ARUP; ISBN: 978-0-9562121-5-3 - CMBBE 2012 - Cardiff University

palliated with a surgical procedure, called ‘Norwood’ [1], reconstructing the hypoplastic
aortic arch and placing a Goretex conduit (i.e. shunt) between a systemic artery or the
right ventricle and the pulmonary arteries. Among the various shunt connections that
can be surgically performed, one option is the modified Blalock-Taussig (mBT) shunt
[2], whose proximal connection generates from the innominate artery. As all the energy
for the system is provided by a single ventricle, the Norwood procedure creates a
parallel arrangement of systemic and pulmonary circulations at the shunt level, enabling
unobstructed perfusion of the former and proper blood oxygenation through lungs.
Few months after palliation, Norwood patients can develop a coarctation (AC) of the
neoaorta, a narrowing of the proximal descending tract which is repaired either with
surgical or endovascular techniques. When AC is present, univentricular circulation is
highly dependent on the hemodynamic behavior of both the shunt and the coarctation.
Literature reports several experimental and/or computational studies on the Norwood
procedure [3-5] and AC [6,7]. However, most of these works focused on stand-alone
three-dimensional (3D) models of idealized or patient-specific geometries. The most
recent studies adopted a multiscale (or multidomain) approach, that couples a 3D
representation of the surgical area to a lumped parameter network (LPN) of the
remaining circulation [5,8]. This way, both local (shunt and AC) and peripheral
(pulmonary and systemic) impedances can be properly described to explore a unique
circulation, such that presented by a Norwood patient with concomitant AC. Not only
downstream flow distribution and pressure tracings, but also local hemodynamic
information within the surgical region can be extracted.
In modeling a circulatory system or device, in-silico results are usually validated against
experimental data [9,10]. Although the latter are considered more reliable, for a
mathematical model a complex physiological behavior may be easily described through
an equation (e.g. the non-linear pressure drop-flow relationship of a hydraulic tap),
whereas a physical component has to be constructed in a mock-loop simulator. Relying
on the respective advantages, a more effective combination of the in-vitro and in-silico
approaches may be achieved, i.e. matching of the results can provide mutual validation
while discrepancies help understand the processes, sometimes occult, that are involved
in the simulations.
The present study combines computational and experimental modeling of a patientspecific
Norwood anatomy with mBT shunt and concomitant AC, using a multiscale
representation of the univentricular circulation. By comparing simulations results, this
work aims to allow a mutual validation of the two methodologies, when matching each
other, and to identify the hemodynamic effects that may be responsible for
disagreement.
3. MATERIALS AND METHODS
3.1 Patient-specific anatomical reconstruction
A patient with HLHS, who underwent a Norwood surgery with mBT shunt, was
selected for this study. Clinical data were collected from magnetic resonance (MR) and
cardiac catheterization performed 3 months after the operation, detecting the presence of
a coarctation of the neoaorta. The coarctation index, as derived from the ratio of the
isthmus and descending aorta diameters, was 0.5. The use of the imaging data for
research purposes was approved by the local Research Ethics Committee. A patientspecific
anatomical model of the aortic arch was reconstructed from the available MR