10th IFAC Symposium on Biological and Medical Systems

10th
10th IFAC
IFAC Symposium
São Paulo,Symposium on
on Biological
Biological
Brazil, September and
and Medical
3-5, 2018 Medical Systems
Systems
10th
São IFAC Symposium
São Paulo,
Paulo, Brazil, on Biological
Brazil, September
September 3-5, Available
and
3-5, 2018
2018 Medicalonline at [Link]
Systems
10th Paulo,
São IFAC Symposium on Biological
Brazil, September and Medical Systems
3-5, 2018
São Paulo, Brazil, September 3-5, 2018
ScienceDirect
IFAC PapersOnLine 51-27 (2018) 299–304
Specific
Specific compliance:
compliance: is is it
it truly
truly independent
independent of of lung
lung volume?
volume?
Specific compliance: is it truly independent of lung volume?
Specific Oliver
compliance: is itJennifer
Kannangara*, truly [Link] of lung
Dickson**, J. Geoffrey volume?
Chase**
Oliver
Oliver Kannangara*,
Kannangara*, Jennifer Jennifer L. L. Dickson**,
Dickson**, J. J. Geoffrey
Geoffrey Chase** Chase**
Oliver Kannangara*, Jennifer L. Dickson**, J. Geoffrey Chase**
Oliver Kannangara*,
*University of Auckland, School of Medicine, Auckland, Jennifer  L. Dickson**, J. Geoffrey
New Zealand (e-mail: Chase**
ollie389@[Link])
*University
*University of
of Auckland,
Auckland, School
School of
of Medicine,
Medicine, Auckland,
Auckland,
 New
New Zealand
Zealand (e-mail:
(e-mail: ollie389@[Link])
ollie389@[Link])
**University
*University of
of Auckland, Canterbury,
School of Dept of
Medicine, Mechanical
Auckland, Eng,New Christchurch,
Zealand (e-mail: New Zealand (e-mail:
ollie389@[Link])
**University
**University
*University of of
of
Auckland, Canterbury,
Canterbury,
School ofDept
Dept of
of
Medicine, Mechanical
Mechanical
Auckland, Eng,
Eng,New Christchurch,
Christchurch,
Zealand New
New
(e-mail: Zealand
Zealand (e-mail:
(e-mail:
ollie389@[Link])
[Link]@[Link];
**[Link]@[Link]; [Link]@[Link])
of Canterbury, Dept of [Link]@[Link])
Eng, Christchurch, New Zealand (e-mail:
[Link]@[Link];
**[Link]@[Link]; [Link]@[Link])
of Canterbury, Dept of [Link]@[Link])
Eng, Christchurch, New Zealand (e-mail:
[Link]@[Link]; [Link]@[Link])
Abstract: Specific pulmonary compliance is a widely used metric in neonatal mechanical ventilation,
Abstract:
Abstract: Specific
Specific pulmonary
pulmonary compliance
compliance is aa widely
isexpandwidely used
used metric in
in neonatal
metricnormalized neonatal mechanical
mechanical ventilation,
ventilation,
capturing the
Abstract: intrinsic
Specific ability ofcompliance
pulmonary the lung to is a widely withused pressure
metric in neonatal bymechanical
the functional residual
ventilation,
capturing
capturingofthe
Abstract: the intrinsic
intrinsic
Specific ability
ability of
pulmonary of the lung
the lung
compliance to expand
tofound
isexpand
a widely with
withused pressure
pressure
metric normalized
normalized
in neonatal by the functional
bymechanical
the functional residual
residual
ventilation,
capacity
capturing the
the lung.
intrinsic Previous
ability studies
of the have
lung to expand specific
with compliance
pressure to
normalizedbe similar
by for
the both neonates
functional and
residual
capacity
capturing of
capacityHowever, the
ofthe lung.
theintrinsic
lung. Previous
Previous
ability studies
studies
of the have
have
lung tofound
found
expand specific
specific
with compliance
compliance
pressure to be
be similar
togrowth
normalized similar
by for both
forlung.
the both To neonates
neonates
functional and
and
residual
adults.
capacity of the lung. it is unclear what this implies about the structure and of the model the
adults.
adults. However,
capacity
factors However,
of the
affecting [Link] isPrevious
it is
specific
unclear
unclear
Previous
studies
what
what
studies
compliance,
this
thishave
have
a simple
foundabout
implies
implies
found model
specific
about
specific the
is
compliance
structure
thedeveloped
structure
compliance andto
and to
be similar
growth
growth
be
separating
of
of the
similar
forlung.
the
for
respiratory
both To
lung.
both
neonates
To model
modelinto
neonates
compliance
and
the
the
and
adults.
factors
factors However,
affecting
affecting it is unclear
specific
specific what this
compliance,
compliance, a
a impliesmodel
simple
simple about is
model thedeveloped
is structure separating
developed and growthrespiratory
separating of the lung.
respiratory To modelinto
compliance
compliance the
into
adults. However,
material-dependent
factors affecting specificit is
and unclear what this
volume-dependent
compliance, implies
a simple about isthedeveloped
components.
model Itstructure
provides and growthrespiratory
reasonable
separating ofestimates
the [Link]
modelinto the
material-dependent
material-dependent
factors
for affecting
experimental
and
and volume-dependent
specific
clinical volume-dependent
compliance,
data. The a
model simple components.
components.
model
demonstrates is It
It provides
provides
developed
specific
reasonable
reasonable
separating
compliance is estimatescompliance
estimates
respiratory
dependent
of compliance
of the
compliance
compliance
on into
alveolar
material-dependent
for
for experimental
experimental and volume-dependent
clinical
clinical data.
data. The model
The ratio,
modelseptal [Link]
demonstrates
demonstrates It provides
specific reasonable
compliance
compliance is estimates on
dependent
isratio,
dependent of the
on compliance
the alveolar
alveolar
material-dependent
tissue
for elastic modulus,
experimental and volume-dependent
clinical Poisson’s
data. The model components.
thickness
demonstrates It toprovides
specificalveolar reasonable
compliance radiusis estimates
dependent of the
and surface
on compliance
tension
alveolar
tissue
tissue
for elastic
elastic
experimental modulus,
modulus,
clinical Poisson’s
Poisson’s
data. The ratio,
ratio,
model septal
septal thickness
thickness
demonstrates to
to
specificalveolar
alveolar
compliance radius
radius is ratio,
ratio,
dependentand
and surface
surface
on the tension
tension
alveolar
effects.
tissue The
elastic model
modulus, also indicates specific
Poisson’sspecific compliance
ratio, septal thickness may not
to be entirely
alveolar independent
radius ratio, and to lung
surfacevolume
tension as
effects.
effects.
tissue The
The
elastic model
model
modulus, also
also indicates
indicates
Poisson’s specific
ratio, compliance
compliance
septal thickness may
may not
not
to be
be entirely
entirely
alveolar independent
independent
radius ratio, andto
to lung
lung
surfacevolume
volume
tension as
asa
the lung
effects. grows, due to surface tension effects varying with diameter, and the possibility of
the lungThe
the lung
effects.
varying The
model
grows,
grows,
model
alveolar
due
due also
also
wall to indicates
to surface
surface
indicates
thickness
specificeffects
tension
tension
specific
to
compliance
effects
compliance
diameter
varying
varying
ratio.
may
maywith
To
notalveolar
with
not
be entirely
alveolar
be entirely
experimentally
independent
diameter,
diameter,
independent and
and the
evaluate
to lung
the
to
trends
volume
possibility
possibility
lung volume
in ofas
of
specific
aa
as
the lung
varying
varying grows,
alveolar
alveolar due wall
wall to surface
thickness
thickness tension
to effects
diameter
to diameter varying
ratio.
ratio. To with
To alveolar
experimentally
experimentally diameter, and
evaluate
evaluate the
trends
trendspossibility
in of
specific
in specific a
the lung alveolar
compliance,
varying grows, datadue from
wall to surface
9 infants
thickness tension
and
to 8effects
adult varying
diameter mechanical
ratio. withexperimentally
To alveolar diameter,
ventilation patients and
evaluatewasthe possibility
recorded.
trends in Medianof a
specific
compliance,
compliance,
varying data
data from
alveolar from
wall 99 infants
infants and
and 88 adult
adult for mechanical
mechanical ventilation
ventilation patients
patients was
was recorded.
recorded. Median
Median
compliance
compliance, values
data from werethickness
0.51
90.51
infants to diameter
mL/cmH2O/kg
and 8 adult for ratio.
adultTopatients
mechanical experimentally
and 0.78
ventilation evaluatewas trends
mL/cmH2O/kg
patients recorded. in neonatal
for specific
Median
compliance
compliance
compliance, values
values were
were 0.51 mL/cmH2O/kg
mL/cmH2O/kg for adult
adult patients
patients and
and 0.78
0.78 mL/cmH2O/kg
mL/cmH2O/kg for
for neonatal
neonatal
patients
compliance =data
(p values from
0.098). 90.51
This
wereThis infants
result and 8 adult for
demonstrates
mL/cmH2O/kg mechanical
mechanically
adult ventilation
patients ventilated
and patients
0.78 was recorded.
neonates
mL/cmH2O/kg and for adults Median
have
neonatal
patients
patients (p
compliance =
= 0.098).
(p similar
values 0.098). were This
0.51 result
result demonstrates
demonstrates
mL/cmH2O/kg for mechanically
mechanically
adult patients ventilated
ventilated
and 0.78 neonates
neonates
mL/cmH2O/kg and
and adults
adults
for have
have
neonatal
statistically
patients (p similar
= 0.098). values
Thisofof specific
result compliance,
demonstrates in agreement
mechanically with
ventilated prior studies.
neonates Treating
and adults specific
have
statistically
statistically
patients (p assimilar
= a0.098). values
values
This of specific
specific compliance,
compliance, in
in agreement
agreement with
with prior
prior studies.
studies. Treating
Treating specific
specific
compliance
statistically
compliance
compliance similar
as
as
constant
aa constantvalues
constant of result
and
and
and
using
specific
using
using
ademonstrates
compliance,mechanically
least-squares
aa least-squares
least-squares
fit,
fit,
fit,
the
in the
the
trend ofventilated
agreement
trend
trend of
neonatal
of with
neonatal
neonatal
neonates
studies. and
compliance
priorcompliance
compliance Treatingadults
increasing
increasing
increasing
have
with
specific
with
with
statistically
body weight
compliance similar
can
as be
a constant values
used to of
and specific
predict typical
usingtypical compliance,
ranges
a least-squares of in
adult
fit, agreement
compliance
thecompliance with
trend of neonatal prior
values, studies.
indicating
compliance Treating
the specific
non-linearity
increasing with
body
body weight
weight
compliance can
can
as a be
be used
used
constant to
to predict
predict
and using typical
a ranges
ranges
least-squares of
of adult
adult
fit, the compliance
trend of values,
values,
neonatal indicating
indicating
compliance the
the non-linearity
non-linearity
increasing with
of compliance
body varying
weight canvarying be usedwith with volume
to predict is likely
typical ranges notofsignificant
adult compliancein reality, supporting
values, indicating thethe usenon-linearity
of specific
of
of
bodycompliance
compliance
weightascan varying
be usedwith with
to volume
volume is
is likely
likely not ofsignificant
notdistensibility.
significant in
in reality,
reality, supporting
supporting the
the use
use of specific
of for
specific
compliance
of compliance avarying
measure of predict
thevolume typical
underlyingis ranges
tissue
likely not adult compliance
significant However,
in reality,values,
the indicating
variation
supporting theinthe non-linearity
values
use of each
specific
compliance
compliance
of compliance as a measure
as avarying
measure of
with the
of the underlying
underlying
volume is tissue
tissue
likely not distensibility.
distensibility.
significant However,
However,
in reality, the variation
the variationthe
supporting in values
in use
values of for
for each
each
specific
cohort
compliance suggest that
as athatmeasure specific compliance,
of the underlyinglike like
tissuelung elastance,
distensibility. can vary
However, significantly,
the variation and and while
in values overall
cohort
trends
suggest
cohort hold,
compliance suggestas a that
centralmeasure specific
specific
values of compliance,
compliance,
the
are underlying
not necessarilylike
tissuelung
lung elastance,
elastance,
distensibility.
indicative.
can
can vary
varymodel
However,
Overall, the
significantly,
significantly,
the variation
provides and
in
a whilefor
while
values
unique for
each
overall
overall
each
analysis
cohort
trends
trends suggest
hold,
hold, that values
central
central specificare
values arecompliance,
not
not necessarily
necessarilylike lung
indicative.
[Link], can the
Overall,
Overall, varymodel
the significantly,
model provides and
provides whileanalysis
aa unique
unique overall
analysis
cohort
and
trends suggest
insight
hold, using that
central specificare
a simplified,
values compliance,
mechanics
not necessarilylike lung
relevant elastance,
model,
indicative. which can
Overall, varymodel
is could
the significantly,
be generalized
provides and
a whilestudies.
tounique
other overall
analysis
and
and insight
trendsinsight
hold, using
using
central aa simplified,
simplified,
values are mechanics
mechanics
not relevant
relevant
necessarily model,
model, which
indicative. which
Overall,is
is could
could
the be
be generalized
model generalized
provides ato other
other studies.
tounique studies.
analysis
© 2018,
and
Keywords:
and IFAC
insight
insight
using(International
Mechanical
using
a simplified, Federation
mechanics
Ventilation,
a simplified, mechanics of Automatic
Pulmonaryrelevant
relevant Control)
model,
Mechanics,
model,
whichHosting
Model,
which by Elsevier
is could
Compliance,
is could Ltd.
be generalized All rights
Elastance,
be generalized
to other
ICU,
to other reserved.
studies.
NICU.
studies.
Keywords: Mechanical
Keywords: Mechanical Ventilation,Ventilation, Pulmonary
Pulmonary Mechanics,
Mechanics, Model, Model, Compliance,
Compliance, Elastance,Elastance, ICU, ICU, NICU.
NICU.
Keywords: Mechanical Ventilation, Pulmonary Mechanics,  Model, Compliance, Elastance, ICU, NICU.
Keywords: Mechanical Ventilation, Pulmonary Mechanics,  Model, Compliance, Elastance, ICU, NICU.
intrinsic elasticity of the lung tissue, independent of lung
1. INTRODUCTION  intrinsic
intrinsic elasticity
elasticity of
of thethe lung
lung tissue,
tissue, independent of of lung
1. INTRODUCTION
1. INTRODUCTION  volume,
intrinsic and is similar
elasticity of thebetween
lung tissue, adultsindependent
and term neonates
independent
lung
of lung
1. INTRODUCTION volume,
volume,
intrinsic and
and is
is
elasticity similar
similar between
between
of However,
thebetween
lung specificadults
adults
tissue, and
and term
term neonates
neonates
Pulmonary compliance
Pulmonary compliance
and its inverse, elastance, are useful (Phelan
1. INTRODUCTION
and its inverse, elastance, are useful volume, et
(Phelan
(Phelan
al., is1969).
etand
et al., similar
al., is1969).
1969). However,
However, adultsindependent
specific
specific
compliance
and term neonates
compliance
compliance
of lung
can
can
be a
be
be aa
can what
Pulmonary
measures compliance
ofcompliance
respiratory and and
system its inverse, elastance,
[Link],
Compliance are useful
is the (Phelan volume,
confusing andconcept.
et concept.
al., 1969).similar
There between
is
However, little adults
in the
specific and
literature
complianceterm neonates
about
can what
be a
Pulmonary
measures
measures of respiratory system its inverse,
capacity. Compliance are useful
is the confusing
confusing concept. There
There is
is little
little in
in the
the literature
literature about
about what
Pulmonary
change
measures in of
of
respiratory
compliance
lung volume per
respiratory
system
and
system unit [Link],
its inverse,
increase
capacity. inCompliance
drivingare
Compliance
is the exact
useful
pressure.
is the
(Phelan
confusing
exact
et concept.
al., 1969).
mechanical
mechanical ThereHowever,
properties
properties is are
little
are
specific
in the compliance
encompassed
literature
encompassed
by can
about
by
be a
specific
what
specific
change
changeacute in lung
in of
lung volume
volume per per unit
unit increase
increase in driving
inCompliance
driving pressure.
pressure. exact
confusing mechanical
compliance. concept.
Further, properties
There
if is are
little
specific inencompassed
the
compliance literature
is the by
about
samespecific
what
from
measures
Both respiratory
respiratory system
distress capacity.
syndrome (ARDS) is the
and exact mechanical properties are encompassed
change
Both
Both in lungrespiratory
acute
acute volume perdistress
respiratory unit increase
distress syndrome
syndrome in driving(ARDS)
(ARDS) pressure.
and
and compliance.
compliance. Further,
Further, if
if specific
specific compliance
compliance is theby
the
isimply same
samespecific
from
from
change
pulmonary in lung volume
fibrosis haveperbeen unitassociated
increase inwith
driving pressure.
decreased in exact infants mechanical
compliance. to adults,
Further, properties
what if does this
specific are encompassed
similarity
compliance is the bysamespecific
about the
Both
pulmonary
pulmonary acute respiratory
fibrosis
fibrosis have
have distress
been
been syndrome
associated
associated with
with (ARDS)
decreased
decreased and in
in infants
infants
compliance.
growth
to
of
adults,
to the
adults,
Further,
lung,
what
what
or if
how
does
does
specific this
patients
similarity
thiscompliance
similarity
should be isimply
imply
the aboutfrom
about
same
ventilated?
the
the
from
Both
compliance
pulmonary acute respiratory
(Hickling,
fibrosis have 1998).distress
been syndrome
Pulmonary
associated with (ARDS)
compliancedecreased is and
alsoin infants
growth to
of adults,
the lung, what
or howdoes this
patients similarity
should be imply about
ventilated? the
compliance
compliance (Hickling,
(Hickling, 1998).
1998). Pulmonary
Pulmonary compliance
compliance is
is also
also growth
infants of
to the lung,what
adults, or how doespatients
this should beimply
similarity ventilated?
about the
pulmonary
used
compliance to fibrosis
guide
(Hickling, have
mechanical been associated
ventilation
1998).ventilation
Pulmonary (MV) with
(MV)
compliancedecreased
therapy is also in
for growth of the lung, or how patients should be ventilated?
used
used to
to guide
guide mechanical
mechanical ventilation (MV) therapy
therapy for
for This study
growth of thederives
lung, a how
or pure patients
mechanics model
should be for pulmonary
ventilated?
compliance
used patients
to with (Hickling,
guideARDS mechanical1998). Pulmonary
(Gattinoni et al., 1984). compliance is alsofor This This study
study derives
derives aa pure
pure mechanics
mechanics model
modelto for for pulmonary
pulmonary
used patients
patients
to with
with
guideARDS
ARDS (Gattinoniventilation
(Gattinoni
mechanical et
et al.,
al., 1984).
ventilation 1984). (MV)
(MV)
therapy
therapy for
compliance
This
compliance
and its relation
study derives
and its
to lung volume
a pure mechanics
relation to lung modelto
volume
relate
for specific
pulmonary
relate specific
patients with ARDS (Gattinoni et al., 1984). compliance
This study
compliance and
derives
to its relation
a pure
mechanical to lung
mechanics
properties volume
model
of to relate
for
the specific
pulmonary
lung. It is
However, compliance
patients with and elastance
ARDS (Gattinoni et al.,are both affected by lung compliance
1984). compliance and
to its relation toproperties
mechanical lung volume of to
therelate
lung. specific
It is
However,
However, compliance
compliance and
and elastance
elastance are
are both
both affected
affected by
by lung
lung compliance
compliance
validated to
and
using mechanical
its relation
clinical data to properties
lung
from volume
adults of
and the
to lung.
relate
neonates It the
specific
in is
volume,
However, ascompliance
best demonstrated
and in patients
elastance are who affected
both have undergoneby lung compliance
validated usingto clinical
mechanical
clinical data properties
from adults of
and the lung. in
neonates It the
in is
the
volume,
volume, as best demonstrated
ascompliance
best demonstrated in
in patients
patients who have
who affected
have undergone validated using data from adults and neonates
However,
avolume,
unilateral
as pneumonectomy.
best and elastance
demonstrated Although
in are both
patients their
who lung undergone
have tissue
undergone and compliance
by lung intensive
validated
intensive
care
using
care
tounitmechanical
(ICU data
clinical
unit (ICU and
properties
and NICU).
from
NICU). adults of
and the lung. inIt the
neonates is
aavolume,
unilateral
unilateral
structure as
may
pneumonectomy.
pneumonectomy.
best demonstrated
be completely
Although
Although
in patients
normal,
their
their
who
their
lung
have
total
tissue
tissue and
lungcompliance
undergone and intensive validated care
using unit (ICU data
clinical and NICU).
from adults and neonates in the
astructure
unilateral
structure may
may pneumonectomy.
completely Although
be completely
be normal, their
normal, theirtotal
their total tissue and intensive
lungcompliance
compliance care unit (ICU and NICU).
awill
unilateral
inevitably
structure may pneumonectomy.
be half
be normal Although
completely because their
normal, a their
single lung
total lung tissue
can only
compliance and intensive care unit (ICU2. 2.
and NICU).
METHODS
take
will
will inevitably
inevitably
structure
in may
half the
be
be
be half
half normal
normal
completely
volume as a
because
becauseset
normal,
complete
aa single
single
their
for total
the
lung
lung can
can
compliance
same
only
only
driving 2. METHODS
METHODS
take will inevitably
in half the be
half the half normal
volume becauseset
as aa complete
complete seta single thelung
for the samecan only
driving 2. METHODS
take
pressure. in
will inevitably
Likewise,volume
be half
infantsas
normal have because
muchset for
a single
lower same
lung
compliance candriving
only
than 2.1 Patient Data 2. METHODS
take
pressure.
pressure. in half the
Likewise,
Likewise,volumeinfants
infantsas a complete
have
have much
much lower
lowerfor the same
compliance
compliance driving
than
than
take in half the volume as
areahave complete set for the same driving 2.1 Patient Data
adults,
pressure.
adults,
adults,
asLikewise,
as
their
their
lungs
their lungs
asLikewise, infants
lungs are
are have
manymuch
many
many
times
times
timeslower smaller.
compliance
smaller.
smaller. than 2.1
Therefore,
Therefore,
Therefore,
Patient Data
2.1 Patient Data
pressure.
when
adults, comparing
as their lungs infants
mechanics are many of much
lungstimeslower
of compliance
different volumes,
smaller. volumes, than
Therefore, it 2.1 Patient pressure
Ventilator Data and flow data was sampled at 125 Hz
when
when comparing
comparing mechanics
mechanics of
of lungs
lungs of
of different
different volumes, it
it Ventilator pressure and
adults,
is
when as
common their
comparing to lungs
divide
mechanics are
the many
of lungstimes
compliance of smaller.
by
different the Therefore,
functional
volumes, it Ventilator
from SLE pressure
500 and flow
ventilators flow
(SLE,
data
dataUK) was
wasfor sampled
sampled
10 infants
at
at 125
125 Hz
Hz
in the
is
is
when common
common
comparing to divide
to divide
mechanics the
theobtain
ofcompliance
compliance
lungs of by the functional
by the volumes,
different functional it Ventilator
from SLE pressure
500 and
ventilators flow
(SLE, data UK)was sampled
for 10 at
infants 125
in Hz
the
residual
is commoncapacity
to (FRC)
divide to
the specific
compliance compliance.
by the functional from
NICU SLE
Ventilatorat 500
pressure
Christchurchventilators
and Women’s(SLE,
flow data UK)was
Hospital, for 10
sampled infants
Christchurch,at in
125 the
Hz
New
residual
residual capacity
capacity (FRC) to
(FRC) totheobtainobtain specific compliance.
specific compliance. from
NICU SLE
at 500 ventilators (SLE, UK) for 10 infants in the
is
residual common to divide
capacity (FRC) to obtaincompliance
specific by the functional from
compliance. NICU
Zealand. at Christchurch
SLE Christchurch
500
Ethics ventilators
approval
Women’s
Women’s(SLE,
was
Hospital,
Hospital,
grantedUK) byfor Christchurch,
Christchurch,
the 10 NZinfants
NortherninNew
New
the
B
Specific
residual compliance
capacity (FRC) is topurported
obtain specific a measure of the NICU
to be compliance. Zealand.
Zealand. at Christchurch
Ethics
Ethics approval
approval Women’s
was
was grantedHospital,
granted by
by theChristchurch,
the NZ
NZ Northern
Northern New B
B
Specific
Specific compliance
compliance is is purported
purported to to bebe aa measure
measure of the NICU
of the at Christchurch Women’s Hospital,
Zealand. Ethics approval was granted by the NZ Northern B Christchurch, New
Specific compliance is purported to be a measure of the Zealand. Ethics approval was granted by the NZ Northern B
Specific ©
2405-8963
Copyright 2018,
compliance
© IFAC (International
2018 IFAC is purportedFederation to be a ofmeasureAutomatic ofControl)
the 299Hosting by Elsevier Ltd. All rights reserved.
Copyright
Peer review©
Copyright 2018
©under IFAC
2018 responsibility
IFAC 299
of International Federation of Automatic
299Control.
Copyright © 2018 IFAC
10.1016/[Link].2018.11.625 299
Copyright © 2018 IFAC 299
IFAC BMS 2018
São Paulo, Brazil, September 3-5, 2018 Oliver Kannangara et al. / IFAC PapersOnLine 51-27 (2018) 299–304
300

Ethics Committee. No interventions were made in this study. Second, plane stress generalised Hooke’s Law links alveolar
Infants were invasively ventilated, and expected to remain on stresses and strains, :
invasive MV for 24-48 hours. Patients range from 25-41
weeks gestational age and 0.77-3.4 kg birth weight.
Data was sampled at 100 Hz from Puritan Bennett 840
ventilators for 8 adult patients in the Christchurch Hospital
ICU, during a pilot trial of the CURE protocol (Szlavecz et
(4)
al., 2014). Patients received synchronous intermittent
mandatory ventilation (SIMV). Ethics approval was granted Where subscripts xx and yy indicate stresses and strains in the
by the NZ Upper South Island Regional A Committee. x and y directions respectively,  and  are shear stress and
Patients were 20-80 years and weighted 48-146 kg. strain, G is the shear modulus, and  is Poisson’s ratio.

2.2 Single Compartment Model If the alveoli undergoes uniform expansion, Equation 4 can
be simplified to obtain an expression for alveolar wall stress
A single compartment lung model was fit for each inspiratory in terms of the circumferential strain, defined:
pressure (Paw) and flow (Q) waveforms to obtain patient-
specific elastance (Ers) and resistance (R) values (Chiew et (5)
al., 2011, Chiew et al., 2015). The model is defined:
Third, initial and final alveolar radii, R0 and Rf, are related to
the circumferential strain by:
(1)
Where PPEEP is the imposed positive end expiratory pressure, (6)
and PETT accounts for the resistance of the endotracheal tube
(ETT) in neonates, separating pressure drop due to ETT Stress, strain and alveolar pressure are 0 at initial radius, R0.
resistance from pressure drop due to neonatal airway Fourth, the equation for alveolar compliance, C, is defined in
resistance, and is calculated (Jarreau et al., 1999): terms of alveolar pressure and volume:

(2) (7)
Where L is the length of the ET tube in cm, D is the diameter Combining Equations 3, 5, 6, and 7 and solving for C yields a
in mm, and Q is the flow rate in mL/s. This term does not spherical linear-elastic alveolar compliance, Calv:
significantly affect identified elastance values, as it trades off
with the resistance value. It is not included in the adult
model, due to its laminar flow assumption being violated at
endotracheal tube diameters greater than 4mm. (8)

Pressure and flow data were used in their raw form with no Equation (8) is the compliance of an individual alveolus. The
compliance of the entire lung, Clung, = Calv*n. Thickness can
filtering or smoothing. Cases where the single compartment
lung model resulted in poor fit, or where identified elastance also be replaced by a dimensionless thickness to alveolar
values were zero, were removed from the analysis. Poor fit radius ratio, tr=t/R0, multiplied by alveolar radius, yielding:
was defined as a modelled pressure error over 15% compared
to actual pressure, leading to a removal of 99,860 breaths of
2,298,257 (4.3%). All data processing was carried out using (9)
MATLAB (r2016b, MathWorks, Natick, USA).
Equation 9 comprises a lung-volume dependent component
based on n, Rf, R0, and a material dependent component a
2.3 Spherical linear-elastic alveolar model function of , alveolar modulus, Y, and tR.

To separate respiratory elastance into material-dependent and

lung volume-dependent components, a first-principles model 2.3 Initial validation of spherical alveolar model
is needed. The main simplifying assumption treats the lung as
n independent spherical alveoli. The model is derived from Before validating the model with clinical data, model shape
was assessed to evaluate if it matched expected macroscopic
four equations. First, stress, , in the alveolar wall is related
lung behaviour. If specific compliance reflects the underlying
to alveolar pressure, P, by assuming the alveoli is a thin
tissue distensibility and is not dependent on volume, then it
walled sphere with constant wall thickness, t, and radius, Rf:
should be able to demonstrate a constant specific compliance
with this simple model. Thus, the modelled lung compliance,
(3) and specific lung compliance, was calculated for a range of
alveolar radii and thus a range of alveolar volumes. Although

300
IFAC BMS 2018
São Paulo, Brazil, September 3-5, 2018 Oliver Kannangara et al. / IFAC PapersOnLine 51-27 (2018) 299–304 301

it is the resting alveolar volume, R0, which is more relevant to Equation (10) can be rearranged where all non-mass terms
this study, the effects of changing the final alveolar volume, are collected into a single constant, km, yielding:
Rf, were also examined to determine if modelled compliance
is approximately constant within a tidal breath, as expected.
Next, experimentally determined lung parameters from the (11)
literature (Table 3) were substituted into Equation 9 to
calculate theoretical values of lung elastance for both adults Where Cs is the specific compliance and km is the specific
and neonates. These theoretical values are compared to elastance, except each is scaled for mass instead of FRC.
median elastance values for neonates and adults obtained Equation (11) is fit using a linear least-squares regression.
from the clinical trials and Equations (1)-(2). First it is fit to only neonatal patients, and then projected to
In Table 1, adult values of elastic modulus and Poisson’s predict the adult values of elastance. This approach is a more
ratio were used for both adults and neonates, which is robust method to test the validity of the model than fitting to
consistent with the assumption of adults and neonates having both adults and neonates, ensuring the model is not ‘forced’
similar specific compliance. Adult values of wall thickness to fit all values. There is thus a fitted data set (neonates), and
were scaled by the alveolar radius ratio, tR, to obtain neonate a validation data set (adults). The median specific compliance
values. Due to the limited relevant human data, animal values values of adult and neonatal patients is also compared using a
are used for alveolar wall thickness and elastic modulus. To two-sample non-equal variance t-test.
convert rat and rabbit alveolar septal thickness to human
septal thickness, a difference in interstitial thickness of 2 µm 3. RESULTS
is used based on Mercer et al. (Mercer et al., 1994).
3.1 Elastance from single compartment model
Table 1: Alveolar parameters from literature for validation
Adult Values The elastance values obtained by fitting the lung model of
Alveolar wall thickness (Cavalcante Equation 1 to ICU and NICU patients are reported in Table 4.
t 7-10 µm Median specific compliance values were 0.51 mL/cmH2O/kg
et al., 2005, Tsunoda et al., 1974)
Poisson’s Ratio (Al‐Mayah et al., for adult patients and 0.78 mL/cmH2O/kg for neonatal
v 0.4 patients (p > 0.098). Note mL/cmH2O/kg is not a SI unit, but
2009)
Elastic modulus (Cavalcante et al., is the normalised specific compliance in SI units.
E 5-11 kPa
2005, Perlman et al., 2014) Table 2: Elastance and specific compliance values obtained
Alveolar radius (Ochs et al., 2004) R0 100 µm from observational trial with 9 neonatal and 7 adult patients
# alveoli (Ochs et al., 2004, Weibel,
n 3-4.8 x 108 Specific
1973)
Elastance median compliance
Neonatal Values
No. [IQR] median [IQR]
Alveolar wall thickness (Cavalcante Pt. breaths (cmH2O/mL) (mL/cmH2O/kg)
t 3.1 -6.2 µm
et al., 2005, Tsunoda et al., 1974) Neonatal Patients
Alveolar radius (Davies et al., 1970) R0 62 µm 1 4501 1.43 [0.92-2.41] x103 0.78 [0.46 - 1.22]
# alveoli (Langston et al., 1984) n 0.2-0.5 x 108 2 23770 0.15 [0.11-0.22] x103 1.95 [1.33 - 2.66]
3 42541 0.41 [0.25-0.69] x103 0.89 [0.53 - 1.45]
2.4 Relation to specific compliance 4 21888 0.43 [0.37-0.52] x103 1.47 [1.22 - 1.71]
5 65738 2.20 [1.91-2.64] x103 0.39 [0.33 - 0.45]
It is well-known that lung volume is approximately 6 50296 1.04 [0.73-1.30] x103 0.58 [0.47 - 0.83]
proportional to body weight (Bartlett et al., 1977). If the 7 59557 1.63 [1.28-2.22] x103 0.80 [0.59 - 1.02]
hypothesis that specific compliance is independent of lung 8 71052 2.78 [1.28-2.22] x103 0.38 [0.34 - 0.47]
volume is correct, then it should also scale essentially linearly 10 134528 1.77 [1.45-2.04] x103 0.71 [0.61 - 0.86]
with body weight. Thus, Equation 9 can be further simplified Adult Patients
for this case by collecting the volume-dependent term into a 1 252908 22.3 [19.7-25.8] 0.31 [0.27 - 0.35]
constant, k, multiplied by body weight (m): 2 18767 20.2 [17.9-24.2] 0.71 [0.53 - 0.64]
3 57067 21.9 [18.5-26.4] 0.51 [0.42 - 0.60]
4 26099 42.9 [41.0-43.9] 0.35 [0.34 - 0.36]
(10) 5 42513 31.8 [29.2-34.8] 0.44 [0.40 - 0.48]
6 258809 19.0 [11.5-23.0] 0.64 [0.53 - 1.07]
Where k is an experimentally determined constant with units 7 1069363 22.9 [19.0-29.9] 0.91 [0.69 - 1.10]
m3/kg, relating lung dimensions to body mass. The units of
the material dependent term are Pa, which is the same as the
specific compliance. Note Equation (10) only holds if the
volume-dependent term [in brackets] is shown in the model
validation to be directly proportional to initial lung volume.

301
IFAC BMS 2018
São Paulo, Brazil, September 3-5, 2018 Oliver Kannangara et al. / IFAC PapersOnLine 51-27 (2018) 299–304
302

3.2 Variation of compliance with lung volume 3.3 Validation of spherical linear elastic lung model

Fixing the final alveolar radius, Rf, and varying the initial Using maximum and minimum values from Table 3, the
alveolar radius, R0, in Equation (9), yields Figure 1. Both spherical linear elastic lung model of Equation (9) is used to
axes are normalised by values from the lowest volume, calculate a range of theoretical elastance values. These
giving unit-less volume and compliance, as the general trends model-derived values are compared with experimental
are desired, rather than the absolute values. The increase in elastance values from the single compartment model and
alveolar volume from infancy to adulthood is approximately literature. Table 3 shows good agreement between literature
four-fold (Ochs et al., 2004, Davies et al., 1970). Hence, this and model-derived elastance, once chest wall elastance is
range has been used as the limits to the volume-axis. accounted for. Note, neonatal chest wall elastance in
literature are normalised for mass. Thus, it was multiplied by
With alveolar thickness held constant, the relationship 2.5kg, the mass of a typical healthy neonate in this case.
between alveolar volume and compliance is non-linear. This
nonlinearity results in a non-constant specific compliance. Table 3: Validation of spherical linear elastic lung model by
However, when thickness is made a proportional function of comparison with experimentally obtained elastance values
alveolar radius, this relationship becomes linear, y = x, and
specific compliance is accordingly constant. Therefore, Elastance value (cmH2O/L) Neonate Adult
modelling alveolar stiffness as a function of radius to capture Experimental Ers: median 1432 22.3
growth better matches physiological and clinical data. [range] [150-2780] [19.0-42.9]
‘Typical’ Ers, in literature
500-667 10.0-16.7
(Donn et al., 2012)
Chest wall elastance, in
130 5.6
literature (Davis et al., 1988)
Lung elastance (w/o chest
370-537 4.4-11.1
wall), in literature
Spherical linear elastic model 80-550 2.8-8.7

3.4 Validation in predicting adult elastance

The specific compliance model is fit to neonatal data and

extrapolated to ‘predict’ adult compliance in Figure 3.
Because Equation (11) accounts for mass and thus lung
volume, any deviation from constant specific compliance
should be due to variation in the underlying tissue stiffness,
independent of volume. The overall prediction is very good
with R2 = 0.85 explaining 85% of the variation.

Figure 1: Variation of lung compliance and elastance with

R0. Values are unitless, normalized to value at lowest volume.
Figure 2 results from fixing resting alveolar radius, R0, and
increasing the Rf in Equation (9), simulating lung expansion
in inspiration. Compliance increases ~20%, where lung
compliance should be relatively constant (dotted red line;
500mL tidal volume over 2.4L FRC). This result indicates
compliance is not very sensitive to final radius, Rf, when R0 is
varied, as long as it is within the region of a tidal breath.

Figure 3: Log-log scale plot of specific compliance model fit

to neonatal data and used to predict adult compliance
To further examine if specific compliance is independent of
lung volume, the neonatal elastance values of specific
compliance were plotted against body weight in Figure 4. If
Figure 2: Variation of compliance as lung expands. Volume there is no statistically significant relationship between mass
and compliance are unitless, normalized by values at Rf = R0. and specific compliance, this outcome would support the

302
IFAC BMS 2018
São Paulo, Brazil, September 3-5, 2018 Oliver Kannangara et al. / IFAC PapersOnLine 51-27 (2018) 299–304 303

assertion that specific compliance is independent of lung greatly depending on the inputs. Some of these inputs,
volume and reflects only the intrinsic material properties of namely elastic modulus and septal thickness, were derived
the lung tissue. However, there is a statistically significant (p from animal data, and have a wide range of quoted values. A
< 0.05) trend in specific compliance, which does not support range of inputs was thus included, producing a range of
this assumption, although it does have a poor R2 value. Thus, elastance values. However, the main purpose of the model is
for each cohort, there is measurable variation showing only a to assess if specific compliance is independent of lung
weak association between specific compliance and weight, volume. Thus, it is not concerned with producing a precise
although the overall trend across all body weights (neonate to and accurate value of elastance. Hence, a good match of
adult) in Figure 4 supports its use on average. values across the range is acceptable for validation to this
purpose.
In comparing the model-derived elastance values with those
from the literature, the chest wall elastance had to be
accounted for by using values from the literature. The chest
wall elastance is also known to vary with lung volume, with
neonates having far higher values of chest wall elastance than
adults (GERHARDT et al., 1980) . If desired, this aspect
could be modelled in a similar manner to how the lung
elastance is modelled, albeit with different geometry. If the
chest wall is modelled as a half cylinder, the same underlying
equations could be used to model the chest wall elastance,
except with Equation (1) being divided by a factor of 1
instead of 2. The overall outcomes of the analysis, would
remain similar.

Figure 4: Variation of neonatal specific compliance with These findings support the assertion that the macroscopic
body weight. pressure-volume properties of mechanically validated adult
and infant lungs are the same when normalised to body
4. DISCUSSION weight as a surrogate of lung size. This study also shows, as a
result, that specific compliance is a reasonable measure of a
patient’s underlying tissue distensibility, independent of lung
4.1 Spherical linear-elastic alveoli model volume. Finally, the model is able to define this specific
compliance as a function of alveolar tissue elastic modulus,
The overall model captures compliance well. However, Poisson’s ratio, alveolar wall thickness to diameter ratio, and
Figure 2 shows the apparent compliance of alveoli increases surface tension effects, creating a foundation model on which
as they expand. In reality, compliance is expected to decrease to assess specific cases or hypotheses in terms of fundamental
as the lung expands, due to distension of the alveolar tissue mechanical properties.
effecting an increase in the material stiffness. However,
because the model assumes the alveolar tissue to be linearly 5. CONCLUSIONS
elastic, this behaviour is not seen, but could be captured with
a more exact model of elastic modulus in Equation (9). A simple model based on fundamental mechanics is
presented demonstrating specific compliance is a measure of
The model does capture the volume-dependent effects of alveolar elastic modulus, Poisson’s ratio, alveolar wall
increasing lung volume. The lung is not becoming less stiff thickness to diameter ratio, and surface tension effects. The
as it expands, but more volume is gained for each incremental model shows mechanically ventilated neonates and adults
increase in tissue strain, due to an increasing volume to radius have statistically similar values of specific compliance, where
ratio. If a more realistic model for changes in lung prior studies only showed this relationship via correlation.
compliance over a breath was desired, the linear elastic Thus, this study is the first demonstration using fundamental
modulus could be replaced by an experimentally determined mechanics, leading to a simple model for assessing lung
function for elastic modulus that varies as a function of mechanical behaviour and response.
alveolar strain that also included large deformations, which
are currently not modelled. However, the models serves its
6. REFERENCES
main purpose to investigate and quantify compliance over
tidal breathing with relatively small deformations. AL‐MAYAH, A., MOSELEY, J., VELEC, M. & BROCK, K.
2009. Sliding characteristic and material compressibility
4.2 Predicting typical neonatal and adult elastance values of human lung: Parametric study and verification.
Medical physics, 36, 4625-4633.
BARTLETT, D. & ARESON, J. 1977. Quantitative lung
Results show the spherical linear elastic alveolar model gives
morphology in newborn mammals. Respiration
reasonable estimates of typical neonatal elastance. This result
physiology, 29, 193-200.
is quite surprising given the number of simplifying
CAVALCANTE, F. S., ITO, S., BREWER, K., SAKAI, H.,
assumptions. However, the modelled values of elastance vary
ALENCAR, A. M., ALMEIDA, M. P., ANDRADE, J. S.,

303
IFAC BMS 2018
São Paulo, Brazil, September 3-5, 2018 Oliver Kannangara et al. / IFAC PapersOnLine 51-27 (2018) 299–304
304

MAJUMDAR, A., INGENITO, E. P. & SUKI, B. 2005. SZLAVECZ, A., CHIEW, Y. S., REDMOND, D.,
Mechanical interactions between collagen and BEATSON, A., GLASSENBURY, D., CORBETT, S.,
proteoglycans: implications for the stability of lung tissue. MAJOR, V., PRETTY, C., SHAW, G. M., BENYO, B.,
Journal of applied physiology, 98, 672-679. DESAIVE, T. & CHASE, J. G. 2014. The Clinical
CHIEW, Y. S., CHASE, J. G., SHAW, G., SUNDARESAN, Utilisation of Respiratory Elastance Software (CURE
A. & DESAIVE, T. 2011. Model-based PEEP Soft): a bedside software for real-time respiratory
optimisation in mechanical ventilation. BioMed Eng mechanics monitoring and mechanical ventilation
OnLine, 10. management. BioMedical Engineering OnLine, 13, 140.
CHIEW, Y. S., PRETTY, C. G., SHAW, G. M., CHIEW, Y. TSUNODA, S., FUKAYA, H., SUGIHARA, T., MARTIN,
W., LAMBERMONT, B., DESAIVE, T. & CHASE, J. G. C. & HILDEBRANDT, J. 1974. Lung volume, thickness
2015. Feasibility of titrating PEEP to minimum elastance of alveolar walls, and microscopic anisotropy of
for mechanically ventilated patients. Pilot and Feasibility expansion. Respiration physiology, 22, 285-296.
Studies, 1, 9. WEIBEL, E. R. 1973. Morphological basis of alveolar-
DAVIES, G. & REID, L. 1970. Growth of the alveoli and capillary gas exchange. Physiological Reviews, 53, 419-
pulmonary arteries in childhood. Thorax, 25, 669-681. 495.
DAVIS, G., COATES, A., PAPAGEORGIOU, A. &
BUREAU, M. 1988. Direct measurement of static chest
wall compliance in animal and human neonates. Journal
of Applied Physiology, 65, 1093-1098.
DONN, S. M. & SINHA, S. K. 2012. Manual of neonatal
respiratory care, Springer.
GATTINONI, L., PESENTI, A., CASPANI, M.,
PELIZZOLA, A., MASCHERONI, D., MARCOLIN, R.,
IAPICHINO, G., LANGER, M., AGOSTONI, A. &
KOLOBOW, T. 1984. The role of total static lung
compliance in the management of severe ARDS
unresponsive to conventional treatment. Intensive care
medicine, 10, 121-126.
GERHARDT, T. & BANCALARI, E. 1980. Chestwall
compliance in full‐term and premature infants. Acta
Paediatrica, 69, 359-364.
HICKLING, K. G. 1998. The pressure–volume curve is
greatly modified by recruitment: a mathematical model of
ARDS lungs. American journal of respiratory and critical
care medicine, 158, 194-202.
JARREAU, P.-H., LOUIS, B., DASSIEU, G., DESFRERE,
L., BLANCHARD, P. W., MORIETTE, G., ISABEY, D.
& HARF, A. 1999. Estimation of inspiratory pressure
drop in neonatal and pediatric endotracheal tubes. Journal
of Applied Physiology, 87, 36-46.
LANGSTON, C., KIDA, K., REED, M. & THURLBECK,
W. 1984. Human lung growth in late gestation and in the
neonate. The American review of respiratory disease,
129, 607-613.
MERCER, R. R., RUSSELL, M. L. & CRAPO, J. 1994.
Alveolar septal structure in different species. Journal of
Applied Physiology, 77, 1060-1066.
OCHS, M., NYENGAARD, J. R., JUNG, A., KNUDSEN,
L., VOIGT, M., WAHLERS, T., RICHTER, J. &
GUNDERSEN, H. J. G. 2004. The number of alveoli in
the human lung. American journal of respiratory and
critical care medicine, 169, 120-124.
PERLMAN, C. E. & WU, Y. 2014. In situ determination of
alveolar septal strain, stress and effective Young9s
modulus: an experimental/computational approach.
American Journal of Physiology-Lung Cellular and
Molecular Physiology, 307, L302-L310.
PHELAN, P. D. & WILLIAMS, H. E. 1969. Ventilatory
studies in healthy infants. Pediatric research, 3, 425-432.

304