PNEUMA Release 3.

0 software described in this document is furnished by the Biomedical Simulations

Resource under the terms of a release agreement.
PNEUMA may be used only under the terms of the release agreement.

PNEUMA User’s Guide

Contact: [email protected]

Release 3.0 – January 2013

Release 2.0 – January 2011
Release 1.1 – March 2003
Release 1.0 – August 2002
Beta Release – September 2001

Supported by: NIH Grant P41-EB001978

© 2013 Biomedical Simulations Resource (BMSR)
University of Southern California
 PNEUMA Release Agreement
Before you use PNEUMA, please read the following conditions for using our package.
Thank you for your cooperation.

 PNEUMA is restricted to non-profit research and instructional purposes aimed at

further knowledge in the area of cardiorespiratory system modeling and
simulation. PNEUMA is supported by University of Southern California (USC)
Biomedical Simulations Resource (BMSR) (NIH Grant P41-EB001978 ).
 Any publications of research results that were obtained in part by the use of
PNEUMA will contain proper acknowledgement of the BMSR at USC. Reprints
of such publications will be sent to the BMSR for the record.
 I will not distribute PNEUMA, in whole or part, to others without the expressed
permission of the BMSR.
 I understand that neither USC nor the BMSR make any warranties, expressed or
implied, that PNEUMA is free of errors or is consistent with any standard of
merchantability, or that it will meet my requirements for any particular
application. I understand that PNEUMA should not be relied on for solving a
problem whose incorrect solution could result in injury to a person or loss of
property, and that if I do use PNEUMA in such a manner it is at my own risk. I
understand that USC, the BMSR and the authors disclaim any and all liability for
direct or consequential damages resulting from my use of PNEUMA.
Contents
Getting Started………………………………………………………………............. 2
Individual Model…………………………………………………………….............. 2
Overall Pneuma Model………………………………………………………............ 2
Open Pneuma…………………………………………………………………........... 2
Constant Parameters…………………………………………………………............. 6
Adjustable Inputs……………………………………………………………............. 8
Interventions…………………………………………………………………............ 9
Block Description……………………………………………………………............ 15
Contact and Support…………………………………………………………............. 15
Blocks Reference……………………………………………………………............. 16
Overall Pneuma………………….……………………………………………........... 17
Reflexes (Reflex_Ursino.mdl)………………………………………………............. 19
Carotid Baroreceptors……………………………………………….……........... 20
Chemoreflex…...…………………………………………………………........... 22
Lung Stretch Receptors Reflex…………………………………………............. 24
Offsets……………………………………………………………………............ 25
Autonomic Control…………………………...….………………………….............. 27
SA Node (SA_Node_Ursino.mdl)…………………..………………………............. 31
-Sympathetic Control…………………………...………………….….............. 33
Parasympathetic Control…………………………….………………….............. 34
-Sympathetic Control of Peripheral Resistance (TPR_Ursino.mdl)….……............ 35
Variable Breathing Period (PNEUMA.mdl)…...……………………………............. 37
Variable Heart Period (PNEUMA.mdl)….………………………….………............. 40
Cardiovascular System (PNEUMA.mdl)……………………………………............. 42
Neuromuscular Drive (NeuroMuscular.mdl)…………………………………........... 48
Respiratory Muscle Activity (Pmus_Flow_Younes.mdl)……………………............ 50
Pleural Pressure (Pleural_Schuessler.mdl)………………………………….............. 53
Gas Exchange and Transport (Gas_Exchange.mdl)…………………………............ 55
Dead Space (Dead_Space_Khoo.mdl)………………..………………………........... 57
Alveolar Gas Exchange (Lungs_Khoo.mdl)……..…………………………............. 60
Cardiovasuclar Mixing, Convection and Dissociation (Cardio_Mix_Lange.mdl,
Dissociation_Spencer.mdl).………………..……..……..…...…….…….…….……. 62
Brain Compartment (Brain_Khoo.mdl)…………………………………….…….…. 66
Body Tissues Compartment (Body_Khoo.mdl)……………………………....…….. 68
Ventilatory Response (Vent_Drive_Khoo.mdl)……………………………....…….. 70
Upper Airway / State Change (State_UA_Khoo_Borbely.mdl)……………....…….. 72
Upper Airway………………………………………………….……….….……. 73
Sleep Mechanism……………………………………………………..…...……. 75
Metabolic Control (PNEUMA.mdl)..……………………………………..…...… 79
Autonomic and Metabolic Interactions………………………………….………. 86
Appendix I: Software Package..……………………………………………….…….. 88
Appendix II: Saved Data Files…...…………………………………………....…….. 91
Appendix III: Save/Load Date for Advance Users………..………………………… 92
Appendix IV: Overall Parameter Set and Initial Conditions……………...………… 94
 2

Getting Started
Thank you for trying out PNEUMA and its modularized component models. Before using
PNEUMA, please take a moment to read the Release Agreement first. To download the
associated files or their updates, please go to bmsr.usc.edu and click on Software. Before
you begin to use PNEUMA or its individual model components, please take a moment to
make sure that you have downloaded the most recent files that you will be using. In
Appendix I, there is a full list of files that are included in zipped format on the BMSR-
PNEUMA web site. After you unzip the downloaded file, please refer to the appendix
and check that you have the correct files.

Individual Sub-Models
PNEUMA is implemented using Simulink and Matlab version R2007b or higher (© The
Mathworks Inc., Natick, MA), which provides a graphical programming environment that
promotes modularization of the overall model into hierarchically smaller subsystems.
This allows the user to customize parts of the overall model in accordance to his/her
simulations needs. Alternatively, the user may also choose to focus on a specific
PNEUMA block and use it to study the corresponding mechanism of interest. Therefore,
depending on the user’s interest and needs, individual component blocks may be
downloaded and used.
Please refer to the reference and the “.m” file for variable names and values of each
compartment. Some of the components are difficult to decompose into smaller modules
and therefore may not be suitable for your application. If you have suggestions or would
like to request modifications to PNEUMA components that would better suit your
simulation needs, please feel free to send us feedback. Contact information is provided in
the Support and Contact section of this manual.

PNEUMA V.3.0: What’s New

In Pneuma Release 3.0, we have incorporated a metabolic component with autonomic-
metabolic interactions into the existing integrative comprehensive simulation model. This
metabolic component of PNEUMA is based on prior models of glucose-insulin regulation
by Bergman et al. (1979) and free fatty acid (FFA) regulation by Roy and Parker (2006).
Changes in sympathetic activity from the autonomic portion of PNEUMA produce
changes in epinephrine output, which in turn affects the metabolism of glucose, insulin
and FFA. Inputs from the dietary intake of glucose and external interventions, such as
insulin injections, have also been incorporated into the model. Also incorporated is
autonomic “feedback” from the metabolic component to the rest of PNEUMA in the
following way: changes in insulin level are assumed to lead to changes in sympathetic
tone. The “Control Panel” along with other input panels have been improved to facilitate
greater user interaction and control of the simulations
 3

References

Bergman, R. N., Ider, Y. Z., Bowden, C.R.,and Cobelli,C.(1979).Quantitative estimation

of insulin sensitivity. Am. J. Physiol. 236, E667–E677.

Roy, A., and Parker, R. S. (2006). Dynamic modeling of free fatty acid, glucose, and
insulin: an extended “Minimal Model”. Diabetes Technol. Ther. 8, 617–626.

Using Pneuma
To begin using Pneuma, unzip the “PneumaRelease3.zip” file and check that you have
all the necessary files. For the list of files in “PneumaRelease3.zip” file, please refer to
“Getting Started” section. After you have unzipped the file and you are ready to run the
program in the MATLAB environment, make sure that you are in the directory where the
unzipped files are located. To Open Pneuma, in the Matlab command prompt, type
“PNEUMA_MAIN_CONTROL_PANEL”.

If you are running Pneuma using a version of Matlab higher than Matlab75 (version
2007b), a series of warnings may appear due to compatibility issues, but these warnings
should disappear after the first time you open PNEUMA.
 4

The Control Panel graphic user interface (GUI) will appear, as shown below.
 5

Next, input the parameter values.

 Start Time: time to start the simulation. (default is zero seconds)
 End Time: end-time (in seconds) of the simulation (for example:
3600*24*7 will end up with 7-day simulation.
 Max Step: the simulation is using variable integration time steps, and it
requires that the user specify the maximum allowable time step. A large
max time step is not recommended (the default is 0.01 second).
 Saved Sample Time: some of the parameter/variable values can be saved
to data files after each simulation and the user has the option of specifying
the sample time of the saved segment. If given value -1, it will be default
sample time of the simulation which can be used for saving data with
“Saved Segment Time” as 0.5 day. The suggested sample time for saving
is 0.1 second for neural-cardio-respiratory system. The sample time for
metabolic system is constant as 6 seconds.
 Saved Segment Time: each data file can be saved as long as the segment
time. The suggested time is 7 days.

The “Run” and “Stop” buttons allow the user to run and terminate the simulation.
Currently, the Real Time Workshop allows the simulation to run using the accelerated
mode even with standard Matlab Simulink package. So the “Run” operation will run in
accelerated mode that allows the simulation to run faster. If the user prefers to execute
the model in normal mode, it will have to be run under Simulink model itself rather than
using that GUI button (see options under Simulation tab in Simulink model window). If
you decide to stop the simulation before the End Time that you have specified, some
data will be stored to files (Saved Sample Time option) and all the variables are in the
workspace which can be saved later. The “Reset” button will reset all variable values to
their defaults.

Under “Open” menu, the user has three options. “Open Pneuma Model Ctrl+O” will
show the Pneuma model in Simulink. User can explore the modules in Pneuma and
incorporate other blocks if needed. “Open Display Panel Ctrl+D” allows the user to see
the output from some of the more common measurements such as arterial blood pressure,
heart rate and so on. Having achieved some familiarity with PNEUMA, the user may
want to add more inputs to the display panel or create new displays. “Open Program
Status Ctrl+P” will show the Pneuma Progress module in Simulink, that displays the
total duration of simulation, current simulation time and percentage of simulation
completed, based on total duration of simulation and current simulation time.

If the user wants to load or save the simulation workspace, click under “File” menu and
two selections will show up. “Load Data Ctrl+L” opens the standard Matlab open file
window, which allows the user to specify the data file and load the data into workspace.
“Save Data Ctrl+S” opens the standard Matlab save file window, which allows saving
the workspace data to the directory of user's choice.
 6

Constant Parameters
When the user clicks on “CONSTANT PARAMETERS” button, another graphical
interface will appear, as shown below:
These are the constant parameters used in the model. The values may be changed before
the simulation, if desired. It is recommended that these parameters be left at their default
values. Each model subsystem is listed along with the constant parameter in that
compartment. Each title button gives user the opportunity to open the Simulink
implementation of that particular subsystem.
 7

If the mouse cursor is placed and held at a particular box with number, the help text for
the corresponding parameter will appear so that the user will know what physiological
entity that parameter represents, as shown below:
 8

Adjustable Parameters
This panel allows the user to vary parameters before or during the simulation. Click on
“ADJUSTABLE PARAMETERS” button and the following panel will appear:

The user can adjust the value either by using the slider bar or by typing directly into the
box. Both the “min” and the “max” values are shown for each slider bar. These values
can be changed as well. But the values that fall within the default spans indicated are
recommended, since these are consistent with physiologically feasible ranges.
 9

The above panel shows the parameters that may be altered in value while the simulation
is being executed. Since the model is continually being revised, the actual parameters that
can be adjusted may be different in different versions of the program.

External Interventions
Here, the user is permitted to apply a variety of external interventions to the model.
Click on “EXTERNAL INTERVENTIONS” and the graphical panel opens up, shown
below:

The panel shows the interventions that have been included in the model at the present
time. Before you run each intervention, please click “Reset” button on “Control Panel” to
reload the original parameter set, then enter your new start/stop time and other parameters
on the Control Panel, then go back to the External Interventions. Again, as this software
gets updated, other interventions will be added.

The followings are some typical examples for the interventions.

A. Hypoxia. To simulate hypoxia, simply enter values into “Start Time” and “Duration
Time” such as start at 800 sec with duration 300 sec, then enter value into “Change in
PIO2” such as “-90” by default, then go back to Control Panel and click on “Run”
button, make sure the simulation “End Time” is equal or longer than the hypoxia end
time, shown below:
 10

B. Normocapnic Hypoxia. To simulate normocapnia in hypoxia, first click on check box

of “Normocapnia”, then define the hypoxia condition as in Hypoxia, then give the
same “Start Time” and “Duration” in CO2 Inhalation part as O2 Inhalation part, then
go back to Control Panel and click on “Run” button, shown below:

C. Non-Normocapnic Hypoxia. To simulate non-normocapnic including hypercapnic

hypoxia, first click on check box of “Non-Normocapnia”, then define the hypoxia
condition as in Hypoxia, then give the same “Start Time” and “Duration” in CO2
Inhalation part as O2 Inhalation part, then enter value into “Change in PICO2” such
as “40” by default, then go back to Control Panel and click on “Run” button, make
sure the simulation “End Time” is equal or longer than the hypercapnia end time,
shown below:
 11

D. Normal Sleep. To simulate normal sleep, simply click on check box “Sleep Enable”,
then go back to Control Panel and click on “Run” button. You can change the
parameter set for sleep to simulation different interventions. For overnight sleep,
make sure your “End Time” in Control Panel is longer than 3600*8+200 seconds (>8
hrs) , shown below:

E. OSA Sleep. To simulate obstructive sleep apnea (OSA) sleep, first click on check box
“Sleep Enable”, then drag the slider button in “Upper Airway Mechanism” or directly
enter value into “Pcrit” such as -2.66 which will simulate a moderate OSA, then go
back to Control Panel and click on “Run” button. For overnight sleep, make sure your
 12

“End Time” in Control Panel is longer than 3600*9+200 seconds (>9 hrs) , shown
below:

F. CPAP with OSA Sleep. To simulate continuous positive airway pressure (CPAP),
first set up OSA sleep as the above example in OSA Sleep. Then click on check box
“CPAP”, enter values into “Start Time” and “Duration”, give values for the positive
pressure such as 15 cmH2O, then go back to Control Panel and click on “Run”
button. You can try 1 hour CPAP shown as below or overnight CPAP for OSA Sleep.
In our model, the default mode is to repeat CPAP every night if the CPAP duration is
longer than 1 day. For example, if the simulation runs for 30-day OSA sleep with 10-
day CPAP, on in the middle of the 30-day run time simulation, then the CPAP “Start
Time” could be 3600*24*10-1800 sec (which is 0.5 hour short than 10 days) and
“Duration Time” could be 3600*24*10+3600*2 sec (which is 2 hours longer than 10
days).
 13

G. Maneuvers. To simulate Mueller Maneuver, click on check box “Mueller Maneuver”,

then use the default setup which can be entered with different values as you desire,
then go back to Control Panel and click on “Run” button. To simulate Valsalva
Maneuver, click on check box “Valsalva Maneuver”, then use the default setup which
can be entered with different values as you desired, then go back to Control Panel and
click on “Run” button, shown below:

H. CSR-CHF Sleep. To simulate central sleep apnea (CSA characterized with Cheney-
Stokes Respiration CSR) with congestive heart failure (CHF), first activate Sleep
as in Normal Sleep. Then to change heart contractility, go to “Adjustable
Parameters”, enter value such as “0.475*0.3” or drag the slider bar for
 14

“Gain_Emaxlv” and enter value such as “2392*0.3” or drag the slider bar for
“Basal_Emaxlv” in “Heart Contractility” area, then increase chemoreflex gain such as
increase “Peripheral Chemo-Gain” by directly entering value as “0.0063*6” (example
value) or dragging the slider bar, then increase “Lung-Chemo Volume” by directly
entering value as “0.588*1.5” (example value) or dragging the slider bar. Lastly, go
back to Control Panel and click on the “Run” button, as shown below:

These are brief descriptions to help the user get started using our package. Please feel free
to explore the model. Since this is an open source environment, contribution of newer
code or model will also help us to improve our implementation and to better suit the
needs of other users as well.
 15

Block Description
For the complete descriptions of all the individual Simulink model blocks, please refer to
the “Blocks Reference” section.

Contact and Support

The whole model and its modularized components will be updated from time to time. So,
please check the website for newer update or if you wish to join the mailing list,
notification will be sent to you regarding our progress on the update.
FAQ will be set up as we get more questions and comments. In the meantime, please
send all your valuable comments and feedbacks to [email protected]. Once we
have the solution, then we will post it in the forum so that other users can benefit from it.
The PNEUMA project is supported by the USC Biomedical Simulations Resource (NIH
Grant P41-EB001978). Comments and feedback on all aspects of this software are
welcome.
 16

Blocks Reference
 17

PNEUMA V.3.0

Description

PNEUMA is implemented using SIMULINK. The open architecture of PNEUMA allows

to group models into hierarchies to create a simplified view of components or
subsystems. High-level information is presented clearly and concisely, while detailed
information is easily hidden in subsystems within the model hierarchy. Current
PNEUMA implementation builds up on 557 model parameters and allows the tracing of
93 model states. It is a hybrid model that simultaneously addresses fast and slow
physiological processes (i.e. single heart beat and circadian rhythm) that are implemented
in mixed discrete and continuous modes.
The modular design of PNEUMA makes it possible to perform simulations in which
specific physiological mechanisms are excluded or added in order to better determine
their contribution to the closed-loop operation of the overall system of interconnected
components. This allows the user to explore alternative models of physiologic function in
silico, which could be very useful in circumventing the challenges of attempting to study
the systems in question experimentally or clinically. As well, the modularity of
PNEUMA enables users to replace one or more of the model blocks with their own
modules of specific physiological components.

General References:

1. Cheng, L., and Khoo, M. C. K. Modeling the autonomic and metabolic effects of
obstructive sleep apnea: a simulation study. Front Physiol 2:111, 2012. doi:
10.3389/fphys.2011.00111.

2. Cheng, L., Ivanova, O., Fan, H., and Khoo, M. C. K. An integrative model of
respiratory and cardiovascular control in sleep-disordered breathing. Respiratory
Physiology and Neurobiology 174, 4-28, 2010.
 18

Simulink Model. Overall Pneuma

Display
Maneuvers 200 Progress
Panel
t_stop
Double Click
to load Initial Conditions

Variable Respiratory Rhythm

Mech Vent Neural

Tidal Volume Vt
RespMus Drive
Sleep/Awake

External Pressure PaO2

Total Ventilatory Drive
PaCO2
Metabolic Control
SAO2 ftas

PbCO2
ftas _v

ftbs

ftp

Central Neural Control

AI -Arousal Index

PaCO2

CaO2

AI -Arousal Index fcs Cardiovascular System

SI-Sleep Wake State Index deltaFtas
Blood Flow

Ppl

REM

Respiratory System
 19

Reflexes (Reflex_Ursino.mdl)

Description

The reflexes model includes the key cardiorespiratory reflexes: baroreflex, chemoreflex
and lung stretch receptor influences on respiration and heart-rate control.

Reflexes

ABP
Carotid fcs
Dstate Baroreceptors

PCO2
Chemoreflex fchemo
PO2

Vt Lung Stretch fls

Receptors Reflex
 20

Carotid Baroreceptors
This block represents the pressor receptors that are located in the carotid sinus. In
response to arterial blood pressure changes, it produces both parasympathetic and the
sympathetic neural activity changes. During sleep, baro-sensitivity is assumed to
increase slightly. The input for this compartment is the arterial blood pressure, ABP, and
the output is the carotid sinus firing frequency, fcs.

Carotid Baroreceptors Equation:

 P  Pn   Pn 
 f cs, min  f cs,maxexp( )
k cs   k cs
f cs   
 P  Pn   Pn 
1  exp( )
 k cs   k cs 
 Pn  Pn _ sleep  (1  AI )  SI
kcs  Kcs _ sleep  (1  AI )  SI

Reference: Ursino, M, Interaction between carotid baroregulation and the pulsating

heart: a mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.

Simulink Model: Baroreceptors

 21

Input: ABP Arterial Blood Pressure

Output: fcs Carotid Sinus firing frequency

Variables: Pn Center pressure for sigmoidal function

kcs Parameter for sigmoidal slope control
fcs,min Lower threshold for sigmoidal function
fcs,max Upper saturation for sigmoidal function
Pn Pressure change in sleep
kcs Slope change in sleep
 22

Chemoreflex

Description

The inputs to the chemoreflexes are Oxygen (O2) and Carbon Dioxide (CO2) levels in
the arterial blood. This reflex affects both the heart rate and the peripheral vasculatures.
Inputs for this block are the oxygen and carbon dioxide partial pressure, PaO2 and
PaCO2. Output is the chemoreceptors firing, fac.

Chemoreflex Equations:

 _____ 
 PaO  PaO 
fchemo,min  fchemo,max exp 2 2
 kchemo     


chemo PaO2 , PaCO2       K  ln  PaCO2   
  ______  f
 _____ 
 Pa 
 PaO  PaO    CO2  
 
1  exp 2 2
 kchemo 

 
where

K H if Pa O 2  80

  Pa O2  80 
K   K H  1.2 
 if 40  Pa O 2  80
  30 
 K  1.6 if Pa O 2  40
 H

df chemo

1
 f chemo   chemo 
dt  chemo

Reference: Ursino, M, A mathematical model of CO2 effect on cardiovascular

regulation. American Journal of Physiology – Heart and Circulatory Physiology,
281:H2036-H2052, 2001.
 23

Inputs: PaCO2 Arterial CO2 partial pressure

PaO2 Arterial O2 partial pressure

Output: fchemo Chemoreceptor firing

Variables: fchemo,max Lower saturation for the sigmoidal function

fchemo,min Upper saturation for the sigmoidal function
_____
PaO2 Center point in the sigmoidal function
kchemo Slope control parameter for the sigmoidal
function
_____
PaCO2 Normalizing PaCO2 value
KH Constant value for the static response
f Constant value for the static response
τchemo Time constant for the chemoreflex

Simulink Model: Chemoreflex

 24

Lung Stretch Receptor Reflex

Lung inflation or deflation can produce changes in heart rate through the lung stretch
receptors. The input for this block is the tidal volume, Vt. The output is the lung stretch
receptor activity, fls.

Lung Stretch Receptors Reflex Equations

 lung  GlungVT

df lung
dt

1
 lung
 f lung   lung 

Reference: Ursino, M, A mathematical model of CO2 effect on cardiovascular

regulation. American Journal of Physiology – Heart and Circulatory Physiology,
281:H2036-H2052, 2001.

Simulink Model: Lung Stretch Receptors Reflex

Inputs: Vt Tidal volume

Output: fls Lung stretch receptors firing rate

Variables: Gls Constant gain

Τls Time constant
 25

Offsets (CNS Response in PNEUMA.mdl)

Offsets for Autonomic Control are the central nervous system response to the partial
blood pressure of carbon dioxide and oxygen in the cerebral circulation. The input for
this block are partial arterial blood pressure PaCO2 and PaO2. The outputs are the
offsets for autonomic control, Offsetres,vein,heart, respectively.

Offsets Equations

Offsetres,vein,heart  san,spn,sbn  O 2 sa ,O 2 sp ,O 2 sb  CO 2 sa ,CO 2 sp ,CO 2 sb

dO 2 sa ,O 2 sp ,O 2 sb 1
 ( O 2 sa ,O 2 sp ,O 2 sb  Wsa ,sp ,sb )
dt  isc
dCO 2 sa ,CO 2 sp ,CO 2 sb 1
 [CO 2 sa ,CO 2 sp ,CO 2 sb  gccsa ,sp ,sb  ( PaCO 2  PaCO 2 n )]
dt  cc
Wsa ,sp,sb  X sa ,sp,sb /(1  exp(( PaO 2 - PO2nsa ,sp,sb ) / kiscsa ,sp,sb ))

Reference: Ursino, M, A mathematical model of CO2 effect on cardiovascular

regulation. American Journal of Physiology - Heart and Circulatory Physiology,
281:H2036-H2052, 2001.

Inputs: PaCO2 Arterial CO2 partial pressure

PaO2 Arterial O2 partial pressure
Output: Offsetres,vein,heart CNS Response as offsets of autonomic
control
Xsa Saturation for the offset of α-sympathetic
Variables: activity on peripheral resistance
θsan Nominal level of offset of α-sympathetic
activity on peripheral resistance
PO2nsa Central point for the sigmoidal function
kiscsa Parameter of α-sympathetic activity on
peripheral resistance
Xsb Saturation for the offset of -sympathetic
activity
θsbn Nominal level of offset of -sympathetic
activity
PO2nsb Central point for the sigmoidal function
kiscsb Parameter of -sympathetic activity
Xsp Saturation for the offset of α-sympathetic
activity on peripheral resistance
θspn Nominal level of offset of α-sympathetic
 26

activity on peripheral resistance

PO2nsp Central point for the sigmoidal function
kiscsp Parameter of α-sympathetic activity on
unstressed volume of veins
τisc Time constant for oxygen response
τcc Time constant for carbon dioxide response

Simulink Model: Offsets (CNS Response)

-C-

theta _sa_n

1 f(u) 1
PaO 2 1/tao _isc s 1
theta _O2_sa
Wsa_Fcn Offset _Resistance
1/tao _isc

2
PaCO 2 -K- 1
1/tao _cc
s theta _CO2_sa
-C- Gain 2
1/tao _cc
PaCO 2_n

-C-
theta _sp_n

f(u) 1
1/tao _isc s 2
theta _O2_sp
Wsp_Fcn1 Offset _veins
1/tao _isc1

-K- 1
1/tao _cc
s theta _CO2_sp
Gain 1
1/tao _cc1

-C-
theta _sb_n

f(u) 1
1/tao _isc s 3
theta _O2_sb
Wsb_Fcn2 Offset _heart
1/tao _isc2

gcc _sb 1
1/tao _cc
s theta _CO2_sb
Gain 3
1/tao _cc2
 27

Autonomic Control (submodels refer to Autonomic.mdl)

Description

Influences from the central respiratory control (RSA), baroreflexes, chemoreflexes and
lung stretch receptors reflexes are integrated in this compartment and these inputs
determine the total -sympathetic, -sympathetic and parasympathetic influences on
heart rate and peripheral resistance. The inputs for this compartment are the central
respiratory drive, Nt, chemoreflex, fchemo, lung stretch receptors reflex, fls, carotid
baroreceptors firing, fcs, and CNS response, Offsets. The outputs are the -sympathetic
response, ftas, -sympathetic response, ftbs and parasympathetic response, ftp. The models
shown below is in PNEUMA.mdl, but the submodel is referred to Autonomic.mdl.

Autonomic Control

Autonomic Integration
Nt
 Central Respiratory Control ftas
fchemo  Chemoreflex
fls  Lung Stretch Receptors ftbs
Reflex
fcs ftp
 Baroreflex
Offsets
 CNS Response

Autonomic Integration Equations:

(a) Alpha-Sympathetic Activity

f tas _ res,vein  f s,  ( f s,0  f s, ) 

 
exp  ks  Gbaro,as f cs  Gchemo,as f chemo  Glung ,as f lung  GRSA,as N t  Offsetres ,vein 


(b) Beta-Sympathetic Activity

f tbs  f s,  ( f s,0  f s, ) 


exp  ks  Gbaro,bs f cs  Gchemo,bs f chemo  Glung ,bs f lung  GRSA,bs N t  Offsetheart 
 
 28

(c) Parasympathetic Activity

  fcs  fcs , 0  
 f para ,0  f para ,  exp  
  kp 
f tp   Gchemo , p f chemo  Glung , p f lung  GRSA, p N t  Offset para _ n
  fcs  fcs , 0  
1  exp  
  kp 

Reference: Ursino, M, A mathematical model of CO2 effect on cardiovascular

regulation. American Journal of Physiology – Heart and Circulatory Physiology,
281:H2036-H2052, 2001.

Simulink Model: Autonomic Control

4 0 1 1
noise
Total Alpha-Symp
Band-Limited Gain2 Gain3 ftas_blocker (ftas_res)
White Noise

4 0.4
Nt Central Respiratory G_CRSA
Neural Drive 0.34

G_lung_asymp
4 f(u) fs
(u<=60)*u+(u>60)*u
fas
G_chemo_asymp Alpha-Sympathetic
Response
5 1
Offset_Alpha-symp1 1 2
G_offset_asymp1 Alpha-Symp Total Alpha-Symp
f(u) f(u) ftas_blocker1
6 1
Integration (ftas_vein)
Offset_Alpha-symp2 Alpha-Sympathetic
1 G_offset_asymp2
Response1
lung feedback
0.24

G_lung_bsymp
2.8 f(u) (u<=60)*u+(u>60)*u 1 3
fbs fs
G_chemo_bsymp Total Beta-Symp
Beta-Sympathetic ftbs_blocker (ftbs)
Response
2 1
7 1
Chemoreflex G_chemo Offset_Beta-symp
G_offset_bsymp Beta-Symp
Integration

0.24
G_lung_para
0.03 (u>=0)*u 1 4
G_chemo_para Total Parasymp
3 1 f(u) ftp_blocker
(ftp)
fcs fp
Carotid Sinus G_fcs
-C- Parasympathetic 1
BaroResponse
theta_para_n G_offset_para Parasymp
Integration
 29

Simulink Model: Alpha-Sympathetic Response

Simulink Model: Beta-Sympathetic Response

Simulink Model: Parasympathetic Baroresponse

 30

Inputs: Nt Respiratory Neural firings

e fchemo Chemoreceptor firings
flung Lung stretch receptors firings
fcs Baroreceptor firings
Offsetres,vein,heart CNS response

Outputs: ftp Total parasympathetic response

ftbs Total β-Sympathetic response
ftas_res,vein Total α-Sympathetic response

Variables: fpara,0 Lower threshold of the parasympathetic baroreflex

sigmoidal function
fpara, Upper saturation of the parasympathetic baroreflex
sigmoidal function
fcs,0 Center point for the sigmoidal function
kp Slope control parameter for the sigmoidal function
GRSA,p Central RSA gain for parasympathetic response
Gchemo,p Chemoreflex gain for parasympathetic response
Glung,p Lung stretch receptor reflex gain for
parasympathetic response
fs,0 Lower limit of the sympathetic exponential decay
function
fs, Upper saturation of the sympathetic exponential
decay function
ks Constant for the exponential function
GRSA,bs Central RSA gain for -sympathetic response
Gchemo,bs Chemoreflex gain for -sympathetic response
Glung,bs Lung stretch receptor reflex gain for -sympathetic
Gbaro,bs Baroreflex gain for -sympathetic
GRSA,as Central RSA gain for -sympathetic response
Gchemo,as Chemoreflex gain for -sympathetic response
Glung,as Lung stretch receptor reflex gain for -sympathetic
Gbaro,as Baroreflex gain for -sympathetic
 31

SA Node (SA_Node_Ursino.mdl)

Description

This module translates changes in -sympathetic and parasympathetic efferent activity

into changes in heart rate. In sleep, the model assumes that the parasympathetic response
increases while there is small decrease in the sympathetic activity. The inputs for this
subsystem are the total -sympathetic firing frequency, ftbs, and parasympathetic firing
frequency, ftp, and the output is the heart period, HP (= reciprocal of instantaneous heart
rate).

SA Node

ftbs
-sympathetic
SI Response
HPbs

ftp
parasympathetic HPp HP
SI Response

Basal Heart
Period HPbasal

SA Node Equation:

HP  HPbs  HPp  HPbasal

Reference: Ursino, M, Interaction between carotid baroregulation and the pulsating

heart: a mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.
 32

Simulink Model: SA Node

3
Gbs(SI )

1 1
ln -0.13
ftbs toux (7).s+1
Transport
Delay Saturation Math Gain _HPbs Transfer Fcn
Function

ftbs_min
HP _basal 1
Constant
HP
Constant 1

2 1
ftp 0.09
Transport toux (8).s+1
Delay41 Gain _HPpara Transfer Fcn 1
Gps(SI )

Inputs: ftbs Total beta-sympathetic firing frequency

ftp Total parasympathetic firing frequency

Output: HP Heart Period (equivalent to RR-interval)

Variable: HPbasal Basal value for HP for denervated heart

HPbs Change in HP modulated by -sympathetic
response
HPp Change in HP modulated by parasympathetic
response
 33

-Sympathetic Control
Description
This response is modeled assuming first-order dynamics. The time-constant and delay
associated with the -sympathetic effect on the heart period is longer than that of the
parasympathetic response. There is slight decrease in -sympathetic response in sleep.
The input for this compartment is the -sympathetic firing frequency, ftbs and the output
is the corresponding component of heart period change, HPbs.

-Sympathetic Control Equations:

Gbs  Gbs ( SI )  ln[ f tbs (t  Dbs )  f tbs min  1], f tbs  f tbs min
 bs (t )  
 0, f tbs  f tbs min
Gbs ( SI )  1  SI  (1  AI )  Gbs _ sleep

d
HPbs 
1
 HPbs (t )   bs (t )
dt  bs

Reference: Ursino, M, Interaction between carotid baroregulation and the pulsating

heart: a mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.

Input: ftbs Total beta-sympathetic firing frequency

SI Sleep Index for sleep wake state
AI Arousal Index

Output: ΔHPbs Heart Period change modulated by -symp.

Variables: Dbs -sympathetic time delay

ftbsIC -sympathetic initial output after time delay
ftbs_min Lower limit for the natural log function
Gbs -sympathetic Gain varied with sleep drive
τbs -sympathetic time constant
delta_HPbsIC Initial input to the -symp first order dynamic
system
Gbs_sleep -sympathetic Gain of sleep factor
 34

Parasympathetic Response

Description

The vagal effect on heart rate is modeled assuming first-order dynamics. During sleep,
parasympathetic activity increases, and this is partially responsible for the decrease in the
heart rate. The input for this compartment is the parasympathetic firing frequency, ftp
and the output is the corresponding component of heart period change, HPp.

Parasympathetic Response Equations:

Gps
 ps (t )   f tp (t  Dps )
Gps ( SI )
d
HPp 
1
 ΔHPp(t)  σp(t)
dt τ para
Gps ( SI )  1  SI  (1  AI )  G para _ sleep

Reference: Ursino, M, Interaction between carotid baroregulation and the pulsating

heart: a mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.

Input: ftp Total parasympathetic firing frequency

SI Sleep Index for sleep wake state
AI Arousal Index

Outputs: ΔHPp Heart Period change modulated by

parasympathetic

Variables: Dpara Parasympathetic time delay

ftpIC Parasympathetic initial output after time delay
Gpara Parasympathetic Gain varied with sleep drive
τpara Parasympathetic time constant
delta_HPpIC Initial input to the parasympathetic first order
dynamic system
Gpara_sleep Parasympathetic Gain of sleep factor
 35

-Sympathetic Control of Peripheral Resistance

(TPR_Ursino.mdl)

Description

This block models -sympathetic control of peripheral vascular resistance, using a first-
order dynamic system as in the case of the -sympathetic component. During sleep in
normals, the accompanying decrease in -sympathetic activity contributes substantially
to a decrease in blood pressure. The inputs are the total -sympathetic firing frequency,
ftas and the state/sleep drive, Dstate. The output is the proportional change in the
peripheral resistance, TPR.

Total Peripheral Resistance

ftas Vascular Resistance

Changes (baro, lung TPR
SI stretch, central, chemo)

Equations for Total Peripheral Resistance Change:

G j  Gas ( SI )  ln[ f tas _ i (t  D j )  f tas min  1], f tas _ i  f tas min

Zj  
 0, f tas _ i  f tas min
d TPR j 1
 (TPR j  Z j )
dt j
TPR j (t )  TPR j  TPR j 0
Gas ( SI )  1  SI  (1  AI )  Gas _ sleep

Reference: Ursino, M, Interaction between carotid baroregulation and the pulsating

heart: a mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.
 36

Simulink Model: Alpha-Sympathetic Modulation on Peripheral Resistance

{Rsp}
1 fes
Rsp
Rsp
Alpha Symp Gas(SI)

(ftas) Peri Circ Rsp

fes
Rep {Rep }
Rep
Gas(SI)
3 1-u Peri Circ Rep
fes
AI (arousal Index) Gas_sleep Rmp_n {Rmpn } 1
Rmp_n
4 1 Gas(SI)
Gas(SI)
1
u
SI (Sleep Index) Gas_sleep Peri Circ Rmp
u
1
fes {Gbp } TPR _change
2 Vusv {Vusv} Gbp u
Vusv
Alpha Symp Gas(SI) Goto
1
(ftas_vein) Venous Circ Vusv {Rmp }
Rmp u
fes
Vuev {Vuev }
Gas(SI) Vuev 1
{Rhp }
Rhp u
Venous Circ Vuev

fes
Vumv {Vumv }
Gas(SI) Vumv
Venous Circ Vumv

Inputs: ftas Total alpha-sympathetic firing frequency

SI Sleep Index for sleep wake state
AI Arousal Index

Outputs: TPR_change TPR change factor

Variables: fasIC -sympathetic initial output after time delay
fas_min Lower limit for the natural log function
Gas_sleep -sympathetic Gain varied with sleep
Gas_sp -sympathetic Gain for splanchnic peripheral resistance
τas_sp -sympathetic time constant
Das_sp Delay -sympathetic time constant
Gas_ep -sympathetic Gain for extra-splanchnic peripheral resistance
τas_ep -sympathetic time constant
Das_ep Delay -sympathetic time constant
Gas_mp -sympathetic Gain for skeletal muscle peripheral resistance
τas_mp -sympathetic time constant
Das_mp Delay -sympathetic time constant
Vusv0 Basal level of unstressed volume of splanchnic venous
circulation
Gas_usv -sympathetic Gain for unstressed volume of splanchnic venous
circulation
τas_usv -sympathetic time constant
Das_usv Delay -sympathetic time constant
 37

Variable Breathing Period (PNEUMA.mdl)

Description

The variable breathing period is controlled from the central neural control system by the
total chemoreflex drive [52]. The inspiratory and expiratory periods of a single breath are
set to be of equal duration. The ventilatory drive is controlled by central and peripheral
chemoreflexes. The combination of ventilatory drive and breathing period determines the
neuromuscular drive.

Reference: Duffin J., R.M. Mohan, P. Vasiliou, R. Stephenson, S. Mahamed, “A model

of the chemoreflex control of breathing in humans: model parameter measurement,”
Respiration Physiology, vol. 120, pp. 13-26, 2000.
 38

Simulink Model: Variable respiratory rhythm generator

Breathing_Enable_Signal

Ventilatory Drive (L/sec)

1
D_VENTILATORY
Breathing_Enable_Signal
2

Breathing_Frequency
Breathing Frequency
(breaths/minute)
F_breathing

D_Vent Variable Breathing Period (sec)

Breathing Period (sec/breath) ENABLE SIGNAL
T_breathing Breathing Period Enable_Breathing_Period

Enable Breathing Signal

VENTILATORY
DRIVE (chemical)1
Breathing Period Enable Breathing Signal

Breathing Period Basal

Breathing_Period_Basal 3.5 Breathing Period Update Variable Breathing Period

Variable Breathing Period Reset

Variable Breathing Period (sec)

RESPIRATORY RHYTHM
VBP Variable Breathing Period Respiratory Rhythm 1

Respiratory Rhythm_Generator
 39

Simulink Model: Ventilatory drive breathing frequency/period

Breathing Period (sec)

1
60 2
u Breathing Period (sec)
T_breathing
Math
Function
1 7500 u BF Breathing Frequency (breaths/minute)
Breathing Frequency (breaths/minute)
D_Vent fcn
-C- para count
1
Control_Constant2 Embedded F_breathing
MATLAB Function BF_BP Scope1
 40

Variable Heart Period (PNEUMA.mdl)

Description

The variable heart period module is modulated by the major reflexes and
cardiorespiratory interactions in a closed loop mode. The sinoatrial node is modeled as a
simple pacemaker, regulated by the parasympathetic and the beta-sympathetic inputs. The
variable heart period is generated from continuous SA output using an
integration/saturation mechanism. The beta-sympathetic branch affects the heart rate
contractility, thus modulating the systolic period. Greater beta-sympathetic tone increases
myocardial elastance and shortens ventricular systole. Each active atria-ventricular
compartment is characterized by a time-varying nonlinear elastance function, describing
the changes in ventricular elastance due to the beta-sympathetic tone input. The diastolic
filling time is the difference between the heart period and systolic period and is thus
controlled indirectly. The activation of the right and left hearts is fully synchronized and
occurs simultaneously.

Reference: Dempsey, J.A., Smith, C.A., Eastwood, P.R., Wilson, C.R., Khoo, M.C.K.
Sleep induced respiratory instabilities. In: Pack, A.I. (Ed.), Sleep Apnea Pathogenesis,
Diagnosis and Treatment. Dekker M., New York. 2002.
 41

Simulink Model: Variable Heart Period

1
Threshold Integrated HP
1
1
u u1 if (u1 <0)
HP_SAnode 1 HPin 1
s If Heart_Period

Integrator
1
1 if { }
Time Constant In 1 Out1 du/dt |u|
s HPout Heart Period
If Action
Subsystem

Integrator 1
Enable signal

Memory
 42

Cardiovascular System (PNEUMA.mdl)

Description

The cardiovascular subsystem is capable of simulating the pulsatile nature of the heart
and blood flow through the pulmonary and systemic circulations. Included in the model
are descriptions of atria-ventricular mechanics, hemodynamics of the systemic and
pulmonary circulations, SA node, change of total peripheral resistance and baroreflex.
The inputs for this combined subsystem are the α-sympathetic firing rates, ftas_res,vein,
β-sympathetic firing rates, ftbs, parasympathetic firing rate, ftp, arterial PaCO2, CaO2,
arousal index, AI, sleep-wake state index, SI, and the pleural pressure, Ppl. To
incorporate the effects of pleural pressure changes on the circulatory system we modulate
basal blood pressure values for systemic and pulmonary components in thoracic cavity
and heart. The output is the arterial blood pressure, ABP, heart period, HP, cardiac
output, CO, and blood flow to lung for gas exchange.

Cardiovascular System

Pulmonary R PA RPC RPV

Circulation C PC C PV

LPA pLA
CLA
R SV R SP
C SV CSP
R LA
QRV CPA
pLV
R EV R EP CLV
C EV CEP
R RV RLV

QMP QLV
CRV

RRA R MV R MP
C MV CMP
C SA
LSA
QBP
pRA
CRA R BV R BP
CBV C BP
RSA
Systemic
QHP pSP Circulation

R VC R HV R HP
CHV C HP
C VC
 43

Reference:

1. Ursino, M. Interaction between carotid baroregulation and the pulsating heart: a

mathematical model. American Journal of Physiology, 275:H1733-H1747, 1998.
2. Ursino, M., Magosso, E. Acute cardiovascular response to isocapnic hypoxia. I. A
mathematical model. American Journal of Physiology – Heart and Circulatory
Physiology, 279, H149-165, 2000.
 44

Simulink Model. Cardiovascular System TPR Change

2 1 Alpha Symp (ftas) {Ppa } {Pla } {Vra } {Vla }

fcs ftas _res

2 Alpha Symp (ftas_vein )
ftas _vein {Vpa } {Pra } {Vrv} {Vlv }
AI (arousal Index )
8 SI (Sleep Index ) {Vpp } {Wlv} {RHeart } {LHeart }
SI Sleep -Wake
State Index
SI Sleep Wake State Index {gPpl } {Wrv} {Vpv } {Vu }

Carotid
3 Beta-Symp (ftbs)
Baroreceptor {Rmpn } {Vusv} {Vuev } {Vumv }
ftbs HP HP HP
4 Parasymp (ftp) {Gbp } {Rhp } {Rmp } {Rsp}
ftp
Arousal _CardIC {Rep } {Vsa} {Psp} {Pvc }

SA-node and Autonomic Control

1
7
9
AI
G_pleural Ppl
ABP
Arousal Index {gPpl }

1
Qrv Blood Flow to Lung
PULMONARY CIRCULATORY SYSTEM for Gas Exchange
Qla

RIGHT HEART LEFT HEART

PaCO2
5
CaO2
Qra Qlv
Local Flow Regulation 6
Qmp

Qbp SYSTEMIC CIRULATORY SYSTEM

Qhp
 45

Simulink Model. SA-Node and Autonomic Control

Vusa
2 f tbs
{Emaxlv} {Emaxrv}
Emaxlv {Emaxlv}
Vusp Beta-Symp (ftbs) Gbs(SI)
Heart Emaxlv
Vuep
Heart Beat
f tbs
[Vusv] Emaxrv {Emaxrv} Right Ventricle
Gbs(SI)
phi Pmaxrv {RHeart}
[Vuev] Heart Emaxrv
Right Ventricle
Vupv
f tbs Heart Beat
Vupa
Left Ventricle
3 f tp
Vupp HP HP phi phi Pmaxlv {LHeart}
{Vu}
Parasymp Gbs(SI)
Vula (ftp) Gps(SI)
phi Left Ventricle

Vura SA Node

Vump
4 1-u
Vubp Gbs_sleep
Arousal_CardIC HP_SAnode Heart_Period 1
Vuhp
1 HP
1
Vubv SI Sleep Wake State Index
RR Interval
Vuhv Gpara_sleep

[Vumv] 1

Simulink Model. Left Heart

Left Heart Vula (Left Atrium Unstressed Volume)
Vulv (Left Ventricle Unstressed Volume)
Vula Vlv (Left Ventricle Volume)
Vla (Left Atrium Volume)
_ Pla (Left Atrium Pressure)
Out 1/Cla Plv (Left Ventricle Pressure)
Qla 1 Pla
1 + {Pla} Qla (Flow into Left Atrium)
s
Qla Qilv (Flow into Left Ventricle)
Qolv (Flow out of Left Ventricle)
{Vla} LHeart (Left Ventricle Function)
[gPpl]

Qilv
1
1/Rla s SV
[ABP]
mitral
Plv [LHeart]
valve

[HP] -K- CO
Pmaxlv
ml/sec
Pmax to l/min
P
Flow 1
Vulv Qolv
[ABP] Pmaxlv--Psa
Qlv
aortic
_ ----------------------
valve
1 Out {Vlv} Rlv
+
s
Qolv {Wlv}
 46

Simulink Model. Right Heart

Right Heart Vura (Right Atrium Unstressed Volume)
Vurv (Right Ventricle Unstressed Volume)
Vrv (Right Ventricle Volume)
Vra (Right Atrium Volume)
Vura
Pra (Right Atrium Pressure)
_ Prv (Right Ventricle Pressure)
Out 1/Cra {Pra} Qra (Flow into Right Atrium)
+ Pra
Qra 1 Qirv (Flow into Right Ventricle)
1
s Qrv (Flow out of Right Ventricle)
Qirv
LHeart (Left Ventricle Function)
{Vra}
[gPpl]

CHF Gain1

1/Rra 1
Qirv [Ppa]
Prv
tricuspid Pmaxrv
[RHeart]
valve
Switch2

Pmaxrv Pmax
P
Flow 1
Vurv
Qrv
[Ppa] Pmaxrv--Ppa
_ pulmonary
----------------------
Qirv Out {Vrv} valve
Rrv
1 +
s
{Wrv}

Product

Simulink Model. Systemic Circulation

Systemic Circulation

Qvc Qvc Qra 1

Qra
Qmp
1 Qlv Qsa Qsa Vena Cava
Qbp
Qlv
Qhp 2
Systemic Qmp
Arteries Systemic 3
Peripheral & Qbp {Vvc}
Venous 4
Circulation Qhp
 47

Simulink Model. Pulmonary Circulation

Pulmonary Circulation Ppa (Pulmonary Arteries Pressure)

Ppp (Pulmonary Peripheral Pressure)
Ppp Ppv (Pulmonary Veins Pressure)
Pla (Left Atrium Pressure)
Vupa {Ppa} Qor (Flow from Right Heart)
_ Qpa (Flow to Pulmonary Aorta)
Qrv Out 1/Cpa Qla (Flow to Left Atrium)
1 + Ppa Vupa (Pulmonary Arteries Unstressed Volume)
1
Qrv Vupp (Pulmonary Peripheral Unstressed Volume)
Qpa s
Vupv (Pulmonary Veins Unstressed Volume)
{Vpa}
Qpa
Rpa
[gPpl]

Qpa 1
1/Lpa
s

{Vpp}
Vupp
Qpa _

1 Out 1/Cpp
+ Ppp
Qpp s
1
Ppp

-K- 1/Rpp

ml/liter Vupv [gPpl]

_
Out 1/Cpv
+ Ppv
1
Qpv s
{Vpv}

Qpv
2 1/Rpv Pla
Qpv [Pla]
Qla

Simulink Model. Local Flow Regulation

4 CaO2
CaO2 PaCO2
6 Qbp Gbp {Gbp}
Qbp Gsleep Brain Peripheral
Cerebral Circulation Resistance
Regulation
3
PaCO2
CaO2
PaCO2
5 Qmp Rmp {Rmp}
Qmp Gsleep Muscular Peripheral
Muscular Circulation Resistance
Regulation

CaO2
PaCO2
7 Qhp Rhp {Rhp}
Qhp Gsleep Coronary Peripheral
Coronary Circulation Resistance
Regulation
1 1-u
AI (arousal Index)
-K-
2
1 Gas(SI)
SI (Sleep Index) Gas_sleep
 48

Neuromuscular Drive (NeuroMuscular.mdl)

Description
Inspiratory muscular activity is produced by neural drive arising from the respiratory
centers. The muscles have to overcome the resistive and elastic forces of the lungs and
chestwall to generate the airflow. The muscular drive is modulated by the
autorhythmicity, chemical and state drives. In the case of the mechanical assisted
ventilation, the internal neural activity will diminish with a period of time. The inputs for
this compartment are the chemical drive, Dchemo, external pressure, Dext and the state-
related drive, Dstate. The output is the neuromuscular drive, Nt.

Neuromuscular Drive

Chemical Drive, State Drive Neuromuscular

Respiratory Autorhythmicity
 Nt

External Assisted Pressure

Drive Profile

Neuromuscular Drive Equations:

Respiratory Autorhythmicity  SquareFunc(TI , TT )

 T
 I Dtotaldt 0  t  TI
N (t )   0

0 TI  t  TT
 49

Simulink Model: Neuromuscular Drive

Resp_Rhythm_Generator
Demux on_off_rhythm_test Mux

Nt Save block 1
RespMus_blocker G_RespMus
Mechanial
1 1 20 Resp Neural Profile Resp_sig Mux Ventilation
RespMus Drive
2
Chemical S_wake State Total Drive
Drive drive
3 0.3 2

Mux f(u)

Inputs: Dchemo Chemical Drive

Dext External drive or pressure
Dstate State related Drive

Output: Nt Neural-Muscular Drive

Variables: Gstate State Drive gain

TTmean Breathing Period
TImean Inspiration Period
Inhale Boolean variable for inhalation
 50

Respiratory Muscle Activity (Pmus_Flow_Younes.mdl)

Description

During the breathing process, the respiratory muscles have to overcome the resistive and
the elastic forces of the respiratory system. By equating the force generated from the
respiratory muscles with the pressure from the respiratory system, the airflow pattern can
be obtained using a simple mechanics model, and tidal volume can be computed from
the flow. During normal breathing, expiratory muscular activity is minimum. The inputs
are the neural signals, Nt, the upper airway conductance, Condua, the expiratory pressure,
Pexp and the external pressure, Pao. The outputs are the airflow, Flow, tidal volume, Vt
and the muscular pressure, Pmus.

Respiratory Muscle Activity

Pexp Pao
Inspiratory
Nt Muscular
Pressure

Vt, Flow, Pmus

Respiratory
Condua Resistive,
Elastic Force

Respiratory Mechanics Equations:

Pisom = Gneuromusc DTotal

Yua
Yrs 
1  ( RAW  RLT  RCW )  Yua
Note: when Yua=0, then Yrs = 0.
Vt / 0.28VC
. (0.25Pisom ( t ) e  Yrs  bVt  Vt ErsYrs  PaoYrs )2  4bVt ( Pisome Vt /0.28VCYrs  Vt ErsYrs  PaoYrs )
Vt 
2
Vt / 0.28VC
0.25Pisom ( t ) e  Yrs  bvt  Vt ErsYrs  PaoYrs

2
 51

.
Pisom e Vt / 0.28VC (bVt  0.25 Vt )
Pmus  .
(Vt  bVt )
.
PPL  RCW Vt  ECWVt  Pmus
.
Palv  RLT Vt  ELTVt  PPL
. (0.25GP(t )e Vt / 0.28VC  bV  GVt E  GPE  GPAO ) 2  4bV (GP(t )e Vt / 0.28VC  GVt E  GPE  GPAO )
Vt 
2
Vt / 0.28VC v
0.25GP(t )e  b  GVt E  GPE  GPAO

2

Reference: Younes, M. and Riddle W. A model for the relation between respiratory
neural and mechanical outputs. II. Methods. Journal of Applied Physiology, 51(4): 979-
989, 1981.

Simulink Model: Respiratory Muscle Activity and Flow Generation

 52

Inputs: Nt Neural-Muscular Drive

Condua Upper Airway conductance
Pexp Expiratory Pressure
Pao External Pressure
Outputs: Vt Lung Volume
Flow Air flow
Pmus Muscle Pressure
Variables: Flowo Initial air flow
tau_resp Inspiratory muscle response time
delta_t Integration step time
VC Vital Capacity
Vo Initial lung volume
pt_frcIC1 Initial condition for respiratory muscle reaction
pt_frcIC2 Initial condition for respiratory muscle reaction
FlowIC Initial condition for airflow
VtIC Initial condition for lung volume
 53

Pleural Pressure(Pleural_Schuessler.mdl)
Description
Pleural pressure influences the arterial blood pressure by increasing the venous return and
decreasing the cardiac output. The combination of the respiratory muscle force and the
static chest wall pressure yields pleural pressure. The inputs are airflow, Flow, muscular
pressure, Pmus and external pressure, Pao. The output is the pleural pressure, Ppl.

Pleural Pressure

Flow
Pleural
Pmus Pressure Ppl
Pao Mechanisms

Pleural Pressure Equation:

.
 .  . PVt (bVt  0.25 Vt )
 
PPL  PAO  K1, AW  K 2, AW Vt Vt 
  .
  (Vt  bVt )
.
 PE  RCW Vt  ECW Vt

Reference: Schuessler, T.F., Gottfried, S.B. and Bates, J.H.T. A model of the
spontaneously breathing patient: applications to intrinsic PEEP and work of breathing.
Journal of Applied Physiology, 82(5): 1694-1703, 1997.
 54

Simulink Model: Pleural Pressure Simulink Model: Chest Wall Mechanics

Simulink Model: Airway Pressure

Inputs: Flow Air flow

Pmus Inspiratory muscle pressure
Pao External Pressure

Output: Ppl Pleural Pressure

Variables: Rcw Chest Wall resistance

Ecw Chest Wall elastance
k1,aw Constant for upper airway pressure
k2,aw Constant for upper airway pressure
 55

Gas Exchange and Transport (Gas_Exchange.mdl)

Description

This subsystem models gas transport through the dead space, CO2 and O2 exchange in the
alveoli, the CO2 and O2 dissociation curves, and the transport of CO2 and O2 in the blood
to the chemoreceptors along with vascular mixing. Also included in this module are CO2
exchange in the brain compartment, gas exchange in the body tissues, conversion of
blood gases into respiratory drive by the chemoreflexes, and chemoreflex effects on
peripheral vascular resistance. The inputs are airflow, Flow, tidal volume, Vt and cardiac
output, CO (in this case, it is synonymous with blood flow, Q. The output is the
chemoreflex-related ventilatory drive, Dchem and chemoreflex modulation of total
peripheral resistance, TPRchemo.

Gas Exchange and Transport

PbCO2
Flow Central Chemoreceptors
Vt Dead
Space
Q
Brain Region Dc
Dchemo

Dp
PdCO2 PdO2 PaCO2

Gas Exchange at PaCO2 Peripheral Chemoreceptors

the Lungs PACO2 Cardiovascular
Mixing, SaO2
PAO2
Convection and
Dissociation Flow TPRchemo
PaCO2
Regulation
PaO2
CvCO2 CvO2 CaCO2 CaO2
Body Tissues
Part
 56

Simulink Model. Gas Exchange

PACO2 PbCO2 f(u)

Pd5CO2 Qb

PAO2 VARIATION OF CEREBRAL Qb

1
Pd5O2 BLOOD FLOW W / PaCO 2
PaCO 2
3 Flow -C- Lung-Chemoreceptor Delay Volume Fcn
Air Flow PaCO2 PaCO2 PbCO2 4
DEAD SPACE Lung -Chemo Volume
PACO2
PaCO 2 PbCO 2

PAO2 SI Sleep Wake State Index

PaO2 2
Q PaO 2 BRAIN COMPARTMENT
Cardiovascular
Mixing and Convection

2
CaCO2 SI Sleep Wake State Index
CVCO 2
PACO2
Pd5CO2 PACO2

Pd5O2
CaCO2
Flow PACO2 SAO2 SI Sleep Wake State Index
1 Vt - Tidal Volume CvCO 2
SAO 2 3
Vt - Tidal Volume CaCO2
Q SAO 2
PAO2 PAO2
PAO2 CaO2 CaO2
CVO2
CvO 2
CaO2
5 Qt
Dissociation
CaO 2
GAS EXCHANGE BODY TISSUES
IN THE LUNGS COMPARTMENT
Blood Flow to Tissues
0.85

4
Blood Flow
 57

Dead Space (Dead_Space_Khoo.mdl)

Description

We assume that no gas exchange with blood occurs in the dead space. The inputs are
airflow, Flow, tidal volume, Vt and blood flow, Q. The outputs are the CO2, PdCO2 and
the O2, PdO2 partial pressure for the dead space.

Dead Space

Flow
PdCO2
Vt Dead Space
CO For CO2 and O2 PdO2

Dead Space Equations:

CO2
Inspiration
. .
Vd (1) P d (1) CO  V [ PI CO  Pd (1) CO ]
2 2 2
. .
Vd (i ) P d (i )CO  V [ Pd (i 1)CO  Pd (i )CO ] 2i5
2 2 2
Expiration
. .
Vd (i ) P d (i )CO  V [ Pd (i 1)CO  Pd (i )CO ] 1 i  4
2 2 2
. .
Vd (5) P d (5)CO  V [ PACO  Pd (5)CO ]
2 2 2
O2
Inspiration
. .
Vd (1) P d (1) O2  V [ PI O  Pd (1) O ]
2 2
. .
Vd (i ) P d (i )O2  V [ Pd (i 1)O  Pd (i )O ] 2i5
2 2
Expiration
. .
Vd (i ) P d (i )O2  V [ Pd (i 1)O  Pd (i )O ] 1 i  4
2 2
. .
Vd (5) P d (5)O2  V [ PAO  Pd (5)O ]
2 2

Reference: Khoo, M.C.K., A model-based evaluation of the single-breath CO2

ventilatory response test. Journal of Applied Physiology, 68(1):393-399, 1990.
 58

Simulink Model: Entire Dead Space Dead Space Compartments for CO2
subsystem

Individual Dead Space Compartment for CO2

Note: Dead Space Compartments for CO2 and O2 designs are the same. Only Part of
CO2 implementations are shown as examples.
 59

Inputs: Flow Air flow

Vt Lung Volume
CO Cardiac Output

Outputs: PdCO2 Dead Space CO2 partial pressure

PdO2 Dead Space O2 partial pressure

Variables: Dead(i),co2I Initial condition for ith CO2 dead space

C
Dead(i),o2I Initial condition for ith O2 dead space
C
Vd(i) ith dead space volume
PI,CO2 Inspiratory CO2 partial pressure
PI,O2 Inspiratory O2 partial pressure
 60

Alveolar Gas Exchange (Lungs_Khoo.mdl)

CO2 and the O2 exchange in the lungs are both modeled assuming first-order dynamics.
The rate of exchange is affected by the gas concentration in the blood, the gas partial
pressure and the blood flow rate. The CO2 storage space is larger than that for O2 to
account for the larger capacity of lung tissue and lung water for CO2. The inputs are the
CO2, PdCO2 and O2, PdO2 partial pressure for dead space, arterial CO2, CaCO2 and O2, CaO2
concentration, venous CO2, CvCO2 and O2, CvO2 concentration, tidal volume, Vt, airflow,
Flow and blood flow, Q. The outputs are alveolar CO2, PACO2 and O2, PAO2 partial
pressure.

Gas Exchange in the Lungs

PdCO2, CaCO2 CO2 exchange in the

Lungs PACO2
CvCO2
Vt, Flow

Q
PdO2, CaO2 O2 exchange in
the Lungs PAO2
CvO2

Gas Exchange in the Lungs Equations:

Inspiration
. . .
Vco2 P Aco2  [863 Q(Cvco2  Caco2 )  V A ( Pd (5)co2  PAco2 )]
. . .
Vo2 P Ao2  [863 Q(Cvo2  Cao2 )  V A ( Pd (5)o2  PAo2 )]

Expiration
. .
Vco2 P Aco2  [863 Q(Cvco2  Caco2 )]
. .
Vo2 P Ao2  [863 Q(Cvo2  Cao2 )]

Reference: Khoo, M.C.K., A model-based evaluation of the single-breath CO2

ventilatory response test. Journal of Applied Physiology, 68(1):393-399, 1990.
 61

Simulink Model: Gas Simulink Model: Gas Exchange in the Lungs for CO2
Exchange in the Lungs

Note: O2 compartment is the same implementation as CO2 and it is not shown here.

Inputs: PdCO2 Dead Space CO2 partial pressure

PdO2 Dead Space O2 partial pressure
CaCO2 Arterial CO2 concentration
CaO2 Arterial O2 concentration
CvCO2 Venous CO2 concentration
CvO2 Venous O2 concentration
Vt Tidal volume
Flow Airflow
CO Cardiac output

Outputs: PACO2 Alveolar CO2 partial pressure

PAO2 Alveolar O2 partial pressure

Variables: Vco2 Lungs storage volume for CO2

Vo2 Lungs storage volume for O2
PAco2IC Initial condition for Partial CO2 pressure
PAo2IC Initial condition for Partial O2 pressure
s Pulmonary shunt
lambda Concentration / Pressure conversion
 62

Cardiovascular Mixing, Convection and Dissociation

(Cardio_Mix_Lange.mdl, Dissociation_Spencer.mdl)
Description
This module includes effects for pulmonary shunt, convection and mixing in the heart
and vasculature, as well as the delay taken for arterial blood to travel from the gas
exchange site to the chemoreceptors. The mixing and convection processes are affected
by the blood flow rate and are modeled assuming a second order dynamic system. Partial
pressures are converted to gas concentration in the blood, using the blood-gas
dissociation equations. The inputs for this compartment are the alveolar CO2, PACO2 and
O2, PAO2 partial pressure. The outputs are the arterial CO2, PaO2 and O2, PaO2 partial
pressure.

Cardiovascular Mixing, Convection and Dissociation

Cardiovascular mixing PaCO2

and convection for
CO2
PACO2

Dissociation CaCO2

Cardiovascular mixing PaO2

and convection for O2
PAO2

Dissociation CaO2

Cardiovascular Mixing Equations:

.. 1 .
P aCO2  [ PACO2 (t  Ta )  (T1  T 2) P aCO2  PaCO2 ]
(T1 * T 2)
.. 1 .
P ao2  [ PAo2 (t  Ta )  (T1  T 2) P ao2  Pao2 ]
(T1* T 2)

Reference: Lange, R.L., Horgan, J.D., Botticelli, J.T., Tsagaris, T, Carlisle, R.P., and
Kuida.H., Pulmonary to arterial circulatory transfer function: importance in respiratory
control. Journal of Applied Physiology, 21(4):1281-1291, 1966.
 63

Simulink Model: Cardiovascular Mixing

CO2 Cardiovascular Mixing
PACO2 and Convection Effects
2
PACO2

4 PaCO2
1
Q
Q PaCO 2
1 Lung-Chemoreceptor Delay Volume
Lung -Chemoreceptor Delay Volume
Q

3 PAO2 PaO2 2
PAO2 PaO 2
Lung-Chemoreceptor Delay Volume
O2 Cardiovascular
Mixing and Convection
Effects

Simulink Model: Cardiovascular Mixing for CO2

PaCO 2secondIC

T 1+T2

1
1 1 PaCO 2
1/(T1*T2)
s s

PaCO 2firstIC

PACO2
1
PACO2 delay

3
Lung -Chemoreceptor Delay Volume To
2 Product 1
Q
PACO2_delayIC

Note: O2 implementation is the same as CO2.

Cardiovascular Convection Equation:

K dp
Ta 
Q

Inputs: PACO2 Alveolar CO2 partial pressure

PAO2 Alveolar O2 partial pressure

Outputs: PaCO2 Arterial CO2 partial pressure

PaO2 Arterial O2 partial pressure

Variables: Kdp Peripheral Chemoreceptors delay time constant

T1 Time constant for cardiovascular mixing
T2 Time constant for cardiovascular mixing
PaO2firstIC Initial condition for first order Pao2 system
PaO2secondIC Initial condition for second order Pao2 system
 64

PaCO2firstIC Initial condition for first order Paco2 system

PaCO2secondI Initial condition for second order Paco2 system
C
PaO2_delayIC Initial condition for O2 convection
Paco2_delayIC Initial condition for CO2 convection

Dissociation Equations:

1 / a1
_ FO
CO 2  C O 2 2
1 / a1
1  FO
2

1 / a2
_ FCO
CCO2  C CO2 2
1/ a
1  FCO 2
2

PAO (1  1 PACO )
FO2  2 2
K1 (1  1 PACO )
2

PACO (1   2 PAO )
FCO2  2 2
K 2 (1   2 PAO )
2

Reference: Spencer, J.L., Firouztale, E., and Mellins, R.B. “Computational Expressions
For Blood Oxygen and Carbon Dioxide Concentrations”, Annals of Biomedical
Engineering, Vol 7, pp. 59-66, 1979.

Simulink Model: Dissociation

 65

Inputs: PACO2 Alveolar CO2 partial pressure

PAO2 Alveolar O2 partial pressure

Outputs: PaCO2 Arterial CO2 partial pressure

PaO2 Arterial O2 partial pressure
Z Molar conversion factor
C1 Maximum concentration of hemoglobin-bound
oxygen
C2 Maximum carbon dioxide concentration
a1 Parameter in O2 dissociation equation
a2 Parameter in CO2 dissociation equation
alpha1 Parameter in O2 dissociation equation
alpha2 Parameter in CO2 dissociation equation
K1 Parameter in O2 dissociation equation
K2 Parameter in CO2 dissociation equation
beta1 Parameter in O2 dissociation equation
beta2 Parameter in CO2 dissociation equation
Saco2_delayI Initial Condition for Oxygen Saturation Delay
C
 66

Brain Compartment (Brain_Khoo.mdl)

Description

Cerebral flow is highly sensitive to changes of the CO2 tension in the brain. The brain
CO2 tension is controlled by the metabolic rate and the blood flow rate. The inputs for
this compartment are the arterial CO2 partial pressure, PaCO2 and blood flow in the brain
region, Qb. The output is the brain arterial CO2 partial pressure, PbCO2.

Brain Tissues and Cerebral Flow

PaCO2 Brain CO2 interaction PbCO2

and cerebral flow

Cerebral Flow Equation:

. .
S bCO2 P bCO2  [ MRbCO2  Qb S CO2 ( PaCO2  PbCO2 )  h]

Reference: Read, D.J.C. and Leigh, J. Blood-brain tissue Pco2 relationships and
ventilation during rebreathing. Journal of Applied Physiology, 23(1):53-70, 1967.

Simulink Model: Brain Tissues

Brain Tissues Equation:

.
. . . .
Q b 2  [1  0.03( Pbco2  40)] Qb0 Q b  0.03(MRbco2  h) Qb0 / S co2  0
Reference: Khoo, M.C.K., A model-based evaluation of the single-breath CO2
ventilatory response test. Journal of Applied Physiology, 68(1):393-399, 1990.
 67

Simulink Model: Variation of Cerebral Blood Flow

Input: PaCO2 arterial CO2 partial pressure

Output: PbCO2 brain arterial CO2 partial pressure

Variables: MRbco2 Metabolic production rate for CO2 in the brain

tissue
Sco2 Dissociation slope for CO2 in the blood
Sbco2 Dissociation slope for CO2 in the brain tissue
Pbco2IC Initial condition for partial CO2 pressure from the
brain
 68

Body Tissues Compartment (Body_Khoo.mdl)

Description

Gas exchange that occurs outside of the lungs and the brain is modeled as taking place in
a single compartment. The rates of O2 consumption and CO2 production are dependent on
the metabolic rate of the body tissues. The inputs are the arterial O2 and CO2
concentrations, CaO2 and CaCO2. The outputs are the venous O2 and CO2 concentrations,
CvO2 and CvCO2.

Body Tissues

CaCO2 CvCO2
Body tissues exchange
CaO2 for O2 and CO2 CvO2

Body Tissues Equations:

. .
Vt CO2 C VCO2  [ MRCO2  Q(C aCO2  CVCO2 )]
. .
Vt O2 C VO2  [ MRO2  Q(CaO2  CVO2 )]

Reference: Khoo, M.C.K., A model-based evaluation of the single-breath CO2

ventilatory response test. Journal of Applied Physiology, 68(1):393-399, 1990.

Simulink Model: Body Tissues Simulink Model: Body Tissues for CO2
 69

Note: O2 compartment is the same as CO2 compartment

Inputs: CAO2 Arterial O2 concentration

CACO2 Alveolar CO2 partial pressure

Outputs: CvCO2 Arterial CO2 partial pressure

CvO2 Oxygen Saturation

Variables: Vtco2 Body tissure storage volume for CO2

Vto2 Body tissure storage volume for O2
MRco2 Metabolic production rate for CO2
MRo2 Metabolic consumption rate for O2
Cvco2IC Initial condition for mixed venous CO2
concentration
Cvo2IC Initial condition for mixed venous O2
concentration
 70

Ventilatory Response (Vent_Drive_Khoo.mdl)

Description

The chemical driven ventilatory response is determined from the central and the
peripheral chemoresponses during sleep-wake state. The central response is driven
primarily by CO2 while the peripheral response is modulated both by oxygen and carbon
dioxide. The inputs for this compartment are the brain CO2 partial pressure, PbCO2,
arterial CO2 partial pressure, PaCO2 and oxygen saturation, SAO2. The output is the
chemical drive for ventilation, Dchem.

Ventilatory Response

PbCO2
PaCO2 Chemical drive DTotal
for ventilation
SAO2
SI

Ventilatory Response Equations:

(a) For normal breathing,

 Y, TH L  Y  TH H
DTotal  
 0, OW
 z p 0 t
Y  X / 2  X / 2  [ z p0  e  u(t )]

2 /(1  e
5 DTotal _ O
)  1, DTotal _ O  0
X 

 0, DTotal _ O  0
(b) For sleep-disordered breathing,

DTotal  DTotal _ O
where

DTotal _ O  (1  0.4 SI ) ( Dvent  Dstate )

Dstate  SI  Swake
Dvent  Dc  D p
G ( P  I ), P  I C
DC   C bCO2 C bCO2
0, Otherwise
G ( P  I pCO2 )  ( I pO2  SAO2), PaCO2  I pCO2 &I pO2  SAO2
DP   P aCO2
 0, Otherwise
 71

Reference: Khoo, M.C.K., A model-based evaluation of the single-breath CO2

ventilatory response test. Journal of Applied Physiology, 68(1):393-399, 1990.

Simulink Model: Ventilatory Drive

SI

2 1
0.075 1 (u>0)*u
PbCO2 Ic
Gc_blocker
Gc 0.3
S_wake

6 Mux f(u)
PaCO2 IpCO 2 4
1 1
Sleep /Awake 5
D_Total
1 for up
0.0063 1 (u>0)*u AI Dchemo 0 for ground
IpO 2 Gp Gp_blocker Switch

3
SAO2 In1 Out1

Dynamic Drive

Inputs: PbCO2 Brain CO2 partial pressure

PaCO2 Arterial CO2 partial pressure
SAO2 Oxygen saturation
SI Sleep-wake state index

Output: DTotal chemical drive for ventilation

Variables: Ic Central apneic threshold

IpCO2 Peripheral apneic threshold for CO2
IpO2 Peripheral apneic threshold for O2
Gc Gain for central chemical drive
Gp Gain for peripheral chemical drive
Swake Factor of wakefulness to sleep
 72

Upper Airway / State Change

(State_UA_Khoo_Borbely.mdl)

Description

Upper airway muscle tone decreases from wakefulness to sleep. This introduces the
possibility of upper airway collapse under certain conditions. The simple model of upper
airway mechanics employed here assumes that upper airway conductance (= reciprocal of
resistance) is directly proportional to the "wakefulness" (or state-related ventilatory)
drive.

Upper Airway and State Change Interactions

Awake / Sleep
State Change SI
Mechanism

SSWA SI
Saw/sleep AI

Pfrc Upper Airway

Yua
Insp Mechanism
Pao
 73

Upper Airway

Description

Upper airway model is driven by pleural pressure, Ppl, total respiratory flow, Qtotal in the
airways, lower airway resistance, Rla and sleep-wakefulness state drive, SD. During the
obstruction, the upper airway is narrowed, therefore the upper airway resistance to the
airflow increases. Because the upper airway is entirely blocked during full obstruction,
and its resistance becomes infinitely large, for modeling purposes we prefer to use the
upper airway conductance, Yua which is the inverse of resistance. We model upper
airway conductance as a function of upper airway opening surface area, A. It is a known
fact that in patients with Obstructive Sleep Apnea the upper airway muscle tone is
reduced and more prone to collapse. Therefore, the upper airway opening surface area
depends on the airway pressure and upper airway compliance, Cua that is in turn a
function of upper airway sensitivity, sua and also depends on the sleep-wakefulness state
drive SD. The upper airway muscle tone is represented by the upper airway sensitivity. In
wakefulness, the sensitivity remains low, but with the sleep onset the sensitivity
increases. The net effect is to impose an additional load on respiratory effort.
All these mechanisms are inter-dependent on each other and connected to lower
respiratory airways as well in a closed loop mode.

Upper Airway Equations:

P 
 

ao
 Pua  Y ua  V V ua

where Yua = 1 / Rua

Pcrit ( SI )  

bua 
Pua   V ua dt  V ua  Ruaw

 0, Pua  Pcrit

Yua ( SI )  kua  Aua , where Aua   A0ua  (1  Pua / Pcrit ( SI )), Pcrit  Pua  0
A , Pua  0
 0ua

 Pcrit _ awake , SI  0 (awake)


Pcrit ( SI )   Pcrit _ awake /(1  Sua  ( SI ) 2 ) /(1  Sleepawake),0  SI  1,
P
 crit _ awake /(1  Sua ), SI  1 (sleep)
(A.40)
where Sleepawake is 0 during sleep and 1 during wakefulness, Sua is upper airway
sensitivity and is directly related to Pcrit .
 74

Simulink Model: Upper Airway Mechanism

Enable

C_ua Upper Airway Copmliance

C_ua_Upper_Airway_Compliance Y_ua_Upper Airway Conductance 1
Y_ua
Upper Airway Conductance
P_crit
P_ua_Upper Airway Pressure P_crit

Upper Airway Conductance3

SI Sleep Wake State Index

C_ua Upper Airway Copmliance SI Sleep Wake State Index
C_ua Upper Airway Compliance 1
SI Sleep Wake State Index
Sleep/awake SI Sleep Wake State Index SI Sleep Wake State Index
6 Ventilatory Drive
Upper_Airway_Compliance 7 Total Ventilatory Drive 2 Respiratory Rhythm
Sleep/awake Respiratory Rhythm P_pl Pleural Pressure
P_ua
C_ua Upper Airway Copmliance
C_ua Upper Airway Compliance P_ua vs. P_cirt
Q_ua Upper Airway Flow Tidal Volume
Q_ua Upper Airway Flow 5
Tidal Volume
P_ua Pressure in Upper Airways Q_ua Upper Airway Flow
Q_la Lower Airway Flow
Flow in Upper Airway C_ua Upper Airway Copmliance
Y_ua Upper Airway Conductance

P_pl Pleural Pressure 4

P_pl Pleural Pressure Upper Airway Scope
P_ua P_ua Pleural Pressure
P_ua Upper Airway Pressure
Q_la Lower Airway Flow
Q_la Lower Airway Flow

Pressure in Upper Airway

Q_ua Upper Airway Flow

Q_total Q_la Lower Airway Flow

Total Respiratory Flow Airways Q_total 1
3
Coefficient 1

Q_total
 75

Inputs: SI Sleep-wake State Drive

Ppl Pleural Pressure
Qtotal Total Respiratory Flow
Rla Lower Airway Resistance
Sleep/Awake Sleep or Awake state
DTotal Total ventilatory drive
Vt Tidal volume
Resp_Rhm Respiratory rhythm

Outputs: Yua Upper Airway Conductance

Variable: Sua Upper Airway sensitivity

Ruaw Upper airway wall resistance
A0ua Maximum area of opening in upper airway
Kua Proportionality coefficient between Aua and
Yua;
Pcrit_awake Critical upper airway pressure in wakefulness
Cua Upper airway compliance
Pua Upper airway pressure


V ua Upper airway flow



V Total flow in airways

Sleep Mechanism

In the sleep mechanism model, the awake/sleep state is determined by a combination of

the circadian rhythm and a sleep propensity index that is correlated with slow wave
activity. The upper circadian threshold marks the point at which sleep onset occurs,
while the lower limit triggers awakening. The circadian rhythm is modeled as a skewed
sine function. The NREM and REM stages during sleep are determined by the slow
wave activity with no activity in REM stage and an overall decaying throughout the night
for the NREM stage. The input for this compartment is the total ventilatory drive, Dvent.
The outputs are state drive, SI, awake/sleep state change, Saw/sleep, sleep stage, SSWA and
arousal, Darousal.

Sleep Mechanism Equations:

 76

Process C:

CH / L  A[0.97sin   0.22sin(2 )  0.07sin(3 )  0.03sin(4 )  0.001sin(5 )]  X

where A  0.12, X  X H  0.9 for CH , X  X L  0.15 for CL.

Process S:

Awake State (S  CL ): S (t )  1  [1  S (t  t )]e( 0.055t / 3600)

dS
Sleep State (S  CH ):   gc  SWA
dt
dSWA SWA
  rc  SWA(1  )   fc  SWA  REMT (t )  SWA  n(t )
dt S
REMT (t )  REM  0.2  AI

1 REM
REM  
0 NREM

1, DTotal  Thre

AI   , where Thre=I vent (0.7+0.3SSWA )
0, Otherwise

SWA
SSWA  1.2 
S

 SSWA , sleep onset transition


Dstate  SI   SSWAcombined , sleep
 0,
 awake

1, AI  0
SSWAcombined  
 0, AI  1

Reference: Khoo, M.C.K. A model-based evaluation of the single-breath CO2 ventilatory

response test. Journal of Applied Physiology, 68(1):393-399, 1990.

Achermann, P., Borbely, A.A. Mathematical models of sleep regulation. Frontiers in

Bioscience, 8, s683-693, 2003.
 77

Simulink Model: Sleep Mechanism

0.9

Circadian _High
Process Circadian H

Sine Wave Process Circadian H

Circadian High Sine Wave XOR

f(u) 1-u 0

Mux Fcn8 Sleep Enable

S between H and L ?

f(u)

S
Process Circadian L
Mux f(u)
0.15
1
Circadian _Low SI Sleep Wake State Index
Sleep Index
Saturation
-0.0005
Process S u1 if (u1 > 0)

If
u1
Step _Size 1/200
f(u) Mux
f(u) Mux
u2 2.1 2 1
u1
Fcn2 s

Mux xo
f(u)
In1 Out1
Fcn4
if { }
In1 Out1 t0
u2 fcn y Triggered
If Action REM
Subsystem
Subsystem t
Clock
1-u
t0
u2 fcn y Fcn1
Diet 5
t
Diet
Process Circadian H
4 u2
Process Circadian L REM S So
REM
Process S Sleep Awake
3 SWA_scaled Sleep/Awake 2
Circadian _Process Mux REM
SWA/S
SWA Airway Cond
Saturation 1 SWA AI - Arousal Index 1
AI - Arousal Index

In1 S/SWA Combined REM

Dstate SWA/Sleep State
Saving CircadianProcess
 78

Simulink Model: SWA/Sleep State

3
SWA

1
s f(u)
2
SWA_scaled

-gc
1 1.2 u>=1
2
s
Sleep /Awake xo Product 1 Gain
1
So
1
u<1
S

rc fc
1
Product 8

4
S/SWA Combined
1-u u<1
Product 10

1st NREM start

Product 11

1-u Fcn3

REM
4

Gain 3
1 0.2 3
AI - Arousal Index
Gain 6

Product 7 Gain 2

Band -Limited
White Noise

Inputs: Dvent Total Ventilatory Drive

Outputs: Darousal Arousal Sate

Dstate State Drive
Saw/sleep Awake/Sleep State change
SSWA Sleep Stage

Variables: A Amplitude of the skewed sine function

XH Bias of the skewed sine function for process CH
XL Bias of the skewed sine function for process CL
gc Constant for sleep decaying
rc Rising rate of slow wave activity
fc Falling rate of slow wave activity
SWAo Initial value of sleep wake activity
 79

Metabolic Control (PNEUMA.mdl)

Description

The metabolic control include the circadian regulation of epinephrine secretion,

epinephrine regulation on dynamic fluctuations in glucose and free-fatty acid in plasma,
metabolic coupling among tissues and organs provided by insulin and epinephrine, as
well as the effect of insulin on peripheral vascular sympathetic activity. The inputs for
this compartment are the alpha-sympathetic activity, ftas,res, sleep state index, SI, REM
sleep, REM, and diet uptake, DIET. The output is the change in the -sympathetic
response, Δftas.

Metabolic Control

Metabolic Control Equations:

Glucose Insulin and Free-fatty Acid Dynamics:

dG(t ) k u (t )  u2 ext (t )
  p1G(t )  p1Gb  p4 X (t )G(t )  p4 X bGb  p6Z (t )G (t )  p6GbZ b  EG 2int
dt VolG
 80

dI (t )
  (G (t  TDi )  Gh )t  n( I (t )  I b )  p5u1 (t )
dt
dX (t )
  p2 ( X (t )  X b )  p3 ( I (t )  I b )
dt
dY (t )
  pF 2 (Y (t )  Yb )  pF 3 ( I (t )  I b )
dt
dF (t ) k u (t )  u3ex t (t )
  p7 F (t )  p7 Fb  p8Y (t ) F (t )  p8Yb Fb  p9GG (t ) F (t )  p9GGb Fb  EF 3int
dt VolF
dZ (t )
 k2 ( Z (t )  Z b )  k1 ( F (t )  Fb )
dt
where p9G  0.00021e0.0055G

Epinephrine Regulation:
 (  E (t )  E (0))2 
Vx,i  Vxo,i 1.0  xE,i E 
  x,i  (  E (t )  E (0))2 

where subscript x = “heart”, “muscle”, “gastrointestinal tract”, “adipose tissue” or “other
tissues”; subscript i = “glucose” (assuming the metabolic pathway: GLC 
G6PGLY) or “FFA” (assuming the metabolic pathway: TGL  FFA  ACoA).
u2int (t )  Vx ,i (t )
x

u3int (t )  Vx ,i (t )
x

E (t )  E (0)  Eb  ( ftas ,meta ) [1.0  exp(t /  E )]

Autonomic and Metabolic Interactions (Forward Pathway):

( ftas,meta )  [ ftas,meta  ftas,meta 0 1]SI G as _ sleep

 (1  bREM  REM )  (1  SI  a )
ftas ,meta  ftas  (1  SI  Gas _ sleep )
ftas ,meta 0  ftas ,0 (1  SI  Gas _ sleep )
E (0)  Eb  KCe,0  ( ftas ,meta  ftas ,meta 0 )  (1  SI )

Autonomic and Metabolic Interactions (Feedback Pathway):

exp[( I  Ib ) / kisc, I ]  1
W ( I )  kas  kas  ftas , I 0 
exp[( I  Ib ) / kisc, I ]  1

ftas  W ( I )  [1  exp(t /  I )]
 81

ftas , FB  ftas  ftas

where ftas  ftas ,res and ftas ,vein , respectively.

Reference:
1. Kim, J., Saidel, G. M., and Cabrera, M. E. Multi-scale computational model of
fuel homeostasis during exercise: effect of hormonal control. An Biomed Eng
35(1): 69-90, 2006.

2. Roy, A., and Parker, R. S. Dynamic modeling of free fatty acid, glucose, and
insulin: an extended “Minimal Model”. Diabetes Tech Therapeu 8(6): 617-626,
2006.

Simulink Model: Metabolic Control

Insulin Input
Time Unit : Second

Epi _Heart Glucose u1(t) Plasma Glucose Concn , G(t)

1 Ftas_r
Epi _Muscle Glucose Glu _Epi -K -
Ftas_r
Epi _Heart FFA u2(t)
Glucose Input Plasma Insulin Concn , I(t)

2 SI Epi _Muscle FFA FFA_Epi -K-

SI Epi _GI FFA u3(t) Plasma FFA Concn , F(t) Display
Gain
Epinephrine Amount Ce (t)
3 REM
REM W_alphaSYMP W_alphaSYMP DeltaFtas 1
Epinephrine Regulation DeltaFtas
4 -K-
on Heart and Muscle Glucose-Insulin-FFA Minimal Model
DIET

Gclamp _in (t)

Glu_in(t)

G(t) Gclamp _in(t)

InsPump _in(t)

Glucose Insulin Interventions

Simulink Model: Glucose-Insulin-FFA Dynamics

2 u2(t)
u2(t)
1
Z(t) G(t) 1 4
s
Plasma DeltaFtas
Integrator 1
X (t) Glucose
Concn , 1/tao _I
Glucose Dynamics_LC G(t)
Gain 1
Sum 2
f(u)

G(t) 2*Fcn+2
Remote
I(t) 2
Insulin 1 u 1(t) Plasma Insulin
In1
X (t) 4
u1(t) Concn , I(t) W_alphaSYMP
Insulin Dynamics _ B&P
X(t)

Remote Saving deltaFtas_WalphaSYMP

Compartment
p3 / (s + p2)1
I(t)

Remote FFA
Z(t)
Z(t)

I(t)

Remote
Ins_FFA
pF3 / (s + pF2)
Y(t)
F(t)

G(t)

F(t) Y (t)

u3(t) 3
FFA Dynamics_LC u3(t)

3
Plasma FFA
Concn , F(t)
 82

Simulink Model: Glucose Dynamics

p1 Gb

1 -K-
u2(t) 1/VolG 1
p6*Gb *Zb
1
s G(t)
G(t)
p4*Xb *Gb

p6

2
Z(t)

p4

3
X(t)

Simulink Model: Insulin Dynamics

1
G(t)
Gamma

Thresholding Gamma Sum 2

Operator 1 Gb

Threshold Glucose
Ti
Concentration (h)
Variable 1
Random Time Delay s
Product Thresholding
Number 1 I(t)
Operator 2 Sum 3 Int 1
1
Int 2
s
n Ib
1 Insulin Destruction Sum 1
Constant Gain , Nu = 0.36

2 p5
u1(t)
 83

Simulink Model: Free-Fatty Acid Dynamics

1 G(t)
f(u )
p9

2
Y(t)
p8
Gb*Fb
Orinally Design
Yb=0.01 *Xb
p7

1
p7*Fb 1
s F(t)
F(t)
3 -K-
u3(t)
1/VolF

p8*0.1*Xb *Fb

Simulink Model: Epinephrine Regulations

1 Ftas _r 1microMol = 1e6 pMol

Ce(t)
Ftas_r
2 SI 1e6 6
SI W_alphaSYMP Epinephrine
3 REM 7 Amount Ce (t)
REM W_alphaSYMP
Epi Dynamics

Epi on Heart (micronmol /min )

1
Epi on Muscle
Epi _Heart
Epi on GI
Ce(t) Glucose
Epi on adipose
4Fluxes of Epi in Heart FLux Scope
5Fluxes of Epi in Muscle (micronmol /min )
Epi Modulation
3
Epi _Heart
FFA
2
Epi _Muscle
Glucose
4
Epi _Muscle
FFA
EPI Modulation
5
(micronmol /min )
Epi _GI
FFA
 84

Simulink Model: Epinephrine Dynamics

1
1
s
Ce (t)
Integrator
Gain

1/tao _e

Sum 1
1
Ftas_r f(u) 9

2 Terminator f(u )
Ftas_REM _Sleeo ABS Fcn
SI REM factor = 0.4 Ftas_Ce0*(1-SI*Gas_SleepM )
Ftas,r0_metabolic
Gas factor _M f(u)
Ftas_Ce0*(1-SI )

f(u)
Ftas_Ce 0
2
f(u)
W_alphaSYMP
Ftas_metabolic

f(u)
3
Ftas_REM _Sleep Fcn Saturation
REM
REM factor = 0.4 Ce_0
f(u) Ftas,r0_metabolic
Gas factor _M
Ftas_r0_metabolic
Neg Power

Simulink Model: Epinephrine Regulations on Heart

1
Ce(t) f(u)
-C-
Fcn
Heart : GLC _to_G6P
1
Epi on Heart (Fluxes )

-C-

Heart : GLY _to_G6P

2
Epi on Heart (Sum )

f(u) Add
-C-
Fcn2
Heart : FFA_to_ACoA

f(u)
-C-
Fcn3
Heart : TGL _to_FFA
 85

Simulink Model: Epinephrine Regulations on Muscle

1
Ce(t) f(u)
-C-
Fcn
Muscle : GLC _to_G6P
1
Epi on Muscle (Fluxes )

f(u)
-C-
Fcn1
Muscle : GLY _to_G6P
2
Epi on Muscle (Sum )

f(u) Add
-C-
Fcn2
Muscle : FFA_to_ACoA

f(u)
-C-
Fcn4
Muscle : PYR_to_ALA

f(u)
-C-
Fcn3
Muscle : TGL _to_FFA

Inputs: ftas,res Alpha-sympathetic Response

SI Sleep State Index
REM REM Sleep Signal
DIET Diet Glucose Uptake

Outputs: Δftas l Change in Alpha-sympathetic Response

Variables: G Plasma Glucose Concentration

I Plasma Insulin Concentration
X Remote Plasma Insulin Concentration
Y Remote Plasma Insulin Concentration that
Promotes FFA Production
F Plasma FFA Concentration
Z Remote Plasma FFA Concentration
E Epinephrine Concentration In Plasma
 86

Autonomic and Metabolic Interactions

Description

PNEUMA is extended from previous Version 2.0, an existing integrative model of

respiratory, cardiovascular and sleep-wake state control, to incorporate a sub-model of
glucose-insulin-fatty acid regulation. This computational model is capable of simulating
the complex dynamics of cardiorespiratory control, chemoreflex and state-related control
of breath-to-breath ventilation, state-related and chemoreflex control of upper airway
potency, respiratory and circulatory mechanics, as well as the metabolic control of
glucose insulin dynamics and its interactions with the autonomic control.

The interactions between autonomic and metabolic control include the circadian
regulation of epinephrine secretion, epinephrine regulation on dynamic fluctuations in
glucose and free-fatty acid in plasma, metabolic coupling among tissues and organs
provided by insulin and epinephrine, as well as the effect of insulin on peripheral
vascular sympathetic activity.

These model simulations provide insight into the relative importance of the various
mechanisms that determine the acute and chronic physiological effects of sleep-
disordered breathing. The model can also be used to investigate the effects of a variety of
interventions, such as different glucose clamps, the intravenous glucose tolerance test and
the application of continuous positive airway pressure on obstructive sleep apnea
subjects. incorporates several key cardiorespiratory reflexes and interactions. The
schematic diagram below shows the overall scheme in which these interactions have been
incorporated.
 87

Scheme 1. Interactions of Autonomic Control and Metabolic Control

 88

Appendix I: Software Package

Here are all the files for either the whole Pneuma or its individual modules. Please check
to make sure that you have downloaded all those files you need.

Overall PNEUMA Package:

PneumaRelease3.zip
FILES
PNEUMA.mdl
pneuma_acc.dll
PNEUMA_MAIN_CONTROL_PANEL.fig
PNEUMA_MAIN_CONTROL_PANEL.m
pneuma_variables.m
pneuma_gains.m
constant_parameters_6.fig
constant_parameters_6.m
adjustable_inputs_6.fig
adjustable_inputs_6.m
About.fig
About.m
directory_list.fig
directory_list.m
directory_list_load.m
directory_list_load.fig
directory_list_save.fig
directory_list_save.m
acquire_data.m
acquire_data_save.m
interventions.fig
interventions.m
cond_check.m
release_note.pdf
modaldlg.fig
modaldlg.m
CNS.bmp
cpap2.bmp
CV_pic.bmp
IC_pic.bmp
Metabolic2.bmp
Respiratory_System.bmp
PNEUMARelease3_MANUAL.pdf

Individual Modules:

Cardiovascular System:
Cardiovascular Cardiovascular.mdl
Cardiovascular_IC.m
 89

Autonomic Control:
Autonomic Autonomic.mdl
Autonomic_IC.m

SA Node:
SA_Ursino SA_Node_Ursino.mdl
SA_Node_Ursino_IC.m

Total Peripheral Resistance change:

TPR_Ursino TPR_Ursino.mdl
TPR_Ursino_IC.m

Respiratory System:
Respiratory Respiratory.mdl
Respiratory_IC.m

NeuroMuscular Profile:
NeuroMuscular NeuroMuscular.mdl
NeuroMuscular_IC.m

Respiratory Mechanics (whole):

Resp_Mech Resp_Mech.mdl
Resp_Mech_IC.m

Respiratory Mechanics (Pmus):

Pmus_Flow_Younes Pmus_Flow_Younes.mdl
Pmus_Flow_Younes_IC.m

Respiratory Mechanics (Pleural Pressure):

Pleural_Schuessler Pleural_Schuessler.mdl
Pleural_Schuessler_IC.m

State/Upper Airway Interaction:

State_UA_Khoo State_UA_Khoo.mdl
State_UA_Khoo_IC.m

Gas Exchange (Overall model):

Gas_Exchange Gas_Exchange.mdl
Gas_Exchange_IC.m

Gas Exchange (Individual):

Dead_Space_Khoo Dead_Space_Khoo.mdl
Dead_Space_Khoo_IC.m

Lungs_Khoo Lungs_Khoo.mdl
Lungs_Khoo_IC.m

Cardio_Mix_Lange Cardio_Mix_Lange.mdl
Cardio_Mix_Lange_IC.m

Dissociation_Spencer Dissociation_Spencer.mdl
Dissociation_Spencer_IC.m
 90

Brain_Khoo Brain_Khoo.mdl
Brain_Khoo_IC.m

Body_Khoo Body_Khoo.mdl
Body_Khoo_IC.m

Vent_Drive_Khoo Vent_Drive_Khoo.mdl
Vent_Drive_Khoo_IC.m

Reflex_Ursino Reflex_Ursino.mdl
Reflex_Ursino.IC

State_UA_Khoo_Borbely State_UA_Khoo_Borbely.mdl
State_UA_Khoo_Borbely_IC
 91

Appendix II: Saved Data Files

Here is the list of saved data files and the corresponding contents inside the files.
Because of the limitations in Matlab® that no data file bigger than 1 GB can be loaded,
for the purpose of saving longer time simulation results, the saved data files for each
group of data are segmented into 10 small data files naming from ***1.mat to ***10.mat
where *** is the data file’s name. Each file must be smaller than 1 GB which is large
enough for a 10-week simulation (3600*24*70 second run time) with sample period 0.1
sec. However, if the simulation time is longer than 12-weeks (3600*24*84 second run
time), then the sampling interval (step duration) must be longer than 0.1 second to ensure
each data file is smaller than its limitation 1 GB.

Data File Name Contents

Autonomic#.mat Autonomic Control Output Data:

ftas_r, ftas_v, ftbs and ftp
BreathingPeriod#.mat Variable Breathing Period Input/Output Data
CARDIO#.mat Cardiovascular System Outputs: Heart Period HP,
Stroke Volume SV, Cardiac Output CO, TPR and
ABP
CARDIORESPIRATORY#.mat Overall Main Outputs of Cardiorepiratory
Interactions: State Drive SI, HR, ABP, Ppl, PaCO2,
SaO2, Breathing Frequency BF, Tidal Volume Vt,
Total Ventilatory Drive DTotal
CircadianProcess#.mat Circadian Process Data in Sleep Mechanism
deltaFtas#.mat Insulin Effects on Peripheral Sympathetic Activity
GIMM_FFA_SEC#.mat Metabolic Model’s Inputs and Outputs: G(t), I(t),
F(t), E(t), Gin(t), I(t)_in
Nt#.mat Central Respiratory Neural Drive Nt
PVleft#.mat Pressure and Volume of Left Atria
Resp_Rhythm#.mat Respiratory Rhythm Resp_Rhythm
stpres#.mat Dynamic Drives for Ventilatory Drive: X, Y and Z
TPR#.mat All the resistances and unstress volumes controlled
by alpha-sympathetic activities
varHeartPeriod#.mat Variable Heart Period Input/Output Data

where # represents number 1 to 10 for each data file.

 92

Appendix III: Saved/Load Data for Advanced Users

For the advanced user, a potentially useful option that is available when PNEUMA V.3.0
is run using Matlab® versions higher than R2009a is the ability to load initial states pre-
simulation and to save final states post-simulation. This allows for a simulation to be
continued starting at the time when the previous simulation run was terminated (assuming
the final states have been saved prior to termination). For example, before clicking on
“RUN” in the Control Panel for 1000 sec simulation, the advanced user can set up the
final state as “yinitial1000” as shown below by opening Configuration in PNEUMA.mdl.

Then, click “Run” to save all the “data” in the workspace when the simulation is stopped
at 1000 sec by using “Save Data” in “File” menu. To load these data for the purpose of
resuming the simulation run from t=1000 sec, modify the configuration as below by
checking the box “Initial State” and change its name into “yinitial1000”; to save the final
state, modify the configuration as below by checking the box “Final States” and change
its name into “yFianl2000” shown as below:
 93

To continuously save/load the states, be sure to provide different names for the initial and
final states. Please note that this feature is only available when using PNEUMA V.3.0 in
Matlab® versions higher than R2009a.
 94

Appendix IV: Overall Parameter Set and Initial

Conditions
Here are the parameters and initial conditions for the complete PNEUMA model. During
simulations and before simulations, some of these parameters and conditions can be
modified. They are also the parameters/variables in the work space that will be saved into
a data file when you select “Save data” under “File”. When you choose “Load data”,
those parameters will be loaded into the workspace and some of these
parameters/variables will be used as initial conditions in the subsequent simulation. All
the saved simulation outputs can be extracted in Matlab® by the user for further plotting
or analysis.

Parameter Definition Values Units

Cardiovascular System
Resistances
RPA Pulmonary arterial flow resistance 0.023 mmHg*s/mL
RPP Pulmonary peripheral flow resistance 0.0894 mmHg*s/mL
RPV Pulmonary venous flow resistance 0.0056 mmHg*s/mL
RSA Systemic arterial flow resistance 0.06 mmHg*s/mL
RSP Splanchnic peripheral flow resistance 3.307 mmHg*s/mL
REP Extra-splanchnic peripheral resistance 3.52 mmHg*s/mL
RMPN Skeletal muscle peripheral flow resistance 4.48 mmHg*s/mL
RBPN Cerebral peripheral flow resistance 6.57 mmHg*s/mL
RHPN Coronary peripheral flow resistance 19.71 mmHg*s/mL
RSV Splanchnic venous flow resistance 0.038 mmHg*s/mL
REV Extra-splanchnic venous resistance 0.04 mmHg*s/mL
RMV Skeletal muscle venous flow resistance 0.05 mmHg*s/mL
RBV Cerebral venous flow resistance 0.075 mmHg*s/mL
RHV Coronary venous flow resistance 0.224 mmHg*s/mL
RVC_0 Nominal vena cava flow resistance 0.025 mmHg*s/mL
RLA Left atrial flow resistance 0.0025 mmHg*s/mL
RRA Right atrial flow resistance 0.0025 mmHg*s/mL
Compliances
CPA Pulmonary arterial compliances 0.76 mL/mmHg
CPP Pulmonary peripheral compliances 5.8 mL/mmHg
CPV Pulmonary venous compliances 25.37 mL/mmHg
CSA Systemic arterial compliances 0.28 mL/mmHg
CSP Splanchnic peripheral compliances 2.05 mL/mmHg
CEP Extra-splanchnic peripheral compliances 0.668 mL/mmHg
CMP Skeletal muscle peripheral compliances 0.525 mL/mmHg
 95

CBP Cerebral peripheral compliances 0.358 mL/mmHg

CHP Coronary peripheral compliances 0.119 mL/mmHg
CSV Systemic venous compliances 61.11 mL/mmHg
CEV Extra-splanchnic venous compliances 20 mL/mmHg
CMV Skeletal muscle venous compliances 15.71 mL/mmHg
CBV Cerebral venous compliances 10.71 mL/mmHg
CHV Coronary venous compliances 3.57 mL/mmHg
CLA Left atrial compliances 19.23 mL/mmHg
CRA Right atrial compliances 31.25 mL/mmHg
Inertances
LPA Pulmonary arterial inertance 0.00018 mmHg*s2/mL
LSA Systemic arterial inertance 0.00022 mmHg*s2/mL
Unstressed Volume
VUPA Pulmonary arterial unstressed volume 0 mL
VUPP Pulmonary peripheral unstressed volume 123 mL
VUPV Pulmonary venous unstressed volume 120 mL
VUSA Systemic arterial unstressed volume 0 mL
VUSP Splanchnic peripheral unstressed volume 274.4 mL
VUEP Extra-splanchnic peripheral unstressed volume 134.64 mL
VUMP Skeletal muscle peripheral unstressed volume 105.8 mL
VUBP Cerebral peripheral unstressed volume 72.13 mL
VUHP Coronary peripheral unstressed volume 24 mL
VUSV Splanchnic venous unstressed volume 1121 mL
VUEV Extra-splanchnic venous unstressed volume 550 mL
VUMV Skeletal muscle venous unstressed volume 432.14 mL
VUBV Cerebral venous unstressed volume 294.64 mL
VUHV Coronary venous unstressed volume 98.21 mL
VVC_0 Vena cava unstressed volume 130 mL
VULA Left atrial unstressed volume 25 mL
VURA Right atrial unstressed volume 25 mL
VULV Left ventricular unstressed volume 16.77 mL
VURV Right ventricular unstressed volume 40.88 mL
Vena Cava
Kr_vc Gain for vena cava flow resistance 0.001 mmHg*s/mL
Vvc_max Maximum volume of vena cava 350 mL
Vvc_min Minimum volume of vena cava 50 mL
D1 Parameter for P-V curve of vena cava 0.3855 mmHg
D2 Parameter for P-V curve of vena cava -5 mmHg
K1_vc Parameter for P-V curve of vena cava 0.15 mmHg
 96

K2_vc Parameter for P-V curve of vena cava 0.4 mmHg

Respiratory System
Pleural Pressure and Alveolar Pressure
Rcw Chest wall resistance 1.03 cmH2O*s/L
RLT Lung transmural resistance 1.69 cmH2O *s/L
Raw Airway wall resistance 1.016 cmH2O *s/L
Ecw Chest wall elastance 5 cmH2O /L
ELT Lung transmural elastance 5 cmH2O /L
k1,aw Constant for upper airway pressure 1.85 cmH2O *s2/ L2
k2,aw Constant for upper airway pressure 0.43 cmH2O *s2/ L2
Gas Exchange and Transport
Dead Space
th
Dead(i),co2IC Initial condition for i CO2 dead space 39.562 L
th
Dead(i),co2IC Initial condition for i CO2 dead space 39.674 L
th
Dead(i),co2IC Initial condition for i CO2 dead space 39.813 L
th
Dead(i),co2IC Initial condition for i CO2 dead space 40.006 L
th
Dead(i),o2IC Initial condition for i O2 dead space 104.36 L
th
Dead(i),o2IC Initial condition for i O2 dead space 104.23 L
th
Dead(i),o2IC Initial condition for i O2 dead space 104.05 L
Dead(i),o2IC Initial condition for ith O2 dead space 103.8 L
th
Vd(i) i dead space volume (i={1,..4} 0.03 L
PI,CO2 Inspiratory CO2 partial pressure 0 Torr
PI,O2 Inspiratory O2 partial pressure 150 Torr
Vt' Respiratory flow variable L/sec
Vt Tidal Volume variable L
PdO2 Dead space O2 partial pressure variable Torr
PdCO2 Dead space CO2 partial pressure variable Torr
Alveolar Gas Exchange
Vco2, VLco2 Lungs storage volume for CO2 3 L
Vo2, VLo2 Lungs storage volume for O2 2.5 L
PAco2IC Initial condition for Partial CO2 pressure 40.943 Torr
PAo2IC Initial condition for Partial O2 pressure 102.52 Torr
PAo2IC Initial condition for Partial O2 pressure 102.52 Torr
PACO2 Alveolar CO2 partial pressure variable Torr
PACO2 Alveolar O2 partial pressure variable Torr
Palv Alveolar partial gas pressure variable Torr
Q Blood flow variable L/sec
Cardiovascular Transport
tauchemo Peripheral chemoreceptors delay time constant 2 s
 97

T1 Time constant for cardiovascular mixing 1 s

T2 Time constant for cardiovascular mixing 2 s
Ta Lung to chemoreceptor circulation delay variable s
LCTV0 Lung to chemoreceptor transportation volume 0.588 liter
constant
PaO2firstIC Initial condition for first order Pao2 system 0.3557 Torr
PaO2secondIC Initial condition for second order P ao2 system 103.14 Torr
PaCO2firstIC Initial condition for first order P aco2 system -0.2465 Torr
PaCO2secondIC Initial condition for second order P aco2 system 40.393 Torr
PaO2_delayIC Initial condition for O2 convection 103.12 Torr
Paco2_delayIC Initial condition for CO2 convection 40.445 Torr
PaCO2 CO2 partial pressure variable Torr
PaO2 O2 partial pressure variable Torr
Cardiovascular Dissociation
C1 Maximum concentration of hemoglobin-bound 9 mL/mL
oxygen
C2 Maximum carbon dioxide concentration 87 mL/mL
a1 Parameter in O2 dissociation equation 0.3836 dimensionless
a2 Parameter in CO2 dissociation equation 1.819 dimensionless
alpha1 Parameter in O2 dissociation equation 0.02598 dimensionless
alpha2 Parameter in CO2 dissociation equation 0.05591 dimensionless
K1 Parameter in O2 dissociation equation 13 dimensionless
K2 Parameter in CO2 dissociation equation 194.4 dimensionless
beta1 Parameter in O2 dissociation equation 0.012275 dimensionless
beta2 Parameter in CO2 dissociation equation 0.03255 dimensionless
Sao2_delayIC Initial Condition for Oxygen Saturation Delay 98.92 sec
Brain Compartment
MRbco2 Metabolic production rate for CO2 in the brain 0.0517 1/s STPD
tissue
Sco2 Dissociation slope for CO2 in the blood 0.0043 mL/(mL*Torr)
Sbco2 Dissociation slope for CO2 in the brain tissue 0.36 mL*100g-
1
/Torr
Pbco2IC Initial condition for partial CO2 pressure from the 48.538 Torr
brain
Body Tissues Compartment
Vtco2 Body tissue storage volume for CO2 6 L
Vto2 Body tissue storage volume for O2 7.7 L
MRco2 Metabolic production rate for CO2 0.0033 1/s STPD
MRo2 Metabolic consumption rate for O2 0.0038 1/s STPD
Cvco2IC Initial condition for mixed venous CO2 0.5247 mL/mL
concentration
Cvo2IC Initial condition for mixed venous O2 0.1639 mL/mL
concentration
Upper Airway Model
 98

Ruaw Upper airway wall resistance 1000000 cmH2O*s/L

A0ua Maximum area of opening in upper airway 1 a.u.
Kua Proportionality coefficient between Aua and Yua; 1 L/(s*cmH2O)
Pcrit_awake Critical upper airway pressure in wakefulness -40 cmH2O
Sua Upper airway sensitivity to collapse 0.01 a.u.
Cua Upper airway compliance variable L/cmH2O
Pua Upper airway pressure variable cmH2O

Upper airway flow variable cmH2O
V ua

Total flow in airways variable cmH2O
V
Respiratory Muscle Activity
FlowIC Initial air flow 0 L/s
VC Vital Capacity 5 L
Pt_frcIC1 Initial condition for respiratory muscle reaction 0 spikes/s
Pt_frcIC2 Initial condition for respiratory muscle reaction 0 spikes/s
FlowIC Initial condition for airflow 0 L/s
VtIC Initial condition for lung volume 0 L
Central Neural Control

Carotid Baroreceptors

Pn Center pressure for sigmoidal function 92 mmHg

Kcs Parameter for sigmoidal slope control 11.758 mmHG
Pn_sleep Parameter for sleep effects 0 mmHg
Kcs_sleep Parameter for sleep effect 0 mmHG
fcs,min Lower threshold for sigmoidal function 2.52 spikes/s
fcs,max Upper saturation for sigmoidal function 47.78 spikes/s
τZ Time constant for baroreflex 6.37 s
τP Time constant for baroreflex 2.076 s
Ventilatory Response
Ic Central apneic threshold 45 dimensionless
IpCO2 Peripheral apneic threshold for CO2 38 dimensionless
IpO2 Peripheral apneic threshold for O2 102.4 dimensionless
Gc Gain for central chemical drive 0.075 dimensionless
Gp Gain for peripheral chemical drive 0.0063 dimensionless
Swake Factor of wakefulness to sleep 0.3 dimensionless
Chemoreflex Control of Variable Respiratory Rhythm
Fb Basal breathing frequency 12.5 Breath
/min
Vb Basal ventilation 6.7 L/min
TD Chemoreflex drive threshold 1539 mL
TP Chemoreflex drive threshold 2879 mL
 99

S1F Scaling factor 0.00518 dimensionless

S1V Scaling factor 0.024 dimensionless
S2F Scaling factor 0.0105 dimensionless
S2V Scaling factor 0.0367 dimensionless
Chemoreflex

fchemo,max Upper saturation for the sigmoidal function 12.3 spikes/s

fchemo,min Lower saturation for the sigmoidal function 0.835 spikes/s
fchemo_control Basal level for the chemoreflex 1.4 dimensionless
Kchemo Slope control parameter for the sigmoidal function 29.27 mmHg
KH Constant value for the static response 3 dimensionless
τchemo Time constant for the chemoreflex 2 s
Lung Stretch Receptors Reflex
Gls Constant gain 23.29 spikes/sec/liter
τls Time constant 2 sec
Offsets
Xsa Saturation for the offset of α-sympathetic activity 6 Torr
on peripheral resistance
θsan Nominal level of offset of α-sympathetic activity 13.2 spikes/sec
on peripheral resistance
PO2nsa Central point for the sigmoidal function 30 Torr
kiscsa Parameter of α-sympathetic activity on peripheral 2 dimensionless
resistance
Xsb Saturation for the offset of -sympathetic activity 21.2 Torr
θsbn Nominal level of offset of -sympathetic activity 3.6 spikes/sec
PO2nsb Central point for the sigmoidal function 45 Torr
kiscsb Parameter of -sympathetic activity 4 dimensionless
Xsp Saturation for the offset of α-sympathetic activity 6 dimensionless
on peripheral resistance
θspn Nominal level of offset of α-sympathetic activity 13.2 spikes/sec
on peripheral resistance
PO2nsp Central point for the sigmoidal function 30 Torr
kiscsp Parameter of α-sympathetic activity on unstressed 2 dimensionless
volume of veins
τisc Time constant for oxygen response 30 s
τcc Time constant for carbon dioxide response 20 s
Autonomic Control
fcs,0 Center point for the sigmoidal function for 25 spikes/s
parasympathetic
fpara,0 Lower saturation of the parasympathetic 3.2 spikes/s
exponential decay function
fpara, Upper limit of the parasympathetic exponential 6.3 spikes/s
decay function
kp Slope control parameter for the sigmoidal function 7.06 dimensionless
G_RSA,p Central RSA gain for parasympathetic response 0.4 dimensionless
Gchemo,p Chemoreflex gain for parasympathetic response 0.03 dimensionless
 100

Glung,p Lung stretch receptor reflex gain for 0.24 dimensionless

parasympathetic response
fs,0 Upper limit of the sympathetic exponential decay 16.11 spikes/s
function
fs, Lower saturation of the sympathetic exponential 2.1 spikes/s
decay function
Ks Constant for the exponential function 0.07 s
G_RSA,bs Central RSA gain for -sympathetic response 0.4 dimensionless
Gchemo,bs Chemoreflex gain for -sympathetic response 2.8 dimensionless
Glung,bs Lung stretch receptor reflex gain for -sympathetic 0.24 dimensionless
G_RSA,as Central RSA gain for -sympathetic response 0.4 dimensionless
Gchemo,as Chemoreflex gain for -sympathetic response 4 dimensionless
Glung,as Lung stretch receptor reflex gain for -sympathetic 0.34 dimensionless
-Sympathetic Response
ftbsIC -sympathetic initial output after time delay 3.8576 spikes/s
ftbs_min Lower limit for the natural log function 2.66 spikes/s
Gbs -sympathetic Gain varied with sleep drive -0.13 dimensionless
Gbs_sleep -sympathetic sleep gain factor 0.2 dimensionless
τbs -sympathetic time constant 2 s
Dbs Delay for -sympathetic time constant 2 s
Parasympathetic Response
ftpIC Para sympathetic initial output after time delay 4.2748 spikes/s
Gpara Parasympathetic Gain varied with sleep drive 0.09 dimensionless
Gpara_sleep Parasympathetic sleep gain factor 0.2 dimensionless
τpara Parasympathetic time constant 1.5 s
Dbs Delay for parasympathetic time constant 0.2 s
Neuromuscular Drive
Inhale Boolean variable for inhalation 1 dimensionless
Sino-Atrial Node
HPbasal Basal value for HP for denervated heart 0.58 s
Maximum End-systolic Elastance
Glv Elastance gain for left ventricle 0.475 mmHg
/ml/v
Dlv Delay for elastance of left ventricle 2 s
τlv Time constant for elastance of left ventricle 8 s
Emax0_lv Basal level of maximum end-systolic elastance of 2.392 mmHg
left ventricle /ml
Grv Elastance gain for right ventricle 0.282 mmHg
/ml/v
Drv Delay for elastance of right ventricle 2 s
τrv Time constant for elastance of right ventricle 8 s
Emax0_rv Basal level of maximum end-systolic elastance of 1.412 mmHg
right ventricle /ml
-Sympathetic Control of Peripheral Resistance
 101

fasIC -sympathetic initial output after time delay 34.793 spikes/s

fas_min Lower limit for the natural log function 2.66 spikes/s
Gas_sleep -sympathetic Gain varied with sleep 0.3 dimensionless
Gas_sp -sympathetic Gain for splanchnic peripheral 0.695 dimensionless
resistance
τas_sp -sympathetic time constant 2 s
Das_sp Delay -sympathetic time constant 2 s
Gas_ep -sympathetic Gain for extra-splanchnic peripheral 1.94 dimensionless
resistance
τas_ep -sympathetic time constant 2 s
Das_ep Delay -sympathetic time constant 2 s
Gas_mp -sympathetic Gain for skeletal muscle peripheral 2.47 dimensionless
resistance
τas_mp -sympathetic time constant 2 s
Das_mp Delay -sympathetic time constant 2 s
Vusv0 Basal level of unstressed volume of splanchnic 1435.4 ml
venous circulation
Gas_usv -sympathetic Gain for unstressed volume of -265.4 ml/v
splanchnic venous circulation
τas_usv -sympathetic time constant 20 s
Das_usv Delay -sympathetic time constant 5 s
Local Blood Flow Control of Peripheral Resistance
PaCO2_n Nominal arterial CO2 partial pressure i 40 Torr
CvO2n_b Nominal venous O2 concentration in cerebral 0.14 dimensionless
peripheral circulation
CvO2n_m Nominal venous O2 concentration in skeletal 0.155 dimensionless
muscle peripheral circulation
CvO2n_h Nominal venous O2 concentration in coronary 0.11 dimensionless
peripheral circulation
Tau_CO2 Time constant for peripheral CO2 response 20 s
Tau_O2 Time constant for peripheral O2 response 10 s
A Parameter for flow regulation equation 20.9 dimensionless
B Parameter for flow regulation equation 92.8 dimensionless
C Parameter for flow regulation equation 10570 dimensionless
GO2_b Gain of local O2 response on cerebral vascular 10 dimensionless
bed
GO2_h Gain of local O2 response on coronary vascular 35 dimensionless
bed
GO2_m Gain of local O2 response on muscular vascular 30 dimensionless
bed
Sleep Mechanism
A Amplitude of the skewed sine function 20.9 dimensionless
XH Bias of the skewed sine function for process CH 0.9 dimensionless
XL Bias of the skewed sine function for process CL 0.15 dimensionless
gc Constant for sleep decaying 0.2/60 dimensionless
rc Rising rate of slow wave activity 0.4/60 dimensionless
fc Falling rate of slow wave activity 0.008/60 dimensionless
 102

SWAo Initial value of sleep wake activity 0.007 dimensionless

Interlink between Metabolic Model and Autonomic Control

KCe,0 Gain for basal level of epinephrine in plasma 9 dimension-less

bREM Gain for REM sleep effect from autonomic control 0.4 dimension-less
on epinephrine regulations
aw Parameter from autonomic control on epinephrine 0.6 dimension-less
regulations
ftas,0 basal firing rate of sympathetic activity 2.1 1/s
Kas Gain of metabolic feedback to change of 2 dimension-less
sympathetic activities
ftas.I0 Parameter of metabolic feedback to change of 1 dimension-less
sympathetic activities
Kisc,I Parameter of metabolic feedback to change of 20 dimension-less
sympathetic activities
τI Time constant of metabolic feedback to change of 30 min
sympathetic activities
Plasma Glucose Dynamics
P1 Utilization rate for plasma glucose concentration 0.068 1/min
P4 Utilization rate for plasma glucose concentration 1.3 mL/min
under the influence of remote insulin /µU
P6 Production rate for remote plasma glucose 0.00006 L/min
concentration that promotes FFA /µmol
Gb Basal level of plasma glucose concentration 124.8 mg/dL
VolG Glucose distribution space 117 dL
KEG Gain from epinephrine to glucose uptake 0.04 dimension-less
Plasma Insulin Dynamics
n Utilization rate for plasma insulin concentration 0.142 1/min
P5 Factor for insulin inputs 0.000568 1/mL
Ib Basal level of plasma insulin concentration 16.6 µU/mL
P3 Production rate for remote insulin concentration 0.000012 1/min
 Insulin sensitivity factor 0.038 µU/mL/min2per
mg/dL
TDi Variable time delay 5±3 sec
Gh Threshold of plasma glucose concentration 125 mg/dL
P2 Utilization rate for remote insulin concentration 0.037 1/min
PF2 Utilization rate for remote insulin concentration 0.17 1/min
that promotes FFA
PF3 Production rate for remote insulin concentration 0.00001 1/min
that promotes FFA
Xb Basal level of remote plasma insulin concentration 0.08125 µU/mL
Yb Basal level of remote plasma insulin concentration 0.008125 µU/mL
that promotes FFA production
Plasma Free Fatty Acid Dynamics
P7 Utilization rate for plasma FFA concentration 0.03 1/min
P8 Utilization rate for remote plasma insulin involved 4.5 mL/
FFA concentration min/ µU
Fb Basal level of plasma FFA concentration 380 µmol/L
Zb Basal level of remote plasma FFA concentration 190 µmol/L
 103

k2 Utilization rate for remote FFA concentration 0.03 1/min

k1 Production rate for remote FFA concentration 0.02 1/min
VolF FFA distribution space 11.7 L
KEF Gain from epinephrine to FFA uptake 0.01 dimension-less
Epinephrine Regulation
Eb Basal level of epinephrine concentration in plasma 198 pM
τE Time constant for epinephrine regulation 30 min
Δ Epinephrine regulation factor for metabolic fluxes 1e6 dimension-less
V0_GLC_Heart Maximum rate coefficient in heart 88 µmol/min
λE_GLC_Heart Epinephrine regulated flux parameter in heart 3 dimension-less
αE_GLC_Heart Epinephrine regulated flux parameter in heart 1000 pM

V0_GLY_Heart Maximum rate coefficient in heart 320 µmol/min

λE_GLY_Heart Epinephrine regulated flux parameter in heart 0 dimension-less
αE_GLY_Heart Epinephrine regulated flux parameter in heart 0 pM
V0_FFA_Heart Maximum rate coefficient in heart 280 µmol/min
λE_FFA_Heart Epinephrine regulated flux parameter in heart 2 dimension-less
αE_FFA_Heart Epinephrine regulated flux parameter in heart 447.2 pM
V0_TGL_Heart Maximum rate coefficient in heart 8 µmol/min
λE_TGL_Heart Epinephrine regulated flux parameter in heart 0.5 dimension-less
αE_TGL_Heart Epinephrine regulated flux parameter in heart 1000 pM
V0_GLC_Muscle Maximum rate coefficient in muscle 398 µmol/min
λE_GLC_Muscle Epinephrine regulated flux parameter in muscle 18 dimension-less
αE_GLC_Muscle Epinephrine regulated flux parameter in muscle 1000 pM
V0_GLY_Muscle Maximum rate coefficient in muscle 1000 µmol/min
λE_GLY_Muscle Epinephrine regulated flux parameter in muscle 0.3 dimension-less
αE_GLY_Muscle Epinephrine regulated flux parameter in muscle 10 pM
V0_FFA_Muscle Maximum rate coefficient in muscle 701 µmol/min
λE_FFA_Muscle Epinephrine regulated flux parameter in muscle 9 dimension-less
αE_FFA_Muscle Epinephrine regulated flux parameter in muscle 447.2 pM
V0_PYR_Muscle Maximum rate coefficient in muscle 80 µmol/min
λE_PYR_Muscle Epinephrine regulated flux parameter in muscle 2 dimension-less
αE_PYR_Muscle Epinephrine regulated flux parameter in muscle 1000 pM
V0_TGL_Muscle Maximum rate coefficient in muscle 260 µmol/min
λE_TGL_Muscle Epinephrine regulated flux parameter in muscle 2.5 dimension-less
αE_TGL_Muscle Epinephrine regulated flux parameter in muscle 1000 pM
V0_TGL_GI Maximum rate coefficient in GI tract 80 µmol/min
λE_TGL_GI Epinephrine regulated flux parameter in GI tract 2 dimension-less
αE_TGL_GI Epinephrine regulated flux parameter in GI tract 1000 pM
V0_TGL_adipose Maximum rate coefficient in adipose 190 µmol/min
 104

λE_TGL_ adipose Epinephrine regulated flux parameter in adipose 2 dimension-less

αE_TGL_ adipose Epinephrine regulated flux parameter in adipose 1000 pM