Fluid Mechanic Assessment of the Total Cavopulmonary Connection Using Magnetic Resonance Phase Velocity Mapping and Digital Particle Image Velocimetry

Ann E. Ensley, Georgia Tech
Agnes Ramuzat, Georgia Tech
Timothy M. Healy, Georgia Tech
George P. Chatzimavroudis, Cleveland State University
Carol Lucas, University of North Carolina
Shiva Sharma, Emory University
Roderic I. Pettigrew, Frederik Philips Magnetic Resonance Research Center
Ajit P. Yoganathan, Georgia Institute of Technology and Emory University


The total cavopulmonary connection (TCPC) is currently the most promising modification of the Fontan surgical repair for single ventricle congenital heart disease. The TCPC involves a surgical connection of the superior and inferior vena cavae directly to the left and right pulmonary arteries, bypassing the right heart. In the univentricular system, the ventricle experiences a workload which may be reduced by optimizing the cavae-to-pulmonary anastomosis. The hypothesis of this study was that the energetic efficiency of the connection is a consequence of the fluid dynamics which develop as a function of connection geometry. Magnetic resonance phase velocity mapping (MRPVM) and digital particle image velocimetry (DPIV) were used to evaluate the flow patterns in vitro in three prototype glass models of the TCPC: flared zero offset, flared 14 mm offset, and straight 21 mm offset. The flow field velocity along the symmetry plane of each model was chosen to elucidate the fluid mechanics of the connection as a function of the connection geometry and pulmonary artery flow split. The steady flow experiments were conducted at a physiologic cardiac output (4 L/min) over three left/right pulmonary flow splits (70/30, 50/50, and 30/70) while keeping the superior/inferior vena cavae flow ratio constant at 40/60. MRPVM, a noninvasive clinical technique for measuring flow field velocities, was compared to DPIV, an established in vitro fluid mechanic technique. A comparison between the results from both techniques showed agreement of large scale flow features, despite some discrepancies in the detailed flow fields. The absence of caval offset in the flared zero offset model resulted in significant caval flow collision at the connection site. In contrast, offsetting the cavae reduced the flow interaction and caused a vortex-like low velocity region between the caval inlets as well as flow disturbance in the pulmonary artery with the least total flow. A positive correlation was also found between the direct caval flow collision and increased power losses. MRPVM was able to elucidate these important fluid flow features, which may be important in future modifications in TCPC surgical designs. Using MRPVM, two- and three-directional velocity fields in the TCPC could be quantified. Because of this, MRPVM has the potential to provide accurate velocity information clinically and, thus, to become the in vivo tool for TCPC patient physiological/functional assessment. © 2000 Biomedical Engineering Society.