Development of a flow estimation and control system of an implantable centrifugal blood pump for circulatory assist

A model for estimating the blood flow of the POLVAD pulsatile ventricular assist device

OBJECTIVE

The aim of this work was to model the blood flow rate of the POLVAD-MEV pulsatile ventricular assist device (VAD). An adequate flow rate is crucial to restore physiological cardiac output. Unfortunately, during clinical heart support, neither blood flows nor pressures can be measured within the device. In general, the flow rate depends on the control parameters and patient conditions. However, the patient's hemodynamic parameters are not constantly monitored. Therefore, blood flow must be evaluated based on the standard measurements from the device control unit.

METHODS

The model identification data were taken from a research stand consisting of a VAD connected to a hybrid cardiovascular simulator. The studies were conducted under different work and control conditions. A compound model of a ventricular assist device was proposed. First, the driving pressure waveform for an idle run of the supply unit is modeled. Next, the blood flow is estimated based on the difference between the measured value of driving pressure and the modeled value for an idle run.

RESULTS

The quality of the developed model is good (R=0.92) and similar for all tested cases, confirming the high versatility of the proposed solution.

CONCLUSION

The blood flow rate is estimated based on standard signals from the device control unit; therefore, no additional measurements are necessary.

SIGNIFICANCE

The developed model application in the VAD control unit will aid the selection of control parameters and might be useful for development of adaptive control system. A preliminary version of this work was reported at the [1].

Outflow monitoring of a pneumatic ventricular assist device using external pressure sensors

BACKGROUND

In this study, a new algorithm was developed for estimating the pump outflow of a pneumatic ventricular assist device (p-VAD). The pump outflow estimation algorithm was derived from the ideal gas equation and determined the change in blood-sac volume of a p-VAD using two external pressure sensors.

OBJECTIVES

Based on in vitro experiments, the algorithm was revised to consider the effects of structural compliance caused by volume changes in an implanted unit, an air driveline, and the pressure difference between the sensors and the implanted unit.

METHODS

In animal experiments, p-VADs were connected to the left ventricles and the descending aorta of three calves (70-100 kg). Their outflows were estimated using the new algorithm and compared to the results obtained using an ultrasonic blood flow meter (UBF) (TS-410, Transonic Systems Inc., Ithaca, NY, USA).

RESULTS

The estimated and measured values had a Pearson's correlation coefficient of 0.864. The pressure sensors were installed at the external controller and connected to the air driveline on the same side as the external actuator, which made the sensors easy to manage.

Viscosity-adjusted estimation of pressure head and pump flow with quasi-pulsatile modulation of rotary blood pump for a total artificial heart

Estimation of pressure and flow has been an important subject for developing implantable artificial hearts. To realize real-time viscosity-adjusted estimation of pressure head and pump flow for a total artificial heart, we propose the table estimation method with quasi-pulsatile modulation of rotary blood pump in which systolic high flow and diastolic low flow phased are generated. The table estimation method utilizes three kinds of tables: viscosity, pressure and flow tables. Viscosity is estimated from the characteristic that differential value in motor speed between systolic and diastolic phases varies depending on viscosity. Potential of this estimation method was investigated using mock circulation system. Glycerin solution diluted with salty water was used to adjust viscosity of fluid. In verification of this method using continuous flow data, fairly good estimation could be possible when differential pulse width modulation (PWM) value of the motor between systolic and diastolic phases was high. In estimation under quasi-pulsatile condition, inertia correction was provided and fairly good estimation was possible when the differential PWM value was high, which was not different from the verification results using continuous flow data. In the experiment of real-time estimation applying moving average method to the estimated viscosity, fair estimation could be possible when the differential PWM value was high, showing that real-time viscosity-adjusted estimation of pressure head and pump flow would be possible with this novel estimation method when the differential PWM value would be set high.

Indirect flow rate estimation of the NEDO PI Gyro pump for chronic BVAD experiments

In totally implantable ventricular assist device systems, measuring flow rate of the pump is necessary to ensure proper operation of the pump in response to the recipient's condition or pump malfunction. To avoid problems associated with the use of flow probes, several methods for estimating flow rate of a rotary blood pump used as a ventricular assist device have been studied. In the present study, we have performed a chronic animal experiment with two NEDO PI gyro pumps as the biventricular assist device for 63 days to evaluate our estimation method by comparing the estimated flow rate with the measured one every 2 days. Up to 15 days after identification of the parameters, our estimations were accurate. Errors increased during postoperation days 20 to 30. Meanwhile, their correlation coefficient r was higher than 0.9 in all the acquired data, and estimated flow rate could simulate the profile of the measured one.

An Intelligent Remote Monitoring System for Artificial Heart

A web-based database system for intelligent remote monitoring of an artificial heart has been developed. It is important for patients with an artificial heart implant to be discharged from the hospital after an appropriate stabilization period for better recovery and quality of life. Reliable continuous remote monitoring systems for these patients with life support devices are gaining practical meaning. The authors have developed a remote monitoring system for this purpose that consists of a portable/desktop monitoring terminal, a database for continuous recording of patient and device status, a web-based data access system with which clinicians can access real-time patient and device status data and past history data, and an intelligent diagnosis algorithm module that noninvasively estimates blood pump output and makes automatic classification of the device status. The system has been tested with data generation emulators installed on remote sites for simulation study, and in two cases of animal experiments conducted at remote facilities. The system showed acceptable functionality and reliability. The intelligence algorithm also showed acceptable practicality in an application to animal experiment data.

Measurement for implantable rotary blood pumps

Rotary blood pumps offer a cost-effective way to assist the failing heart. Relative to their pulsatile cousins, they can consist of remarkably few moving parts, with attendant advantages in reliability. These advantages are realized in full only if the entire assist system is kept maximally simple. Control of the pump must therefore be based on a minimum number of measurement devices. This paper reviews the measurements that are made in the wide range of implantable rotary blood pump designs that are in development for ventricular assist. In a number of these, fluid-mechanical variables are estimated indirectly from measurements of motor speed and current or power. The introduction explains the goals of rotary blood pump control by comparison to the innate properties of the natural heart. Then motor and fluid-mechanical variables that may be transduced are discussed. Methods of indirect estimation of pressure drop and flow-rate are dealt with, followed by ways of detecting unusual states such as inflow obstruction. It is found that detection of these alone can be the basis of an adequate control strategy. Some groups have estimated variables pertaining to the heart that is being assisted, and there has also been work on monitoring the ongoing health of the assist system itself. The review concludes with a brief look at the wider measurement context for the intensive-care facility that proposes to use such devices to provide circulatory support.

Estimation of coupled assist flows in a moving-actuator bi-ventricular assist device using interventricular pressure

An assist flow estimation scheme for a moving-actuator biventricular assist device (MA-BVAD) using interventricular pressure (IVP) has been developed. The scheme uses a waveform feature parameter of IVP, peak IVP time (PIT), for estimation of the filling volumes of both left and right blood sacs simultaneously. In a regression analysis on data from an in vivo test in an 85 Kg male calf for 20 days, the PIT was found to have high correlation with the blood sac filling volume (R=0.883: left filling volume, R=0.967: right filling volume). A conceptual equation hypothesizing this correlation between PIT and filling volume was established based on the observation and the unknown parameters were identified using least squares parameter optimization. The estimation equation identified proved highly accurate (R=0.916 for left flow, R=0.970 for right flow). The accuracy of the estimation scheme promises very good practical applicability.

In vivo test of pressure head and flow rate estimation in a continuous-flow artificial heart

To avoid using sensors with low biocompatibility and low durability in implantable total artificial heart (TAH) systems, the authors previously proposed a new method for estimating instantaneous values of flow rate and pressure head on the basis of voltage, current, and rotational speed in a motor driven centrifugal pump. The previous in vitro experiments showed that the proposed estimator could automatically compensate for the effect of the change in blood viscosity on the estimation accuracy by employing two kinds of autoregressive exogenous models. In this study, validity and reliability of this estimation method were ascertained in an acute animal experiment. In the experiment, two centrifugal blood pumps were implanted into an adult goat as a total artificial heart. Results of estimation were compared with true values when blood viscosity was changed by injecting physiological saline. The results indicated that the system could successfully estimate pressure head by compensating the change of viscosity, although the estimation accuracy of the in vivo estimation was not so high as that of the previous in vitro tests.

Sensorless estimation of pressure head and flow of a continuous flow artificial heart based on input power and rotational speed

The present study has proposed a new method for estimating the pressure head (P(t)[mm Hg]) and flow (Q(t)[L/min]) of a centrifugal pump on the basis of voltage (V(t)[V]), current (I(t)[A]), and rotational speed (N(t)[k(rpm)]) of the DC motor for a pump without any additional sensors. In the proposed estimation method, two auto-regressive exogenous (ARX) models are employed. One ARX model has an output, P(t) or Q(t), and three inputs, VI(t) = V(t)I(t) and N(t) and the steady state gain (K) of the system from VI(t) to N(t). It can be assumed that K may include the information on viscosity of blood. The coefficient parameters of this ARX model are identified in an off-line fashion before implantation of the pump. After implantation, P(t) or Q(t) is estimated by the same ARX model with the already identified parameters. The other ARX model is used to identify Kon the basis of VI(t) and N(t) in an on-line fashion every time the viscosity of blood may change. In the experiment, a mock circulatory system consisting of a centrifugal pump and a reservoir with 37% glycerin or water was employed. The root mean square error between measured Q(t) and its estimate obtained from the proposed method was 1.66L/min. On the other hand, a different method based on a single ARX model with inputs of VI(t) and N(t), but without the additional input of K, yielded the corresponding estimation error of 2.22L/min. This means that the proposed method can reduce its estimation error by about 25% in comparison with a method that cannot cope with the change in blood viscosity.

The role of diastolic pump flow in centrifugal blood pump hemodynamics

We tried to verify the hypothesis that increases in pump flow during diastole are matched by decreases in left ventricular (LV) output during systole. A calf (80 kg) was implanted with an implantable centrifugal blood pump (EVAHEART, SunMedical Technology Research Corp., Nagano, Japan) with left ventricle to aorta (LV-Ao) bypass, and parameters were recorded at different pump speeds under general anesthesia. Pump inflow and outflow pressure, arterial pressure, systemic and pulmonary blood flow, and electrocardiogram (ECG) were recorded on the computer every 5 ms. All parameters were separated into systolic and diastolic components and analyzed. The pulmonary flow was the same as the systemic flow during the study (p > 0.1). Systemic flow consisted of pump flow and LV output through the aortic valve. The ratio of systolic pump flow to pulmonary flow (51.3%) did not change significantly at variable pump speeds (p > 0.1). The other portions of the systemic flow were shared by the left ventricular output and the pump flow during diastole. When pump flow increased during diastole, there was a corresponding decrease in the LV output (Y = -1.068X + 51.462; R(insert)(2) = 0.9501). These show that pump diastolic flow may regulate expansion of the left ventricle in diastole.

Relationship of blood pressure and pump flow in an implantable centrifugal blood pump during hypertension

The purpose of this study was to evaluate the real time relationship between pump flow and pump differential pressure (D-P) during experimentally induced hypertension (HT). Two calves (80 and 68 kg) were implanted with the EVA-HEART centrifugal blood pump (SunMedical Technology Research Corp., Nagano, Japan) under general anesthesia. Blood pressure (BP) in diastole was increased to 100 mm Hg by norepinephrine to simulate HT. Pump flow, D-P, ECG, and BP were measured at pump speeds of 1,800, 2,100, and 2,300 rpm. All data were separated into systole and diastole, and pump flow during HT was compared with normotensive (NT) conditions at respective pump speeds. Diastolic BP was increased to 99.3+/-4.1 mm Hg from 66.5+/-4.4 mm Hg (p<0.01). D-P in systole was under 40 mm Hg (range of change was 10 to 40 mm Hg) even during HT. During NT, the average systolic pump flow volume was 60% of the total pump flow. However, during HT, the average systolic pump flow was 100% of total pump flow volume, although the pump flow volume in systole during HT decreased (33.1+/-5.7 vs. 25.9+/-4.0 ml/systole, p<0.01). In diastole, the average flow volume through the pump was 19.6+/-6.9 ml/diastole during NT and -2.2+/-11.1 ml/diastole during HT (p<0.01). The change in pump flow volume due to HT, in diastole, was greater than the change in pump flow in systole at each pump speed (p<0.001). This study suggests that the decrease of mean pump flow during HT is mainly due to the decrease of the diastolic pump flow and, to a much lesser degree, systolic pump flow.

Continuously maintaining positive flow avoids endocardial suction of a rotary blood pump with left ventricular bypass

This study showed the usefulness of maintaining positive pump flow to avoid endocardial suction and as an assist bypass. Three calves were implanted with centrifugal pumps. Hemodynamics and pump parameters were measured at varying pump speeds (from 1,100 to 2,300 rpm). In each test pump, speed was adjusted to create 3 hemodynamic states: both positive and negative flow (PNF), positive and zero flow (PZF), and continuously positive flow (CPF). The pump flow volume was determined during systole (Vs) and diastole (Vd). Vs in PNF was 29.6 ml and was not significantly different from Vs in PZF (p > 0.15). Vd in PNF was significantly different from Vd in PZF (p < 0.05). All bypass rates of PNF were over 30% of pulmonary flow. All PZF bypass rates were between the PNF rate and the CPF rate. These data showed that PZF satisfied the minimum requirement of assist flow and was under 100% bypass. Thus, PZF may avoid endocardial suction.

Estimation of pump flow rate and abnormal condition of implantable rotary blood pumps during long-term in vivo study

The control system for an implantable rotary blood pump is not clearly defined. A detection system is considered to be necessary for pump flow monitoring and abnormal conditions such as back flow or a sucking phenomenon where the septum or left ventricle wall is sucked into the cannula, etc. The ultrasound flowmeter is durable and reliable but the control system should not be totally dependent on the flowmeter. If the flowmeter breaks, the rotary blood pumps have no control mechanism. Therefore, the authors suggest controlling the pumps by an intrinsic parameter. One left ventricular assist device (LVAD) calf model was studied where the flow rate and waveform of the pump flow proved to identify the sucking phenomenon. Thus, the pump flow rate was calculated from the required power, motor speed, and heart rate. The value of the coefficient of determination (R2) between the measured and estimated pump flow rate was 0.796. To estimate this abnormal phenomenon, 2 methods were evaluated. One method was the total pressure head in which the pump flow rate and motor speed were estimated. During normal conditions the total pressure head is 79.5 +/- 7.0 mm Hg whereas in the abnormal condition, it is 180.0 +/- 2.8 mm Hg. There was a statistical difference (p < 0.01). Another method is using a current waveform. There is an association between the current and pump flow waves. The current was differentiated and squared to calculate the power of the differentiated current. The normal range of this value was 0.025 +/- 0.029; the abnormal condition was 11.25 +/- 15.13. There was a statistical difference (p < 0.01). The predicted flow estimation method and a sucking detection method were available from intrinsic parameters of the pump and need no sensors. These 2 methods are simple, yet effective and reliable control methods for a rotary blood pump.