of a patient’s disease state over time in an effort to ward off sudden clinical presentation; aid in determination of functional post-surgical outcomes in complex CHD; avoid the risks of standard interventional procedures; evaluate prostheses, such as aortic grafts, in the cardiovascular domain; and, most notably, enable “virtual surgery”, or surgical planning to model and predict post-treatment hemodynamics toward comparison and selection of the optimal surgical technique.

Integration of 4D flow MRI and CFD may allow for more accurate computation and depiction of flow patterns within human vasculature and ultimately improve the tools and methodologies used in analysis, simulation and prediction of cardiovascular hemodynamics. 4D flow MRI is still limited in various factors such as longer image acquisition time, lower resolution, image processing/image segmentation, lower signal-to-noise ratio (e.g., VENC – velocity sensitivity encoding parameter), signal void in the presence of metallic stents, etc. However, CFD methods for computing and visualizing flow is limited by “assumptions” made in the geometry derived from patients’ prior imaging studies such as rigid wall, input boundary conditions, patients’ physiological conditions, etc.

Current efforts to correlate flow visualization with high-fidelity anatomical datasets involve manual orientation of image-based three-dimensional (3D) reconstruction of patient anatomy alongside 4D flow MRI. However, we note the potential of CFD to augment this process through automatized correlation of 3D geometry and 4D flow streamlines. We suggest the use of both CFD and 4D MRI in conjunction, i.e., the use of 4D MRI parameters as input boundary conditions for CFD, allowing automatic overlay of 3D anatomy and 4D flow streamlines with decreased expending of time, resources, and technician expertise in manual correlation. Thus, while past literature has focused on the potential of CFD to co-validate or reflect flow patterns in 4D MRI, we propose that efforts shift to combining the potential of 4D MRI and CFD, i.e., using baseline parameters from pre-procedural 4D MRI for patient-specific CFD, toward the ultimate goal of simulation of post-procedural hemodynamics to better predict post-procedural scenarios and thus infer optimal interventional techniques.

Presenting Author: Lamees I. El Nihum Texas A&M University

Authors:

Lamees I. El Nihum Texas A&M University
Ponraj Chinnadurai Siemens Medical Solutions USA Inc.
C. Huie Lin Houston Methodist DeBakey Heart & Vascular Center ; Weill Cornell Medical College
Debjyoti Banerjee Texas A&M University