Physiologic control of cardiac assist devices

Simulation of aortic valve dynamics during ventricular support

Existing commercially-used left ventricular assist devices (LVADs) make no attempt to automatically detect the aortic valve condition in their control methods to optimize ventricular assistance. An important design goal for LVADs is the ability to reliably and accurately detect aortic valve (AV) states during heart pump support that can cause harmful effects on AV structure and function. In this paper, we have investigated the correlation between AV performance and LVAD motor current as well as speed set points, simulating aortic valve blood flow, pressure, pump flow and LV mechanics using a simplified two-dimensional fluid-structure interaction finite-element model of AV dynamics.

NONINVASIVE DETECTION OF SUCTION IN AN IMPLANTABLE ROTARY BLOOD PUMP USING NEURAL NETWORKS

Granting those heart failure patients who are recipients of an implantable rotary blood pump (iRBP) greater functionality in daily activities is a key long-term strategy currently being pursued by many research groups. A reliable technique for noninvasive detection of the various pumping states, most notably that of ventricular collapse or suction, is an essential component of this strategy. Presented in this study is such a technique, whereby various indicators are derived from the noninvasive pump feedback signals, and a suitable computational methodology developed to classify the pumping states of interest. Clinical telemetry data from ten implant recipients was categorized (with the aid of trans-oesophageal echocardiography) into the normal and suction states. These data are used to develop a pumping state classifier based on an artificial neural network (ANN). Nine indices, derived from the noninvasive impeller speed signal, form the inputs to this ANN classifier. During validation, the resulting ANN classifier achieved a maximum sensitivity of 98.54% (609/618 samples of 5 s in length) and specificity of 99.26% (12,123/12,213 samples) for correct detection of the suction state. The ability to detect the suction state with such a high degree of accuracy provides a critical parameter both for control strategy development, and for clinical care of the implant recipient.

Classification of Physiologically Significant Pumping States in an Implantable Rotary Blood Pump: Patient Trial Results

An integral component in the development of a control strategy for implantable rotary blood pumps is the task of reliably detecting the occurrence of left ventricular collapse due to overpumping of the native heart. Using the noninvasive pump feedback signal of impeller speed, an approach to distinguish between overpumping (or ventricular collapse) and the normal pumping state has been developed. Noninvasive pump signals from 10 human pump recipients were collected, and the pumping state was categorized as either normal or suction, based on expert opinion aided by transesophageal echocardiographic images. A number of indices derived from the pump speed waveform were incorporated into a classification and regression tree model, which acted as the pumping state classifier. When validating the model on 12,990 segments of unseen data, this methodology yielded a peak sensitivity/specificity for detecting suction of 99.11%/98.76%. After performing a 10-fold cross-validation on all of the available data, a minimum estimated error of 0.53% was achieved. The results presented suggest that techniques for pumping state detection, previously investigated in preliminary in vivo studies, are applicable and sufficient for use in the clinical environment.

Classification of physiologically significant pumping states in an implantable rotary blood pump: effects of cardiac rhythm disturbances

Methods of speed control for implantable rotary blood pumps (iRBPs) are vital for providing implant recipients with sufficient blood flow to cater for their physiological requirements. The detection of pumping states that reflect the physiological state of the native heart forms a major component of such a control method. Employing data from a number of acute animal experiments, five such pumping states have been previously identified: regurgitant pump flow, ventricular ejection (VE), nonopening of the aortic valve (ANO), and partial collapse (intermittent [PVC-I] and continuous [PVC-C]) of the ventricle wall. An automated approach that noninvasively detects such pumping states, employing a classification and regression tree (CART), has also been developed. An extension to this technique, involving an investigation into the effects of cardiac rhythm disturbances on the state detection process, is discussed. When incorporating animal data containing arrhythmic events into the CART model, the strategy showed a marked improvement in detecting pumping states as compared to the model devoid of arrhythmic data: state VE--57.4/91.7% (sensitivity/specificity) improved to 97.1/100.0%; state PVC-I--66.7/83.1% improved to 100.0/88.3%, and state PVC-C--11.1/66.2% changed to 0.0/100%. With a simplified binary scheme differentiating suction (PVC-I, PVC-C) and nonsuction (VE, ANO) states, suction was initially detected with 100/98.5% sensitivity/specificity, whereas with the subsequent improved model, both these states were detected with 100% sensitivity. The accuracy achieved demonstrates the robustness of the technique presented, and substantiates its inclusion into any iRBP control methodology.

Identification and Classification of Physiologically Significant Pumping States in an Implantable Rotary Blood Pump

In a clinical setting it is necessary to control the speed of rotary blood pumps used as left ventricular assist devices to prevent possible severe complications associated with over- or underpumping. The hypothesis is that by using only the noninvasive measure of instantaneous pump impeller speed to assess flow dynamics, it is possible to detect physiologically significant pumping states (without the need for additional implantable sensors). By varying pump speed in an animal model, five such states were identified: regurgitant pump flow, ventricular ejection (VE), nonopening of the aortic valve over the cardiac cycle (ANO), and partial collapse (intermittent and continuous) of the ventricle wall (PVC-I and PVC-C). These states are described in detail and a strategy for their noninvasive detection has been developed and validated using (n = 6) ex vivo porcine experiments. Employing a classification and regression tree, the strategy was able to detect pumping states with a high degree of sensitivity and specificity: state VE-99.2/100.0% (sensitivity/specificity); state ANO-100.0/100.0%; state PVC-I- 95.7/91.2%; state PVC-C-69.7/98.7%. With a simplified binary scheme differentiating suction (PVC-I, PVC-C) and nonsuction (VE, ANO) states, both such states were detected with 100% sensitivity.

Automated non-invasive detection of pumping states in an implantable rotary blood pump

With respect to rotary blood pumps used as left ventricular assist devices (LVADs), it is clinically important to control pump flow to avoid complications associated with over-or under-pumping of the native heart. By employing only the non-invasive observer of instantaneous pump impeller speed to assess flow dynamics, a number of physiologically significant pumping states may be detected. Based on a number of acute animal experiments, five such states were identified: regurgitant pump flow (PR), ventricular ejection (VE), non-opening of the aortic valve (ANO), and partial collapse (intermittent and continuous) of the ventricle wall (PVC-I and PVC-C). Two broader states, normal (corresponding to VE, ANO) and suction (corresponding to PVC-I, PVC-C) were readily discernable in clinical data from human patients implanted with LVADs. Based on data from both the animal experiments (N=6) and the human patients (N=10), a strategy for the automated non-invasive detection of significant pumping states has been developed and validated. Employing a classification and regression tree (CART), this system detects pumping states with a high degree of accuracy: state VE -87.5/100.0% (sensitivity/specificity); state ANO - 98.1/92.5%; state PVC-I - 90.0/90.2%; state PVC-C - 61.2/98.0%. With a simplified binary scheme differentiating suction and normal states, both states were detected without error in data from the animal experiments, and with a sensitivity/specificity, for detecting suction, of 99.2/98.3% in the human patient 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.

Non-invasive flow estimation in an implantable rotary blood pump: a study considering non-pulsatile and pulsatile flows

Non-invasive estimation of flow was investigated in an implantable rotary blood pump (iRBP) with a hydrodynamic bearing. The effects of non-pulsatile and pulsatile flows were studied using in vitro mock loops, and acute (N = 3) and chronic (N = 6) ovine experiments. Using the non-pulsatile and pulsatile mock loops an average flow estimation algorithm was derived from root mean square (RMS) pump impeller speed and RMS input power. These algorithms were programmed into the iRBP controller for subsequent validation in vivo. In the acute experiments, venous return and systemic vascular resistance were adjusted through pharmacological intervention and exsanguination to produce an average range of pump flows from 0.0 to 2.6 l min(-1). Over this range the RMS estimation error was 88 +/- 12 ml, with a linear correlation slope of 0.992 +/- 0.006 (R2 = 0.986 +/- 0.004). In the chronic experiments, animals were monitored daily for up to three months and an average range of flows from 2.8 to 4.8 l min(-1) recorded. A linear correlation between the estimated and measured pump flows yielded a slope of 1.005 +/- 0.006 (R2 = 0.966 +/- 0.004). The RMS estimation error was 120 +/- 11 ml. Using this algorithm it is possible to effectively estimate flow in a rotary blood pump without implanting additional invasive sensors.

A safe automatic driving method for a continuous flow ventricular assist device based on motor current pulsatility: in vitro evaluation

We previously reported that detection of two specific points (the t-point and the s-point) in the relationship between pump speed and Motor Current Amplitude index (ICA) indicates the safe driving range for a continuous-flow ventricular assist device (CFVAD). During the first stage of the present experiment, the characteristic curves relating pump speed and ICA were determined by varying preload (left atrial pressure: -6 to 30 mm Hg), afterload (total circuit resistance: 890 to 3,180 dyne x sec x cm(-5)), and contractility of the left ventricle (total circuit flow: 0.5 to 2.1 L/min). These data showed that an ICA value of 0.18 was always located between the t- and s-points. During the second stage of the experiment, we developed an automatic driving program to control pump speed by maintaining ICA at 0.18. This program was able to drive the CFVAD, without exhibiting regurgitant flow or sucking, under various driving conditions in the mock circulation. Pump speed stabilized within 1 minute after varying the drive conditions. This sensorless method of driving the CFVAD by using a target ICA proved feasible and effective for safe automatic control, within our mock circulation.

Control strategy for biventricular assistance with mixed-flow pumps

A left ventricular assist device (LVAD) is an effective method to rescue severe heart failure. Although some require a biventricular assist, the control method for the biventricular assist device (BVAD) with a rotary pump is rarely shown. The objective of this study was to investigate the strategy for controlling BVAD with rotary pumps by in vivo studies. Using 5 piglets, we set a BVAD through a left thoracotomy and made global ischemia for 30 min by clamping the base of the ascending aorta. After unclamping, the analysis of pumping performance acted for 6 h reperfusion. We set the target flow of the LVAD and set the right ventricular assist device (RVAD) speed limit as less than when the atrial collapse occurs. To detect the ventricular collapse without any specific sensor, we calculated the index of current amplitude from motor current waveform and simultaneous mean current value. In all cases, over 6 h of observation was performed, and the RVAD was weaned almost automatically.

Sensorless controlling method for a continuous flow left ventricular assist device

We originated a novel control strategy for a continuous flow left ventricular assist device (LVAD). We examined our method by acute animal experiments to change the left ventricular (LV) contractility or LV end-diastolic pressure (LVEDP). To estimate the pump pulsatility without any specific sensor, we calculated the index of current amplitude (ICA) from motor current waveform. The ICA had a peak point (t-i point) that corresponded closely with the turning point from partial to total assistance, and a trough (s-i point) that corresponded with the beginning point of ventricular collapse. The pump flow at the t-i point (Qt-i) had no component of flow regurgitation. In the evaluation of the effects of preload LVEDP, afterload (mAoP), and contractility (max LV dp/dt), we found that preload was the only parameter that significantly influenced Qt-i. We concluded that our method could well control continuous flow LVAD by preventing reversed flow and ventricular collapse.