You are on page 1of 5

Computational Aerodynamics of Long Segment

Congenital Tracheal Stenosis with Bridging Bronchus


Limin Zhu1#, Jinlong Liu1,2#, Weimin Zhang1, Qi Sun1,
Haifa Hong1, Zhou Du1, Jinfen Liu1*
1

Department of Cardiothoracic Surgery


2
Institute of Pediatric Translational Medicine
Shanghai Childrens Medical Center
Shanghai Jiao Tong University School of Medicine
Shanghai, China

Qian Wang
Department of Medical Imaging
Shanghai Childrens Medical Center
Shanghai Jiao Tong University School of Medicine
Shanghai, China

Yi Qian

Mitsuo Umezu

Australian School of Advanced Medicine


Macquarie University
Sydney, Australia

Center for Advanced Biomedical Sciences


TWIns, Waseda University
Tokyo, Japan

AbstractLong segment congenital tracheal stenosis (LSCTS)


is one of most severe malformation with high mortality rate and
dismal prospective. Hypoventilation is the main issue for the
death of children with LSCTS after surgical correction. However,
currently, little information is available on local aerodynamics to
disclose the reasons for the improvement of such therapies,
especially LSCTS with distal bronchus stenoses. Here, we
investigated a patient-specific model of LSCTS with complete
tracheal rings and bridging bronchus (BB). Computational fluid
dynamics (CFD) was applied to analyze the local aerodynamics
around BB before and after tracheal surgery in inspiratory phase
and expiratory phase. Average pressure drop, wall shear stress,
streamlines and energy loss were calculated to evaluate the
surgical outcomes. The results indicate the airflow at the trachea
and BB become more turbulent in expiration phase than that in
inspiration phase. The turbulence increases the workloads of
respiration in expiration phase. It should be the cause for postoperative hypoventilation. To study the local aerodynamics is
helpful for the improvement of surgical therapies of the LSCTS.
Keywordscongenital tracheal stenosis; computational fluid
dynamics; bridging bronchus; airflow; aerodynamics

I. INTRODUCTION
The left pulmonary arterial sling (PAS) is usually
associated with diffuse primary or secondary tracheobronchial
abnormalities. It is one of most common congenital tracheal
stenosis (CTS) presenting of complete tracheal rings and
bridging bronchus (BB). The presentation of these anomalies
ranges from no symptoms to severe respiratory distress,
recurrent respiratory tract infections and dysphagia. Although
the slide tracheoplasty (STP) has been reported as a versatile
and reliable technique associated with low morbidity and
mortality recently, the most challenging form of CTS is long
segment congenital tracheal stenosis (LSCTS) with
compromise of the carina and main stem bronchi, sometimes
result in high mortality in neonates and young infants.
*Research supported by the Project funded by China Postdoctoral Science
Foundation (No. 2014T70420, P.I.: Jinlong Liu), the Fund of The Shanghai
Committee of Science and Technology (No.14411968900, P.I.: Jinlong Liu),
the Medicine-Engineering Project of Shanghai Jiao Tong University (No.
YG2014MS63, P.I.: Jinlong Liu) and the Fund of Shanghai Jiao Tong
University School of Medicine (No. 14XJ10039, P.I.: Jinlong Liu).
Jinfen Liu is the corresponding author. Phone: +86-21-5881-5377; Fax:
+86- 21-5089-1405; E-mail: liujinfen2007@aliyun.com
# Co-first Author: Limin Zhu and Jinlong Liu
,(((

A patient-specific investigation based on the analysis of


tracheal aerodynamics was performed. The three-dimensional
(3D) tracheal models before and after surgical modified STP
were reconstructed according to medical images of a patient
who suffered from PAS associated with LSCTS, BB and a
distal bronchial stenosis. Computational fluid dynamics (CFD)
was applied to evaluate the tracheal aerodynamics around the
BB at the inspiration phase and expiration phase. Average
pressure drop, wall shear stress (WSS), streamlines and energy
loss were calculated to quantitatively estimate the local
aerodynamics in trachea and bronchus. The aim of the present
study is not to investigate the aerodynamic characteristics of
STP in the complex CTS, but also to find the reasons for postoperative hypoventilation on such patients.
II. MATERIAL AND METHODS
A. Generation of Geometric Models
This study was approved by the local institutional review
board and regional research ethics committee of Shanghai
Childrens Medical Center (SCMC) Affiliated Shanghai Jiao
Tong University School of Medicine. The informed consent
was obtained from the parents of a 15-month-old child, when
he was diagnosed PAS associated with LSCTS with complete
tracheal rings, BB with distal part stenosis and a distal part of
right upper bronchial stenosis.
The patient-specific computed tomography (CT) images
before and after operations were obtained for the reconstruction
of the 3D tracheal models by using 16-slice multi-detector row
enhanced CT scanner (Bright Speed Elite, GE Medical System,
General Electric, America). Image resolution was 512512
pixels and the slice thickness was 0.625 mm. Medical
imagining software Materialise-Mimics Innovation Suite 17.0
was used to segment CT images and reconstruct the 3D
tracheal geometry. Figure 1 depicts the tracheal geometry after
surface refinement.

We found the maximum Reynold value was about 3500.


The airflow in this patient-specific tracheal geometry should be
turbulence flow. To solve the problem of turbulence flow, we
used the Wilcox model k [3], which was validated
perfectly for the complex airflow in trachea [4]. The Wilcox
k model can be described by the equations of the kinetic
energy of turbulence (Eq. (3)) and the specific dissipation rate
(Eq. (4)) defined below:
Eddy Viscosity

T =

(2)

Turbulence Kinetic Energy


(a) Pre-operation

k
k
u

k (3)
*
+ u j
= ij i * k +
+ T

t
x j
x j
x j
x j

Specific Dissipation Rate

u
(4)

= ij i 2 +
+ u j
( + T )

x j
k x j
x j
x j
t
Closure Coefficients

9 ,
5,
1
1
3
= , * =
= , * =
40
9
2
2
100

where ij is the Reynolds stress tensor. It is given by,


(b) Post-operation

2
3

ij = 2 T S ij k ij

Figure 1. 3D patient-specific tracheal models before and after surgical


corrections.

In Model 1, there were three stenoses in patients trachea.


Due to the difficulty of surgeries, the main stenosis, LSCTS, in
trachea was surgically corrected. However, the other two
stenoses were uncorrected; Stenosis 1 and Stenosis 2 in Model
2.
B. CFD Analysis
1) Governing equations of flow
The 3D incompressible Navier-Stokes (N-S) equation and
continuity equation governs the airflow in trachea. The motion
of airflow can be described by the following equations defined
below.

( u i ) +
( ui u j ) = p +
t
x j
xi x j

+
( u j ) = 0
t x j

u i u j
+

x j xi

+ f i

(1)

where i , j = 1, 2 , 3 , x1 , x2 , and x3 represent coordinate axes,


ui , u j and p are the velocity vector and the pressure in the

point of the fluid domain, and are airflow density and


viscosity, t is time. The term f i expresses the action of body
forces. We assumed airflow to be a Newtonian fluid with
constant density ( = 1.161 kg/m3) [1] and viscosity (
=1.86410-5 kg/m s) [2] and the body forces were omitted.

(5)

1, if i = j
where Sij is the mean strain-rate tensor, i j =

, is
0, if i j
Kronecker delta.
Wall shear stress (WSS), which is difficult to be acquired
by direct measurements, is accepted as one reason for the
damage to the trachea [5, 6]. It is a manifestation of the
interaction between the airflow and trachea. The equation for
WSS in a Newtonian fluid is given by:

wall =

u x
y

(6)
y =0

where is viscosity, u x is the velocity of the fluid near the


boundary, and y is the height above the boundary.
Energy loss (EL), the energy difference between the
tracheal inlet and the outlet, is useful for evaluating inspiration
workload. Given pressure and flow rate of the inflow and
outflow, EL can be calculated by:

EL = E inlet E outlet
=

P + 2 u
i

inlet

2
i

2
Q i Po + u o Q o
2

outlet

(7)

where P is the static pressure, u is the velocity, and Q is the


flow rate. i indicates the inlet. o indicates the outlet.
2) Mesh generation
The grid-generation software, ANSYS -ICEM 14.0, was
applied to produce mixed grids. Three-layer body-fitted
prismatic grids were created in the near-wall regions with an
average nodal space that increased by a ratio of 1.2. The
distance of the first prismatic layer to the tracheal surface was
fixed at 0.0024 mm. This scheme accurately measured WSS
and improved the resolution of the relevant scales in airflow
motion. A tetrahedral mesh covered the remainder of the
domain.
To find the best mesh for CFD analysis, grid-sensitivity
verification were performed and found that grid numbers of
about 0.5 million in steady simulation would make the most
efficient mesh. Table 1 lists the mesh information for each
model used in the present study.

4) Calculations
The finite volume solver package, ANSYS-FLUENT 14.0,
was applied to solve the problem of unsteady airflow in each
model. We assumed tracheal wall consisted of rigid surfaces
with no-slip conditions. The semi-implicit (SIMPLE) method
was chosen to solve the discretized 3D incompressible N-S
equations. A second-order upwind scheme was used in the
calculations. For convergence criteria, the relative variation of
the quantities between two successive iterations was smaller
than the pre-assigned maximum, 10-5.
III. RESULTS
Figure 3 compared the average pressure drop of Stenosis 1
and Stenosis 2 at the phase of inspiration and expiration. An
obvious decrease was obtained at Stenosis 1 after the surgical
correction of main stenosis in trachea. However, an increase of
average pressure drop was found at Stenosis 2. This indicates
the surgical correction has a large effect on the BB.

TABLE I. MESH INFORMATION FOR EACH MODEL


MODEL

MODEL 1

MODEL 2

TOTAL ELEMENTS

793,126

786,573

TOTAL NODES

240,192

262,620

3) Boundary conditions
The pulsatile velocity profile of airflow in one respiration
cycle was used as inlet condition for the simulation of airflow
in the patient-specific model; Model 1. To compare the
geometric effects caused by LSCTS on stenoses, we imposed
the same velocity curve as inlet condition of Model 2. Figure 2
shows these data in one respiration cycle.
(a) Average pressure drop at Stenosis 1

Figure 2. Airflow velocity in one respiration cycle

To fully develop the flow boundary layer, we extended the


inlet domain upstream to twenty times the size of the trachea.
At the outlets, we extended airway diameter forty times in a
normal direction to allow sufficient recovery of air pressure in
each branch. A zero pressure was assumed at all the outlets
according to Ho CY et al. [2].

(B) Average pressure drop at Stenosis 2


Figure 3. The comparison of average pressure drop of Stenosis 1 and
Stenosis 2 at the phase of inspiration and expiration.

Local aerodynamics was investigated. Table 2 displays the


WSS and streamlines at the maximum velocity of airflow at the
inspiration phase and expiration phase. High WSS of trachea
was relieved after the surgical correction. It indicates the

damage of the airflow to the trachea was redduced. The air can
be flow smoothly. However, a relatively large still can be
found at Stenosis 2, especially at the phase off expiration.
Flow separation occurred around Stenosis 2 before surgical
correction. Large rotating flow was created at the connection
area of BB to trachea. This implied an abnormal
a
pressure
gradient may create to increasing flow rotaation especially at
expiration. By the surgical correction of thee main stenosis in
trachea, the rotating flow was weakened around
a
Stenosis 2.
However, the airflow in the trachea became revolved
r
especially
at expiration.

Figure 4 graphs the average EL at the phase of inspiration


and expiration. After the corrrection of the main stenosis, a
sharp decrease of EL was achiieved. It means the workload of
breath was reduced.

TABLE II. AERODYNAMIC ANALYSIS (WSS AND


D STREAMLINES)
INSPIRATION

EXPIRATION

Figure 4. The comparison of Averaage EL at the phase of inspiration and


expiration.

IV. DISC
CUSSION
With the development of thhe technique of STP, the survival
of CTS has improved recently [7-10]. However, the surgery in
these children can be challengged due to their heterogeneous
presentation. The presence of
o bronchial stenosis is still a
significant risk factor for deathh [7, 10]. Based on the clinical
experience in our medical center, the operative survival rate of
TS with distal bronchia stenosis,
STP in the complex case of CT
which is unfeasible to operationn, was significantly low, because
the complex cases with severre distal bronchial stenosis can
cause severe hypercapnia durinng the early postoperative period
after STP.
The results of present studdy disclosed local aerodynamics
after the surgical correction of main
m stenosis in the trachea. The
average pressure drop in Stennosis 1 and Stenosis 2 shows
different results. The reason of the increasing of the pressure
drop in Stenosis 2 after surgerry should be the release of flow
restriction and flow redistributiion to the distal part of the main
airway after main stenosis enllargement. The results of WSS
and streamlines indicate that the turbulence of airflow have
great effects in the expiratory phase, which will result in the
significant restriction of airfloow from the distal parts of the
stenosis, CO2 retention and hypercapnia.
h
According to our
clinical experience, this phenoomenon will be aggravated and
become an obvious problem during the early postoperative
period in gas exchange, especcially in the passive expiratory
phase during positive pressuree mechanical ventilation. To the
patient with residual main distal
d
bronchial stenosis, such
expiratory limitation will be thee reason of failure of ventilation,
respiratory acidosis and needd of extracorporeal membrane
oxygenation. It seemed that thee flow turbulence should be the
major issue for immediately reecovery after STP and long-term
life quality.

In clinic, it will be confused for the decision-making of


surgical strategy of complex CTS with distal bronchus stenosis
because of the unpredictable outcome for these children. The
calculated results are reasonable to explain the clinical
outcomes and disclose the reasons why several children died
from severe hypoventilation after STP. Recent methods of
diagnosis for CTS, including bronchoscopy, bronchography
and multidetector computed tomography (MDCT) [11-13], can
only provide the static imagine modalities of the main airway.
The dynamic information is unavailable, especially in the
diagnosis by bronchography and MDCT. Therefore, to predict
the outcome for surgical intervention of the stenotic bronchus
remained still unclearly. The technique of CFD will be a
promising tool for the evaluation of the severe abnormalities at
trachea and bronchus by providing local aerodynamics, such as
pressure drops, WSS, airflow distribution and energy loss rate.
Combined with computer-aided design (CAD), CFD can be
potentially used for the investigation of virtual surgeries [1416]. Based on the results of virtual surgeries, the possible
results of the aero flow dynamic after surgical repair will be
evaluated before surgical procedure. The patient-specific
enlargement of trachea lumen will be studied in future, from
which the boundary diameter of the main airway lumen
surgical treatment will be indicated In the case of complex CTS
with main distal bronchial stenosis, an additional therapy for
the stenotic bronchus, such as balloon dialation and/or stent
implantation, should be applied more initiatively. In the case of
high risk of sever residual stenosis of the distal branches of the
airway, contingency plans such as ECMO should be prepared
before surgical procedure performed. It can be performed
depending on the information of patients status and local
aerodynamics provided by CFD prediction.

REFERENCES
[1]

[2]

[3]

[4]

[5]
[6]

[7]

[8]

[9]

[10]

[11]

V. CONCLUSION
The approach of computational aerodynamic analysis by
using CFD technique is a potential noninvasive tool to provide
airflow information in trachea for the evaluation of surgical
outcomes. Different bronchial stenosis has influence on airflow
dynamics after surgical corrections, especially at expiration
phase. The computational aerodynamics can be served as a
helpful tool to disclose local detailed airflow information for
further surgical therapies.

[12]

[13]

[14]

ACKNOWLEDGMENT
We would like to express our great gratitude for the support
of the Project funded by China Postdoctoral Science
Foundation (No.2014T70420, P.I.:Jinlong Liu), the Fund of
The Shanghai Committee of Science and Technology
(No.14411968900, P.I.:Jinlong Liu), the Medicine-Engineering
Project of Shanghai Jiao Tong University (No. YG2014MS63,
P.I.:Jinlong Liu) and the Fund of Shanghai Jiao Tong
University School of Medicine (No. 14XJ10039, P.I.:Jinlong
Liu).

[15]

[16]

O Mimouni-Benabu, L Meister, J Giordano, P Fayoux, N Loundon, et al.,


A preliminary study of computer assisted evaluation of congenital
tracheal stenoses: a new tool for surgical decision-making. Int J Pediatr
Otorhinolaryngol, 2012, 76: 1552-1557.
CY Ho, HM Liao, CY Tu, CY Huang, CM Shih, MY Su, JH Chen, TC
Shih, Numerical analysis of airflow alteration in central airways
following tracheobronchial stent placement. Experimental hematology &
oncology,2012, 1, 23.
DC Wilcox, Reassessment of the scale determining equation for
advanced turbulence models. Journal of american institute of aeronautics
and astronautics, 1988, 26 (11), 1299-1310.
G Mylavarapu, S Murugappan, M Mihaescu, M Kalra, S Khosla, E
Gutmark, Validation of computational fluid dynamics methodology used
for human upper airway flow simulations. Journal of biomechanics,
2009, 42, 1553-1559.
RS Cotran, T Collins, V Kumar, W Schmitt, 1999, Robbins Pathologic
Basis of Disease, 6th ed. Saunders, London.
R Chowdhary, V Singh, AE Tattersfield, SD Sharma, S Kar, AB Gupta,
Relationship of flow and cross-sectional area to frictional stress in
airway models of asthma. The Journal of asthma: official journal of the
Association for the Care of Asthma, 1999, 36, 419-426.
CR Butler, S Speggiorin, EM Rijnberg, DJ Roebuck, N Muthialu, RJ
Hewitt, MJ Elliott, Outcomes of slide tracheoplasty in 101 children: A
17-year single-center experience, J Thorac Cardiovasc Surg 2014,
147:1783-1790.
HC Grillo, CD Wright, GJ Vlahakes, TE MacGillivray, Management of
congenital tracheal stenosis by means of slide tracheoplasty or resection
and reconstruction, with long-term follow-up of growth after slide
tracheoplasty. J Thorac Cardiovasc Surg. 2002, 123:145-152.
B Xue, B Liang, S Wang, L Zhu, Z Lu, Z Xu, One-Stage surgical
correction of congenital tracheal stenosis complicated with congenital
heart disease in infants and young children, J Card Surg. 2014, 17.
JL Antn-Pacheco, JV Comas, C Luna, MI Benavent, M Lpez, V
Ramos, MD Mndez, Treatment strategies in the management of severe
complications following slide tracheoplasty in children, Eur J
Cardiothorac Surg. 2014, 46:280-285.
W Baden, J Schaefer, M Kumpf, N Tzaribachev, T Pantalitschka, A
Koitschev, G Ziemer, J Fuchs, M Hofbeck, Comparison of imaging
techniques in the diagnosis of bridging bronchus. Eur Respir J. 2008,
31(5):1125-1131.
YM Zhong, RB Jaffe, M Zhu, W Gao, AM Sun, Q Wang, CT
assessment of tracheobronchial anomaly in left pulmonary artery sling.
Pediatr Radiol. 2010, 40(11):1755-1762.
B Leonardi, A Secinaro, R Cutrera, S Albanese, M Trozzi, A
Franceschini, V Silvestri, R Tom, A Carotti, G Pongiglione, Imaging
modalities in children with vascular ring and pulmonary artery sling.
Pediatr Pulmonol. 2014, Jun 30.
JL Liu, Y Qian, K Itatani, A Murakami, R Shiurba, K Miyaji, T
Miyakoshi, M Umezu, Image-Based Computational Hemodynamics of
Distal Aortic Arch Recoarctation Following the Norwood Procedure,
Proceeding of The 4th International Congress on Image and Singal
Processing and 4th International Conference on BioMedical Engineering
and Informatics, 2011.10.
Q Sun, JL Liu, Y Qian, HB Zhang, Q Wang, YJ Sun, HF Hong, JF Liu,
Computational haemodynamic analysis of patient-specific virtual
operations for total cavopulmonary connection with dual superior venae
cavae, European Journal of Cardio-Thoracic Surgery, 2014, 45(3):564569.
JL Liu, Q Sun, M Umezu, Y Qian, HF Hong, Z Du, Q Wang, YJ Sun, JF
Liu, Influence of Conduit Angles on Hemodynamics of Modified
Blalock-Taussig Shunt: Computational Analysis of Patient-specific
Virtual Procedures, Life System Modeling and Simulation
Communications in Computer and Information Science, 2014, 461: 6271.

You might also like