A hybrid mock circulatory system: development and testing of an electro-hydraulic impedance simulator.

Mock circulatory systems are used to test mechanical assist devices and for training and research purposes; when compared to numerical models, however, they are not flexible enough and rather expensive. The concept of merging numerical and physical models, resulting in a hybrid one, is applied here to represent the input impedance of the systemic arterial tree, by a conventional windkessel model built out of an electro-hydraulic (E-H) impedance simulator added to a hydraulic section. This model is inserted into an open loop circuit, completed by another hybrid model representing the ventricular function. The E-H impedance simulator is essentially an electrically controlled flow source (a gear pump). Referring to the windkessel model, it is used to simulate the peripheral resistance and the hydraulic compliance, creating the desired input impedance. The data reported describe the characterisation of the E-H impedance simulator and demonstrate its behaviour when it is connected to a hybrid ventricular model. Experiments were performed under different hemodynamic conditions, including the presence of a left ventricular assist device (LVAD).

Modelling of cardiovascular system: development of a hybrid (numerical-physical) model.

Physical models of the circulation are used for research, training and for testing of implantable active and passive circulatory prosthetic and assistance devices. However, in comparison with numerical models, they are rigid and expensive. To overcome these limitations, we have developed a model of the circulation based on the merging of a lumped parameter physical model into a numerical one (producing therefore a hybrid). The physical model is limited to the barest essentials and, in this application, developed to test the principle, it is a windkessel representing the systemic arterial tree. The lumped parameters numerical model was developed in LabVIEW environment and represents pulmonary and systemic circulation (except the systemic arterial tree). Based on the equivalence between hydraulic and electrical circuits, this prototype was developed connecting the numerical model to an electrical circuit--the physical model. This specific solution is valid mainly educationally but permits the development of software and the verification of preliminary results without using cumbersome hydraulic circuits. The interfaces between numerical and electrical circuits are set up by a voltage controlled current generator and a voltage controlled voltage generator. The behavior of the model is analyzed based on the ventricular pressure-volume loops and on the time course of arterial and ventricular pressures and flow in different circulatory conditions. The model can represent hemodynamic relationships in different ventricular and circulatory conditions.

Study of systolic pressure variation (SPV) in presence of mechanical ventilation.

Systolic pressure variation (SPV) and its components (dUp and dDown) have been demonstrated to be of interest in assessing preload in mechanically ventilated patients. The aim of this paper is to analyse the sensitivity of these variables to preload and volemic changes during mechanical ventilation in different conditions of the cardiovascular system. Computer simulation experiments have been done using a modular lumped parameter model of the cardiovascular system. The effect of mechanical ventilation has been reproduced operating on intrathoracic pressure. Experiments have been performed varying preload through filling pressure. Sensitivity of SVP dUp and dDown is described varying separately left ventricular elastance (Ev), systemic arterial resistance (Ras) and systemic arterial compliance (Cas). The sensitivity of SPV and dDown to preload and filling pressure is appreciable for high values of Ev and for a wide variation of Ras. Preliminary clinical data concerning the three parameters show good correlation with simulation results.

A hybrid (numerical-physical) model of the left ventricle.

Hydraulic models of the circulation are used to test mechanical devices and for training and research purposes; when compared to numerical models, however, they are not flexible enough and rather expensive. The solution proposed here is to merge the characteristics and the flexibility of numerical models with the functions of physical models. The result is a hybrid model with numerical and physical sections connected by an electro-hydraulic interface - which is to some extent the main problem since the numerical model can be easily changed or modified. The concept of hybrid model is applied to the representation of ventricular function by a variable elastance numerical model. This prototype is an open loop circuit and the physical section is built out of a reservoir (atrium) and a modified windkessel (arterial tree). The corresponding equations are solved numerically using the variables (atrial and arterial pressures) coming from the physical circuit. Ventricular output flow is the computed variable and is sent to a servo amplifier connected to a DC motor-gear pump system. The gear pump, behaving roughly as a flow source, is the interface to the physical circuit. Results obtained under different hemodynamic conditions demonstrate the behaviour of the ventricular model on the pressure-volume plane and the time course of output flow and arterial pressure.

Mono and bi-ventricular assistance: their effect on ventricular energetics.

When mono- and bi-ventricular mechanical assistance is used for heart recovery, its control strategy and circulatory variables affect ventricular energetics (external work-EW, oxygen consumption-VO2, cardiac mechanical efficiency-CME). This study is based on the data obtained in vitro and presents an analysis of the effects of the mono- and bi-ventricular mechanical assistance on ventricular energetics. The assistance was conducted on the principle of counterpulsation with atrio-arterial connection. It includes the following stages: 1) the characterisation of the isolated ventricle model in terms of EW, VO2 and CME as a function of the filling pressure and peripheral resistance, 2) modelling of left ventricular and pulmonary dysfunction, followed by left ventricular and bi-ventricular assistance. Experimental data enable us to draw the following conclusions: * in general, the greatest hemodynamic improvement does not correspond to the highest energetic improvement, * LVAD assistance deteriorates left ventricular CME while its effect on right ventricular energetics depends on the value of right ventricular elastance (Emax). Right ventricular CME is deteriorated by BVAD assistance irrespective of right Emax, * the energetics optimisation in bi-ventricular assistance is closely related to the right Emax, which could probably be a deciding factor in the choice of the assistance mode.

IABP assistance: a test bench for the analysis of its effects on ventricular energetics and hemodynamics.

IABP assistance is frequently used to support heart recovery, improving coronary circulation and re-establishing the balance between oxygen availability and consumption. Hemodynamic and energetic parameters (endocardial viability ratio, ventricular energetics) are used to evaluate its effectiveness which depends on internal (timing, balloon volume and position) and external factors (circulatory conditions). Considering short, medium and long-term effects of IABP, the first depends on its mechanical action, the latter on the changes induced in circulatory parameters. The analysis of the first is important because conditions for the onset of a virtuous cycle able to support ventricular recovery are created. Simulation systems could be helpful in this analysis for the implicit reliability and reproducibility of the experiments, provided that they are able to reproduce both hemodynamic phenomena and energetic relationships. The aim of this paper is to present a system originally developed to test mechanical heart assist devices and modified for IABP testing. Data reported here are obtained from in vitro experiments. A partial verification, obtained from the literature is presented.

A physical model of the human systemic arterial tree.

A physical model of the human arterial tree has been developed to be used in a computer controlled mock circulatory system (MCS). Its aim is to represent systemic arterial tree properties and extend the capacity of the MCS to intraortic balloon pump (IABP) testing. The main problem was to model the aorta simply and to accurately reproduce aortic impedance and related flow and pressure waveforms at different sections. The model is composed of eight segments; lumped parameter models are used for its peripheral loads. After the numerical simulation, the physical model was reproduced as a silicon rubber tapered tube. This rubber was chosen for its stability over time and the acceptable behaviour of its Young's modulus (Ey = 22.23 gf x mm(-2)) with different loads and in comparison with data from the literature (Ey approximately 20.4 gf x mm(-2)). The properties of each segment of the aorta were defined in terms of compliance, resistance and inertance as a function of length, radius and thickness. The variable thickness was obtained using positive and negative molds. Total static compliance of the aorta model is about 1.125 x 10(-3) g(-1) x cm4 x sec2 (1.5 cm3 x mmHg(-1)). Measurements were performed both on numerical and physical models (in open and closed loop configuration). Data reported show pressure and flow waveforms along with input impedance modulus and phase. The results are in good agreement with data from the literature.

A virtual instrument (VI) for haemodynamic management in ICU and during surgery.

Systolic pressure variations (SPV) during mechanical ventilation and its single components, related to short apnea, reflect changes of the volemic condition of the patient. To introduce their determination during clinical monitoring for different fluid states and for different tidal volumes, they must be computed on-line without introducing interference with standard activities. A system computing on-line systolic pressure variation during mechanical ventilation, connected to standard monitoring devices, has been proposed. It is based on a notebook PC implemented with graphical software comprising a user panel in the form of a virtual instrument and is able to acquire, process and present signals from different instruments utilized in ICU and during surgery. It can be used as a base to assess the ability of computed parameters helpful in clinical decision. The use of a notebook PC and open software allows operators, even if non-expert in computer science, to test and implement this, as well as other innovative tools in clinical practice.

A simple method for Emax trend evaluation: in vitro and in vivo results.

The aim of this study is the evaluation of end systolic ventricular elastance trend (as a measure of heart contractility) by hemodynamic variables available in intensive care units or during heart surgery: heart rate, cardiac output, left atrial, mean and diastolic arterial pressure. Its basic assumption is the description of ejection as the interaction between variable left ventricular and arterial compliances (reciprocal of the corresponding elastances) connected in parallel. As pressure is the same in each compliance at systole beginning and ending, ventricular elastance can be estimated by assuming that energy variation is the same on both compliances. The algorithm has been tested on a numerical simulator of the circulatory system and on six sheep at basal conditions and during drug infusion. Correlation function in numerical simulation, between true and computed ventricular elastance (range 0.45 divided by 5 mm Hg-cm(-3)), yields 0.985. In vivo comparison between computed ventricular elastance trend and ventricular dp/dt trend yields a correlation function ranging between 0.87 and 0.99. The result of the algorithm cannot be assumed to be Emax value. However, it can be considered a contractility index as it closely follows any change in dp/dt. It can be computed by simple calculations and needs no variables other than those usually measured in intensive care . It allows the extrapolation of useful information for evaluating the trend in heart contractility and for setting up a control strategy for mechanical or pharmacological assistance during heart recovery.

A computer controlled mock circulatory system for mono- and biventricular assist device testing.

The clinical use of heart assist devices for heart recovery, implies the problem of their in vitro testing and training to use. In a mock circulatory system developed to this aim, the main problem is reproducing interaction among the device, the ventricle and the circulatory network. This can be analysed by the position, on the p-v plane, of the working point defined by the intersection between end systolic ventricular (ESPVR) and arterial elastance lines. The system developed on this basis, connectable to mono- and biventricular parallel assist devices, was a closed loop model including systemic and pulmonary circulation. The arterial trees were reproduced by two windkessels with adjustable peripheral resistance, and the Starling's law of the heart by a variable elastance model. The software controls and monitors circulatory parameters and variables. Results showed the behavior of the system with preload or afterload changes. Further, the reproduction of physiological, pathological (obtained by modifying slope and volume intercept of the ESPVR line) and LVAD assisted circulatory conditions was shown. The assistance effect was underlined by the changes in the ventricular work cycle and in hemodynamics variables. The evaluation of the effect of device control strategy on the ventricle and its energetics (on p-v plane) were among the main characteristics of this system, which ought to be further improved to test devices such as the IABP, which requires a different aortic model.

In vitro testing of a left ventricular assist device. Study of the effect of its control strategy on energetic relationships inside the left ventricle.

In this study an original left ventricular assist device is tested on an open loop modular physical circuit reproducing Starling's law of the heart to set an optimal control strategy for heart recovery. It is assumed that the goals of the assistance are reduction of oxygen consumption, external work and improvement of cardiac mechanical efficiency. The assistance is evaluated by the position of a working point on the characteristic surfaces of the ventricle defined by peripheral resistance, atrial pressure and selected variables pertaining to energy (pressure-volume area, external work and cardiac mechanical efficiency). In this frame an optimal assistance for heart recovery is a compromise among different requirements corresponding to a restricted set of control parameters values: driving pressure PZB = 35 kPa and timing values T1 and T2 (systole beginning and systole ending in relation to QRS complex and cardiac cycle duration T) T1 = 0.55.T and T2 = 0.73.T.

Mock circulatory system for in vitro reproduction of the left ventricle, the arterial tree and their interaction with a left ventricular assist device.

A modular physical circuit for testing monoventricular and biventricular heart assist devices (HAD) is under development. The modules now available are assembled in an open-loop circuit and reproduce the function of the left ventricle and the systemic arterial tree. The left ventricle model reproduces Starling's law of the heart and can be easily controlled to modify other parameters such as contractility and timing (i.e. heart rate and systole/diastole ratio). This circuit, in connection with a left ventricular assist device (LVAD), can be used to evaluate the LVAD performance, its effect on the circulatory system and as a training system. This paper is devoted to a description of the circuit and of its interaction with a LVAD, which is analysed after the simulation of a low contractility pathology of the ventricle. Results obtained in these experiments are reported.

A desk-top computer model of the circulatory system for heart assistance simulation: effect of an LVAD on energetic relationships inside the left ventricle.

The study of the interaction between a pneumatic left ventricle assist device (LVAD), driven with different control strategies, and the cardiovascular system is the subject of this paper. It is performed by a modular numerical model of the cardiovascular system connected to a numerical model of the LVAD. The circulatory system is simulated by a lumped parameter numerical model. The ventricle is represented by a time-varying elastance model to reproduce the Starling law of the heart. The effect of the LVAD on the cardiovascular system is evaluated, on the left ventricle alone, by an open-loop circuit consisting of the models of the ventricle, the LVAD and the arterial tree. The analysis is performed in terms of energy variables (such as external work and oxygen consumption and cardiac mechanical efficiency versus control strategy. The LVAD is driven by different control strategies: a fixed heart rate (with different delays from the onset of the natural ventricle contraction) and a variable heart rate.

A modular numerical model of the cardiovascular system for studying and training in the field of cardiovascular physiopathology.

A modular numerical model of the cardiovascular system has been developed to reproduce the most important circulatory phenomena in terms of pressure and volume relationships. It is an easy tool to use, designed to be used with a friendly approach on any IBM or compatible personal computer: it offers a wide selection of graphical and numerical outputs and can be rearranged easily for a particular experiment. A set of subroutines related to different circulatory phenomena has been developed; they can be assembled easily together and communicate with each other by two variables. A full description of the existing subroutines is presented in this paper with three different application examples resulting from the rearrangement of the existing software modules: the first concerns the behaviour of the natural ventricle model itself and can be regarded as a ventricle stand-alone characterization test in terms of preload and afterload sensitivities, the second is related to the use of a pneumatic ventricle instead of the model of the natural ventricle, and the third is a full model of the cardiocirculatory system.

Effect of magnetic fluids on the efficiency of an electromagnetic actuator prototype.

An electromagnetic actuator to substitute ventricular function presents some advantages compared with different energy converters. The drastic reduction of the moving parts leads to greater reliability and an accurate control system can be set up. Its major limitations concern weight, heat dissipation and, finally, the overall pump efficiency which is usually rather low. We investigated the possibility of using magnetic fluids in an electromagnetic actuator. Limitations intrinsic to the magnetic fluids prevent their being used as pumping elements but they can be useful to increase the pump efficiency by reducing losses in the magnetic circuit. A remarkable increase in pump efficiency was attained. It is necessary to point out that in designing the electromagnetic actuator the focus was on the pump efficiency with and without the magnetic fluids more than on its performance.

Hemodialysis with "adequate" sodium concentration in dialysate.

This investigation was undertaken to define the "adequate" sodium concentration in the dialytic fluid allowing to maintain a stable plasma effective osmolality during dialysis. Isonatric dialysate is shown to miss this aim by inducing a predictable postdialytic hypernatremia. To avoid this effect a new approach was made. 17 clinically stabilized patients, previously dialyzed over a period of at least 2 years with a dialysate sodium concentration of 133 mEq/l, underwent dialysis with the "adequate" sodium concentration in the dialysate for over 3 years. During dialysis cramps, headache, hypotension, hypertensive crises and postdialytic weakness were reduced in frequency and nearly disappeared. No deterioration in blood pressure control occurred and improvement in some general parameters (hematocrit, glucose and insulin metabolism, well-being) was reported after prolonged treatment.

[Clinical coronarographic characteristics and pathogenetic mechanism of angina pectoris with s-t elevation (author's transl)]. / Caratteristiche cliniche, coronarografiche e meccanismo patogenetico dell'angina pectoris con sopraslivellamento del tratto S-T

The study of 46 patients with frequent anginal episodes characterized by S-T elevation (so called "variant angina pectoris") demonstrated that this type of electrocardiographic pattern does not characterize a homogeneous group of patients. In fact, while in some patients angina occurred only at rest, in others it occurred also on exercise. Sometimes ecgraphic alterations characterized by S-T depression were observed on the same leads which on other occasions had shown S-T elevation. The angiographic picture revealed: absence of significant coronary alterations in 10% of cases, stenosis greater than 75% in one main branch in 29%, in two branches in 39% and in three branches in 22% of cases. The hemodynamic monitoring carried out on 14 of these patients demonstrated that the ecgraphic modifications occur before the onset of the hemodynamic parameters which control myocardial O2 consumption. This suggests a primitive reduction of regional myocardial blood supply as a cause of the ischaemic episodes. The study of the regional myocardial perfusion with 201Tl technique in 6 patients confirmed this hypothesis. Coronary angiography carried out during an ischemic episode showed that the reduction of myocardial blood supply was caused by a spasm of a large coronary artery involving a long segment of the vessel, reversible by nitroglycerin administration. Aorto-coronary by-pass operation performed on 6 patients was followed by the disappearance of pain in two patients, even though the "by-pass" patency was angiographically proved in two patients.