A PRELOAD AND AFTERLOAD SENSITIVE

ARTIFICIAL VENTRICLE TO TEST

Paper ID: 4458
CARDIOVASCULAR PROSTHESES
M. Arabia1, F.M. Colacino1, G.A. Danieli1, F. Moscato1, S. Nicosia2, F. Piedimonte2.
1Dip. di Meccanica; Università della Calabria, Arcavacata di Rende, CS, Italia;
2Dip. Informatica, Sistemi, Produzione; Università TorVergata, Roma, Italia;

E-mail to contact author: piedimonte@robot2.disp.uniroma2.it

Methods Hydraulic layout and control scheme

Hydraulic mock circulatory systems have low flexibility to allow tests of
different cardiovascular devices and low precision in reproducing the
reference model. A new bench is described.
It combines the computer modeling of the environment in which the device
will operate, and electro-hydraulic interfaces (E-H-I). The latter are used to
link the software module with the inputs and outputs of the device under
test.
The hydraulic afterload and venous return have been reproduced by
Windkessel and Guyton’s models, respectively. This closed loop setup
represents an hydraulic simulator of the systemic circulation with left atrium.
LV
Cas

The control strategy proposed in this paper is based on the regulation of

the mean ventricular and atrial pressures. This result is obtained by
Ras
choosing a real-time simulated flow rate waveform in one cardiac cycle as
input to the system. Some parameters of this simulation are tuned to
properly estimate some physical variables of the hydraulic circuit.

Cvs
If the model of the system is enough accurate, it will be possible to use the Cla

flow rate waveforms as input signal of the open loop system, and get the
correct pressure waveforms as output. In order to achieve this result, all the Layout to test the E-H-I Electric analog of the test layout
components of the loop have been characterized, including those spurious
(always present in physical systems). Nevertheless, since the mean
systemic pressure Pms and the arterial systemic resistance Ras are not
known a-priori, it becomes necessary to continuously estimate them while
the experiment is running. The real time variations of Pms and Ras in the
simulated model lead to the correct sensitivity of the test bench to preload
and afterload changes. The unknown parameters are continuously
estimated by means of the difference between simulated ( P _ S ) and
measured (P ) mean pressures, by using the following relation:
F(k  1)  F(k )  KF P (k )  P _ S (k )
where F is the parameter to be estimated and KF is the control parameter. Control scheme

Results of the elastance-based mock loop: Afterload and Preload Sensitivity

High frequency signal represents data continuosly recorded during the experiment whose waweforms have been zoomed once steady state is reached to
show agreement with physiological data. PV loops show the Starling behavior of the hydraulic simulator. Valves have high leakage.
HR = 90 [bpm]; Cas= 1.2, Cvs= 80, Cla= 10 [ml/mmHg]; Rid= 0.035, Rii= 10, Rod= 0.065, Roi= 6 [mmHg*s/ml], and by varying Pms from 11.5 to 13.5 [mmHg] and Ras from 1.35 to 1.15 [mmHg*s/ml].

Pms increase Pms decrease Ras increase Ras decrease

Pressure data (bleu: simulated values for EHI; red: measured values into E-H-I; green: measured values into Cas)

Results in short

Pms ↑ Ras↑
Q↑ Q↓
Ves↑ Ves↑↑
Ved↑↑ Ved↑
SV↑ SV↓
Vmean↑ Vmean↑
Pas-mean↑ Pas-mean↑
Pas-min↑ Pas-min↑↑
Estimates (upper row); mean Pas-max↑↑ Pas-max↑ P-V Loops (bleu: reference, red: P-V Loops (bleu: reference,
pressures (lower row); bleu: measured,
Pms increase) red: Ras increase)
green: simulated