See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/348857418

MATLAB/Simulink Medical Fluid Pump Model with a Flow PID Controller

Conference Paper · December 2020

DOI: 10.1109/ICM50269.2020.9331807

CITATION READS

1 490

12 authors, including:

Mohamed Wadaane Saeed H. Bamashmos

CHU El-maarouf Lebanese International University
6 PUBLICATIONS 11 CITATIONS 9 PUBLICATIONS 15 CITATIONS

SEE PROFILE SEE PROFILE

Mohamad Abou Ali Abdallah Kassem

Universidad del País Vasco / Euskal Herriko Unibertsitatea Notre Dame University
35 PUBLICATIONS 113 CITATIONS 113 PUBLICATIONS 1,025 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Mohamad Abou Ali on 12 April 2024.

The user has requested enhancement of the downloaded file.

 MATLAB/Simulink Medical CO2 Insufflator Model with PID-PQT &
MPC-PQT Controllers
Kinana Rashwani1, Hussein Wehby1, Mariam Khayreldeen 1, Abdulhalim Mohamad1, Mohamad I.C.
HajjHassan 1, Mohamed Wadaane1, Ahmad ElSayed1, Ahmed N. Al-naggar2, Saeed Bamashmos2, Hassan
Wehbi4, Mohamad HajjHassan1, Mohamad Abou Ali1, Abdallah Kassem3
1
Department of Biomedical Engineering, Lebanese International University, Lebanon
2
Department of Biomedical Engineering, Lebanese International University, Yemen
3
ECCE Department, Faculty of Engineering, Notre Dame University-Louaize, Lebanon
4
Research & Development, INGENIOUS Medical GmbH, Germany

Abstract—Medical CO2 insufflator is used in a created by the CO2 insufflator; this area is called
minimally invasive surgery creating abdominal cavity pneumoperitoneum [2].
to provide surgeons with a direct view of internal Hybrid pneumatic-electronic laparoscopic insufflators
organs. In our previous work, MATLAB/Simulink consist of three parts: Pneumatic part (High Pressure
medical CO2 insufflator has been developed with only Unit, and Low Pressure Unit), electronic part, and
abdominal cavity’s pressure via a PID controller. This applied part (heated hose system, trocar…) [5]. Such a
work permits the simultaneous control of three device requires inclusive control of all compartments,
parameters pressure, flow and temperature through a parameters and variables.
PID controller. Then, an MPC controller has been Since 2017, Food & Drug Administration (FDA)
implemented to enhance the precision, accuracy, and strongly demands in-silico (computer) modeling to
the response time of the system. The results are accelerate and facilitate the costly long-time process of
promising due to the efficiency of MPC controller in getting an FDA approval. Because of what is mentioned,
multi-input/multi-output system over PID controller. In modeling is needed for any device before its launch in
the future work, high flow CO2 insufflator will be the market, that is to avoid expense of prototype trials as
developed. well as to supply more computerized accuracy and
precision, and MATLAB/SIMULINK is a great tool to
Keywords— Medical CO2 insufflator; computational attain the target. Furthermore, this offers abundance of
modeling; MATLAB/SIMULINK-Simscape; PID and time and economy before the hardware implementations.
MPC controllers. In the previous paper [3], the medical CO2 insufflator-
patient system is modeled and controlled via PID
I. INTRODUCTION controller using MATLAB/SIMULINK-Simscape tool
Minimally Invasive Surgery (MIS) utilizes an regarding only the pressure, which was insufficient since
endoscope to reach internal organs through very small there are many other variables than pressure to be
incisions [1]. Laparoscopic MIS is a platform that considered, to keep the device-patient system under full
permits operating within abdominal-pelvic parts [2]. safety. Moreover, a PID controlled- “proportional
Medical carbon dioxide (CO2) insufflator is widespread pressure regulator PPR” is utilized instead of a
involved in such surgeries, dilating the peritoneal cavity, traditional pressure regulator (pneumatic pressure
aiming to provide better and easier view as well as regulator; the knob).
workspace for surgeons [2]. This work is developed as According to what is recommended by FDA and the
an enhancement of the previous paper outcome, where rapid growth of modern technology, a fully integrated
only pressure variable was considered and controlled by control over the whole system is to be generated
PID controller [3]. including more variables (pressure, flow, temperature).
Since the dawn of medical and operational surgeries, System modeling via same tool is then developed adding
traditional surgeries were based on large cuttings to flow and temperature variables under the control of
visualize internal parts of the patient, followed by a long- PID(s). Later on, PID controllers are eventually replaced
term recovery. Today, MIS is excelled by tiny tools, by MPC controller to rule, monitor and regulate all what
cameras, and light sources which grant small incisions, is said above, since MPC controller is way better than
less hemorrhage, and short-term recovery [4]. PID regarding precision outcomes, short-term response,
Laparoscopic surgery, also known as MIS, is applied multi-input / multi-output signals… etc.
at the abdominal-pelvic site for diagnosis or treatment Literature review and system design are
purposes by the aid of medical insufflator, affording elaborated in sections II and III respectively. Results are
more room inside patient’s abdomen. Those medical analyzed and discussed in section IV. Finally, section V
insufflators use two major types of gases: CO2 and N2O; offers a brief conclusion and future work.
in this work, the medical CO2 insufflator is on the spot.
At the level of peritoneal cavity, a gas-filled space is
 II. LITERATURE REVIEW to produce the most appropriate one attaining the highest
A fractional related work is published by World of safety and performance.
Medicine (WOM) Company, an endoscope-specialized
German company. “WOM Reduces Time-to-Market for
Surgical Device Control Software with Model-Based
Design”; yet, such work is not published [6]. A compact
description of this task is to be found exclusively in the
MathWorks Company website page. WOM Company
did promote medical CO2 insufflator, where WOM team
used measured input-output data to have a non-linear
abdominal cavity mathematical model developed using Figure 1: Medical CO2 Insufflator SIMULINK-
system identification tool in MATLAB. Besides, two Simscape Model.
cascaded PI controllers were involved; one for pressure
The system includes five fundamental parts: High
control and another for flow control “PI-PF”.
Pressure Unit HPU, Low Pressure Unit LPU,
This piece of work is proposing medical CO2
Proportional Pressure Regulator PPR, Trocar (subsystem
insufflator-patient system modeled by
of the applied part), and the abdominal cavity model.
MATLAB/SIMULINK-Simscape that must have
lookalike block diagrams. Whilst, our focus is going to A) PID-P Control System:
be on the application of PID-PQT (PID controller for In the preceding paper [3], the five fundamental parts
Pressure, Flow, and Temperature) & MPC-PQT (MPC (blocks) are modeled, and all detailed compartments are
controller for Pressure, Flow, and Temperature) illustrated with their specifications and parameters. The
controllers to boost the safety, efficiency, and PPR part includes PID controller driving a controllable
performance of the insufflator. Thus, additional pressure source and a pressure sensor, thus named PID-
parameters are to be added to reach the concept of P. Coming forward, it is reconsidered that controlling
“system-of-systems” in a standalone medical CO2 only the pressure is insufficient, so PID-PQT control
insufflator. On the top of that, a reality-imitating system comes to solve the issue.
simulation is to be tested by the addition of leakage on
the abdominal cavity model. B) PID-PQT Control System:
In the current paper, some variations exist where new
III. SYSTEM DESIGN compartments are added instead of others (i.e.; constant
For the modeling to be exemplary, we shall grasp the temperature sources are substituted by controllable
functionality of the real system, in addition to the basis temperature ones). Over and above, two PID controllers
of the existing models in the market. For that purpose, all are added to control CO2 temperature and CO2 gas flow,
prerequisite facts and documentations are compiled for hence called PID-PQT. So, two more parts are new to
the “40L HIGH FLOW INSUFFLATOR” model the system:
presented by STRYKER to be mimicked [7]. a) PID-T: Temperature Control Subsystem
For a well-polished model of the system, MATLAB
In Figure 2, the “constant temperature source”
tool turns up to execute the desired model.
prescribed in HPU previously is substituted with
MATLAB/SIMULINK is a very dominant tool to
“controllable temperature source”. As well, a PID
simulate various physical systems in real-time to obtain
controller and 2 temperature sensors are added to
the required results. In SIMULINK, blocks and
formulate a closed loop temperature control subsystem.
connections are used to model a problem, where
SIMULINK allows users to surpass the knowledge of
programming languages in order to model a physical
system in MATLAB and perform a real-time simulation.
More specifically, new library features (i.e.; SIMSCAPE
GAS SYSTEMS) are available in recent SIMULINK
versions. The MATLAB version used in this work
should be at least R2019b version in order to be able to
run the SIMULINK-Simscape built model.
Within model implementation, every compartment
must be calibrated and its parameters are to be
determined. Such parameters are gathered from several
resources and spreadsheets.
Figure 2: Temperature Control Subsystem
The medical CO2 insufflator-patient system
SIMULINK-Simscape Model
SIMULINK-Simscape model, which is made based on
the standard medical CO2 insufflator offered by b) PID-Q: Flow Control Subsystem
STRYKER, is shown in Figure 1. However, three control As previewed in Figure 3, a controllable volumetric
systems (PID-P, PID-PQT and MPC-PQT) will be used flow block with a volumetric flow sensor and PID
controller are added to the whole system to produce a
flow control subsystem. The block stated (-K) is only are connected to measured outputs port, 3 manipulated
utilized for unit’s conversion. variables connected to manipulated variables port and 3
references connected to references port. In addition, 3
components from the previous model are replaced with
new ones. The controllable temperature source is
replaced with variable thermal resistance, the
controllable pressure source is replaced with variable
local restriction, and the controllable volumetric flow
rate source is replaced with gate valve.

Figure 3: Flow Control Subsystem SIMULINK-

Simscape Model
The PID-PQT system model, is then represented in
Figure 4:

Figure 5: Overall SIMULINK-Simscape System

Model
In essence, three system models have been developed
using different control systems (PID and MPC) for
different variables (Pressure, Flow and Temperature).
Firstly, we started with a pressure control using a PID
controller (PID-P control system). Secondly, we used
Figure 4: PID-PQT System SIMULINK-Simscape three PID controllers for three variables (Pressure, Flow
Model and Temperature “PQT”) instead of pressure only,
C) MPC-PQT Control System: producing PID-PQT controller principle in the same
SIMULINK-Simscape model. Lastly, as a final step, we
The last model of the total system still had some have changed the control system from PID-PQT into
conflicts, concerning the inputs/outputs of the PID MPC-PQT controller. We have noticed that MPC-PQT
controller, since the latter is characterized by “SISO” controller is easier in implementation more than the PID-
functionality (SISO stands for Single-Input Single- PQT controller using the SIMULINK tool. On the other
Output). Also, massive oscillations in the output of the hand, MPC-PQT controller is more complex than PID-
system will occur, complex and continuous tuning by PQT one concerning the tuning process.
time, and non-linearity exists in the system such as
directional dependent actuators or plant dynamics. Since IV. RESULTS AND DISCUSSION
that, an MPC controller is proposed as an alternative to MATLAB/SIMULINK-Simscape is a developed
PID controllers for the sake of its ability to take MIMO simulation tool that is widely spread used to model
(MIMO stands for Multi-Input Multi-Output). Moreover, dynamical systems. There are two main control models
MPC has lower rising time, settling time and overshoots (PID-PQT and MPC-PQT) in this system to be
compared to PID controllers. Furthermore, MPC compared, and SIMULINK-Simscape favors an explicit
eliminates dangerous oscillations and provides smooth comparison between both models concerning response,
operation in transient period. Despite being very simple oscillations, fluctuations, tuning and stability.
to design and implement, MPC algorithms can control
large scale systems with many control variables, and, I) PID-PQT System Model Results:
most importantly, MPC provides a systematic technique A) Tuning Results:
of dealing with constraints on inputs and states. At last, Figure 6 shows PID controller response prior to tuning
nonlinear plant dynamics can be incorporated in MPC it for optimum response, showing obvious fluctuation
control. before reaching optimum desired signal.
Because of what is mentioned previously, the MPC Auto tuning is a unique feature in
controller in this system is placed instead of PID MATLAB/SIMULINK for tuning the PID controller.
controllers (Figure 5). It has 3 measured outputs, which Using this feature, the desired trajectory of PID
controller response can be reached with minimal T1 and T2 curves in Figure 9 represent two
response time and without excessive fluctuation to have temperature sensors’ readings: T1 for internal heating
smooth trajectory response as in Figure 7. within the device, while T2 for external heating at the
MATLAB/SIMULINK grants the user flexibility to interface-patient level. T2 is for CO2 gas temperature
adjust the parameters regarding the system’s needs. (Figure 9) which is approximately 37°C before entering
the abdominal cavity.

Figure 6: Signal before Tuning

Figure 9: Temperature Sensors’ Readings
D) Flow Results:
Flow readings are to be considered constant at 15
L/min which must remain constant all the time, with
some little range of error (± 0.5 L/min).
E) CO2 Volume Results:
The abdominal CO2 volume reaches 1.2 liters as
shown in Figure 10 (the abdominal CO2 volume unit is
expressed in liters); the model represents a limitation
without modeling the leakage and the blood absorption
of CO2 gas.

Figure 7: Signal after Tuning

B) Pressure Results:
P1 pressure sensor measures the pressure coming
from the CO2 tank (Figure 8), which must be constant
over time till it drops to zero which means emptiness of
the tank.

Figure 10: Abdominal CO2 Volume Readings

II) MPC-PQT System Model Results:
A) Pressure Results:
The pressure curve shows that stability -with
acceptable range of error ± 0.5 mmHg- is reached after a
given time (with a slight increase above stability as
displayed in Figure 11); such stability (15 mmHg) is
Figure 8: HPU-P1 Sensor Readings essential while processing.
A second P2 pressure sensor measures the pressure
after the PPR unit. The pressure drops from 88 mmHg to B) Temperature Results:
15 mmHg (set point) before entering the abdomen, that Stability of temperature is approximately reached
is the safe pressure needed; i.e. set point pressure. (around 36°C) after specific time as presented in Figure
12. This temperature is set as the reference value which
C) Temperature Results:
is significant throughout the whole procedure.
 controller has very specific ability, which is not
presented in PID controller. MPC controller has
constraints on inputs and output plus weighting them, a
very useful ability for controlling systems in reality.

Figure 11: Pressure Sensor Readings

Figure 14: Abdominal CO2 Volume Readings

V. CONCLUSION AND FUTURE WORK
In a nutshell, MIS laparoscopic surgery permits
better operations with small incisions, less hemorrhage,
and short-term recovery. It is applied at the abdominal-
pelvic site for examination or treatment purposes by the
help of medical CO2 insufflator, giving more abdominal
space in a patient.
Figure 12: Temperature Sensor Readings In the previous paper [3], the medical CO2 insufflator-
patient system is modeled and controlled via PID
C) Flow Results:
controller using MATLAB/SIMULINK-Simscape tool
The flow sensor readings using MPC controller is regarding only the pressure. Later on, two more PID
extremely appreciated. The referenced set point signal controllers, within some modifications to the system
measurement is 15 L/min, and readings in Figure 13 (additions and replacement of compartments), were
shows that it slightly exceeds the reference, but in connected to the system for temperature and flow, then
acceptable range of error (± 0.5 L/min). this total control was eventually replaced by MPC
controller to rule, monitor and regulate all variables at
once.
In this work, our focus was on the application of PID-
PQT & MPC-PQT controllers to boost the safety,
efficiency, and performance of the insufflator.
In essence, results of those three system models
manifest the significance of MPC over PID, where MPC
MIMO advantage outstands SISO one of PID,
concerning smoothness of signal, weighting constraints
on inputs and outputs, response pace, more beneficial for
dynamic control systems, and much more genuine signal
Figure 13: Flow Sensor Readings outputs.
D) CO2 Volume Results: As for future work, high flow medical CO2 insufflator
Regarding improvement of signal output, the model (>15L/min) is to be developed for better
abdominal CO2 volume signal of this model (using MPC performance. Also, more variables can be added to
controller) is much better than that of the prior model represent more realistic laparoscopic operations and
(using PID). Figure 14 shows that CO2 volume reaches 2 output signals; i.e. different leakage types can be added
liters after a given amount of time. into the abdominal cavity site model to study its impact
Summing up, it is evident that the results of the on the pressure, flow and temperature stabilities.
diverse control systems (PID-P, PID-PQT & MPC-PQT) Furthermore, the 3D camera information from
exhibit significant improvements in the safety of the laparoscopic operations which supply precise data about
medical CO2 insufflator through the use of MPC abdominal size variations can be utilized to control the
controller. By comparing these results, we can conclude medical CO2 insufflator hardware simultaneously with
that MPC-PQT controller is more superior to PID-PQT MPC-PQT controller. This demands the usage of
controller with the multi-input multi-output (MIMO) artificial intelligence (AI) and fast imaging processors
feature in medical CO2 insufflator. In addition, MPC leading to fully automated control system, and
 consequently enhancing the safety via elimination of
human errors.
REFERENCES
[1] “Types of Minimally Invasive Surgery”, JOHNS
HOPKINS MEDICINE. [Online]. Available:
https://www.hopkinsmedicine.org/minimally_invasive_rob
otic_surgery/types.html. [Accessed April 3, 2021].
[2] G. A. Vilos, A. Ternamian, J. Dempster, and P. Y. Laberge,
“No. 193-Laparoscopic Entry: A Review of Techniques,
Technologies, and Complications,” Journal of Obstetrics
and Gynaecology Canada, vol. 39, no. 7, 2017.
[3] A. ElSayed, M. I. C. HajjHassan, Z. Ali El Hawi, M.
Wadaane, A. Mohamad, H. Wehby, L. Hamawy, M.
Khayreldeen, M. Hajj-Hassan, M. A. Ali, and A. Kassem,
“MATLAB/Simulink Medical CO2 Insufflator Model with
a Pressure PID Controller,” 2020 IEEE 5th Middle East and
Africa Conference on Biomedical Engineering (MECBME),
2020.
[4] Ejaz, T. Sachs, J. He, G. Spolverato, K. Hirose, N. Ahuja,
C. L. Wolfgang, M. A. Makary, M. Weiss, and T. M.
Pawlik, “A comparison of open and minimally invasive
surgery for hepatic and pancreatic resections using the
nationwide inpatient sample,” Surgery, vol. 156, no. 3, pp.
538–547, 2014.
[5] G S. Litynski, “Kurt Semm and an automatic insufflator,”
Journal of the Society of Laparoendoscopic Surgeons, vol.
2, pp. 197-200, 1998.
[6] R. Pätznick, WOM, “WOM Reduces Time-to-Market for
Surgical Device Control Software with Model-Based
Design”. [Online]. Available:
https://es.mathworks.com/company/user_stories/wom-
reduces-time-to-market-for-surgical-device-control-
software-with-model-based-design.html. [Accessed
November 28, 2019].
[7] “Stryker 40L High Flow Insufflator Manual”, User manual.
[Online]. Available:
https://manualzz.com/doc/6284998/stryker-40l-high-flow-
insufflator-manual. [Accessed November 13, 2019].

View publication stats