MONITORING OF A CARDIAC ASSIST DEVICE

Abstract

A control system for a cardiac assist device includes a sensor implantable in the body at the heart or at an implanted pump of the cardiac assist device, the sensor being for detecting motion of the pump within the body and hence being for monitoring movement of the pump, where the control system is arranged to, in use: receive signals from the sensor, the signals providing information on the movement of the pump; and to process the signals to monitor the pump speed and/or to identify pump malfunction and/or cardiac assist treatment complications.

Claims

1-21. (canceled)

22. A method for cardiac assistance of a patient, the method comprising using a control system with a cardiac assist device, the cardiac assist device comprising an implanted pump and a graft, and the control system comprising one or more sensors implanted in the body at the heart or at the implanted pump or the graft of the cardiac assist device, wherein at least one of the one or more sensors is a motion sensor, the method including the steps of: detecting, with the motion sensor, a motion of the pump within the body and thereby monitoring the motion of the pump relative to the body; receiving signals by the control system from the motion sensor, the signals providing information on the motion of the pump relative to the body; determining information about pump function, including vibrations caused by an impeller and blood flow patterns through the pump, using the signals from the sensor; processing the signals received from the motion sensor in order to determine an acceleration signal associated with the cardiac assist device, in order to identify a pump malfunction or to identify complications to cardiac assist treatment; identifying a potential pump malfunction or complication based on the acceleration signal; determining, based on an analysis of the acceleration signal, when there is a failure or malfunction of the pump and the type of failure; and providing, by the control system, an alert to a user or operator when the potential pump malfunction or complication is identified.

23. A method as claimed in claim 22, wherein the potential pump malfunction or complication is due to at least one of clotting, embolism, tube dislodgement, suction, acute Atrial Septal Defect (ASD), and Ventricular Septal Defect (VSD).

24. A method as claimed in claim 22, wherein the method comprises measuring the corrective effect of changes in a flow rate of the cardiac assist device and determining that there is a problem in addition to an underlying heart defect when corrective adjustments to the flow rate do not result in an expected improvement in heart function.

25. A method as claimed in claim 22, wherein the method comprises monitoring of the cardiac pump and/or cardiac function and/or guidance of medical treatment in the acute phase or the follow-up phase.

26. A method as claimed in claim 22, wherein the step of processing the signals includes a raw signal analysis, frequency analysis, pattern recognition analysis, data streaming analysis and/or acoustic analysis.

27. A method as claimed in claim 26, wherein the method comprises identifying, based on the signal analysis, pump parameters that can indicate the potential pump malfunction or complication.

28. A method as claimed in claim 26, wherein the method comprises identifying, based on the signal analysis, changes in blood flow patterns thorough the cardiac assist device caused by inflow and outflow problems and changes in blood viscosity and embolies caused by air, blood or tissue fragments.

29. A method as claimed in claim 27, wherein the method comprises identifying, by the control system, the potential pump malfunction or complication when the pump parameter(s) derived from the motion sensor signals differs from an expected or normal pump parameter by more than a given amount.

30. A method as claimed in claim 26, wherein the method comprises identifying the potential pump malfunction or complication due to the presence or absence of one or more predetermined frequencies in the sensor signal.

31. A method as claimed in claim 22, wherein the method comprises comparing the sensor signal and/or the results of a frequency analysis thereof to historical motion sensor data to identify when a failure or malfunction of the pump and to identify the type of failure.

32. A method as claimed in claim 22, wherein the method comprises optimising the performance of the cardiac assist device by increasing a flow rate if the flow rate is determined to be inadequate or decreasing the flow rate if the flow rate is determined to be excessive.

33. A method as claimed in claim 32, wherein the cardiac assist device comprises a pulsatile pumping characteristic and the optimising comprises adjusting a pulsing speed of the cardiac assist device.

34. A method as claimed in claim 22, wherein the method comprises monitoring for a progressive reduction in an afterload of a ventricle of the heart and/or monitoring for an acute increase in the afterload of the ventricle; determining that there is a potentially excessive flow rate when one or both of these occurs; and adjusting the flow rate downwards if it is determined that there is a potentially excessive flow rate.

35. A method as claimed in claim 22, wherein the method comprises monitoring for a progressive reduction in contractility and/or monitoring for heart motion indicating a reduced systolic contraction and an increased post systolic contraction; determining that there is a potentially inadequate flow rate when one or both of these occurs; and adjusting the flow rate upwards if it is determined that there is a potentially inadequate flow rate.

36. A method as claimed in claim 22, wherein the method includes identifying and monitoring diastolic dysfunction by measuring, using the one or more sensors, early and atrial inflow patterns and relations.

37. A method as claimed in claim 22, wherein the method comprises processing signals from the one or more sensors to determine a level of physical activity of the patient and adjusting a flow rate of the cardiac assist device in response to changes in physical activity.

38. A method as claimed in claim 22, wherein the method comprises wirelessly transmitting data within the body and/or to outside the body.

39. A control system for a cardiac assist device, the system including: one or more sensors implantable in a body of a patient at a heart or an implanted pump or graft of the cardiac assist device, wherein at least one of the one or more sensors is a motion sensor, wherein the motion sensor is operative to detect a motion of the pump within the body and thereby to detect the motion of the pump relative to the body; the control system comprising a processor operative to, in use: receive signals from the sensor, the signals providing information on the motion of the pump relative to the body; determine information about pump function, including vibrations caused by an impeller and blood flow patterns through the pump, using the signals from the sensor; determine an acceleration signal associated with the cardiac assist device by processing the signals received from the motion sensor, in order to identify a pump malfunction or to identify complications to cardiac assist treatment; identify a potential pump malfunction or complication based on the acceleration signal; and determine, based on an analysis of the acceleration signal, when there is a failure or malfunction of the pump and the type of failure, wherein the control system is configured to provide an alert to a user or operator when the potential pump malfunction or complication is identified.

40. A control system for a cardiac assist device, the system including: one or more sensors implantable in a body of a patient at a heart or an implanted pump or graft of the cardiac assist device, wherein at least one of the one or more sensors is a motion sensor, wherein the motion sensor is operative to detect a motion of the pump within the body and thereby to detect the motion of the pump relative to the body; the control system configured to operate in accordance with the method of claim 22.

41. A cardiac assist device comprising the control system of claim 39.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Certain preferred embodiments of the invention will now be described by way of example only and with reference to the accompanying drawings, in which:

[0047] FIG. 1 shows an example of the use of implanted sensors in conjunction with a LVAD device for the human heart;

[0048] FIG. 2 shows a similar example where the implanted sensors are on the heart and at the pump of the LVAD device;

[0049] FIG. 3 is a plot of accelerometer readings obtained during LVAD treatment of a patient;

[0050] FIGS. 4a and 4b show an example of signal analysis using frequency distribution in accelerometer signals, in this case during adrenalin infusion;

[0051] FIG. 5 is a diagram with a cross-section of the heart showing possible locations for motion sensors at the heart;

[0052] FIG. 6 shows possible locations for motion sensors at the pump and/or graft;

[0053] FIG. 7 illustrates an experimental set-up use for in vitro testing of the proposed sensor system with an accelerometer at a VAD in a simulated circulation system;

[0054] FIG. 8 is a fast Fourier transform of data from the accelerometer of FIG. 7 showing detection of VAD RPM;

[0055] FIG. 9 is a fast Fourier transform showing detection of a simulated thromboembolism using the accelerometer;

[0056] FIG. 10 illustrates changes in the acceleration signal when afterload is increased;

[0057] FIG. 11 shows changes in the acceleration signal when preload is decreased;

[0058] FIG. 12 shows the effects of injection of thrombus on the acceleration signal;

[0059] FIG. 13 is a close up view of changes during injection of a solid thrombus;

[0060] FIG. 14 is similar to FIG. 12 and shows the effects of simulated air embolisms;

[0061] FIG. 15 is a frequency spectrum for acceleration signals from a sensor implanted in a pig for in vivo testing involving infusion of viscous material into the left ventricle and LVAD; and

[0062] FIG. 16 shows the acceleration signals for the same in vivo testing.

DETAILED DESCRIPTION OF THE DRAWINGS

[0063] The LVAD device of FIGS. 1 and 2 is similar to conventional devices as regards its basic function in pumping blood to assist cardiac function. The LVAD comprises a controller 2, batteries 4 and a pump 6. The batteries 4 are held on the patient's body along with the controller by a harness. The controller 2 is linked to the batteries 4 by wires and control wires 8 link the controller 2 to the pump 6. The pump 6 is implanted inside the body and is connected between the left ventricle and aorta in order to provide ventricular assistance to the heart. The control wires 8 connect to the pump within the body and to the controller 2 outside of the body. They supply power and control signals from the controller 2 to the pump 6.

[0064] The example arrangement of the FIG. 1 embodiment further includes motion sensors 10, 12. A first motion sensor 10 is connected to the wall of the right ventricle, and a second motion sensor 12 is connected to the wall of the left ventricle. The control of the pump 2 by the controller 2 involves the use of data from the motion sensors 10, 12. The motion sensors 10, 12 can be any suitable sensor, such as 3-axis accelerometers, miniaturized ultrasound sensors, inertia based sensors, electromechanical position sensors and/or gyrosensors, and may for example be sensors of a type similar to those disclosed in EP 1458290.

[0065] The motion sensors 10, 12 provide signals for functional assessment of the right and left ventricle to guide therapy management (cardiac assist device settings and medical therapy). Processing of these signals is integrated into the control system of the controller 2 to thereby enable backward supervision (closed loop feedback control) to optimize the treatment of heart failure and the operation of the cardiac assist device. The control system may for example use position, motion and/or acceleration data from the sensors to determine heart movement and then monitor for changes in afterload, contractility, heart rate and other parameters of heart movement in order to identify heart dysfunction indicative of potential sub-optimal operation of the cardiac assist device. Various examples of this are set out above. The control system can also take account of other parameters including those measured at the pump such as blood pressure and so on.

[0066] The possibility to provide continuous hemodynamic monitoring (contractility and pumping capacity) and hemodynamic feedback to cardiac devices to optimize pump settings, guide the effects of medical therapy, effects of physical activity (increased demand) and to detect complications (ventricular failure, device malfunction etc.) during use of cardiac devices. Known cardiac devices do not have direct feedback systems for evaluating cardiac performance.

[0067] Motion sensor systems as described herein, for example attached to the walls of right and left ventricle, will deliver highly clinical relevant signals on myocardial contractility and ventricular performance. The sensors have been tested in various models aimed to induce both global and regional ventricular dysfunction by inducing changes in contractility (ischemia, betablocade, septic and hypotermic cardiomyopathy), preload (volume unloading and pharmacological intervention) and afterload (outflow obstruction and pharmacological intervention). The sensors are capable of detecting heart failure earlier than routine hemodynamic monitoring, and with high sensitivity. The sensors provide information about heart function very similar to echocardiography, but have an obvious advantage as continuous monitoring is possible. Signals from such sensors may also be used for guidance of treatment with implantable cardiac devices. Automated signal analysis has proved feasible with the described sensor systems and hence is implemented in the proposed control system.

[0068] The second sensor 12 in the above embodiment could be used to detect signals reflecting operation of the pump 6, in particular the speed of the pump. FIG. 2 shows an embodiment focused on monitoring of the pump 6 and it should be understood that the second sensor 12 of the embodiment above could be utilized for pump monitoring in the same way as the equivalent second sensor 12 in FIG. 2. As will be seen, the embodiment of Figure is broadly similar to that of FIG. 1 except that the first sensor 10 on the right ventricle is not present and further motion sensor, which is a pump sensor 14, is located at the pump 6.

[0069] The motion sensor 10 at the left ventricle and/or the pump sensor 14 can be used to monitor pump speed and also to detect pump malfunction as a consequence of problems such as thrombus/clotting, embolism and impeller or tube malfunction.

[0070] Pump failure is life threatening, and so is ischemic stroke due to clotting/thrombus formations and embolism. In case of thrombus formation in the LVAD it may often be necessary to change the entire pump. This is both hazardous and costly. The cost for a LVAD pump is approximately USD 120,000. However, the cost related to the operation and intensive care stay far exceeds this amount. Pump exchange is associated with a mortality of 25%. To reduce the risk of thrombus formation in the pump, the patients are anticoagulated and receive platelet inhibitors. However, too much anticoagulation infers the risk of life threatening bleedings related to both the device or to intracranial bleeding (hemorrhagic stroke). Thus, these patients are frequently monitored for level of anticoagulation (INR 2-3), hemolysis due to destruction of red blood cells by the pump, and with echocardiographic assessment of possible thrombus in left ventricle. Thus the patient frequently needs to be in contact with the hospital.

[0071] In known cardiac assist devices there is a continuous analysis of the power needed to drive the impeller within the pump. The rotation speed of the impeller is related to the pump speed (RPM) settings on the controller. If large embolis or clotting occur within the pump then energy or power needed to maintain RPM is increased. Changes in power are logged in the controller. However, studies have shown that this may not always detect device malfunction (see, for example, PMA No.P110047, and INTERMACS registry).

[0072] The cardiac assist device of FIG. 2 uses accelerometers 12, 14 to monitor the pump 6 itself. An accelerometer 12, 14 placed on the pump 6 and/or on the ventricular wall close to the pump 6 can be used to monitor complications with cardiac assist device treatment for end stage heart failure. This has been tested with three patients, where right and left ventricular function was monitored during implantation of a left ventricular assist device (LVAD). From these patients during LVAD treatment it was possible to extract information on pump mechanics, such as rotation speed (RPM) by analyzing the accelerometer signals from the accelerometer 2-4 cm from the device, which corresponds to the motion sensor 12 on the left ventricle.

[0073] FIG. 3 shows the monitored accelerometer signals for various usages of a heart and lung machine (HLM) and LVAD. The accelerometer is a three axis device and in FIG. 3 plot A: acceleration signal in the longitudinal direction of the heart, plot B: the circumferential direction, and plot C: the radial direction. In the acceleration signals there are oscillations that correspond to LVAD RPM. The frequency distribution of the acceleration signals can be used to detect LVAD pump failure (change in higher frequencies will indicate failure).

[0074] There are distinct spikes that correspond to the RPM settings on LVAD. This means that the LVAD pump caused the left ventricle to vibrate in the same frequency as the RPM settings. An accelerometer is an ideal sensor for monitoring such vibrations. From previous studies it has been shown that accelerometers can be used for monitoring heart sounds due to heart wall vibrations caused by heart valve closure. By analyzing the frequency distribution of the vibration signals it is possible to gain more information than just looking on the raw acceleration signal. This has been done to detect regional myocardial ischemia during coronary artery occlusion, but also to detect changes in global heart function (as illustrated in FIGS. 4a and 4 b, which show a signal analysis for an accelerometer signal obtained during adrenaline infusion). Similarly, there will be a change in the frequency distribution of the vibration signals detected by an accelerometer placed on the ventricle if VAD malfunction/failure occurs. An accelerometer can also detect similar changes if integrated as part of the implanted VAD.

[0075] By careful analysis of accelerometer signals it is possible to determine when there is a failure or malfunction of the pump and also to determine the type of failure. This can be done, for example, by identifying certain frequencies of motion that are associated with certain failure modes and/or by comparison of the measured signals with historical accelerometer data. The historical data can include accelerometer signals for pumps without failure and also accelerometer signals for pumps that malfunctioned with a known failure mechanism. It is expected that similar types of failures will produce similar irregularities in the accelerometer signals and therefore comparison with past known failures will allow future failures to be identified.

[0076] In the above embodiments the surgical implantation of the pump 6 and the internal part of the control wires 8 can be carried out by conventional surgical techniques. The implantation of the sensors 10, 12 can be done in conventional fashion.

[0077] Possible locations for the motion sensor(s) used for the invention are shown in FIGS. 5 and 6. FIG. 5 shows a cross-section of the heart through the left ventricle (LV) and right ventricle (RV). Three general locations are shown for a motion sensor at the left ventricle, where A is an epicardial/subepicardial sensor, B is a myocardial sensor and C is an endocardial/subendocardial sensor. FIG. 6 shows a pump 6 and graft 16 and indicates three general locations for a sensor at the pump 6 or graft 16, where D is a sensor on the pump 6, E is a sensor within the pump 6 and F is a sensor on the graft 16.

[0078] It will readily be understood that although the above discussion and the Figures relate to the implanted sensors in the context of an LVAD device the sensors and control system could equally well be applied to aid the operation of an RVAD or BiVAD device, or any similar cardiac assist device for human or animal cardiac assistance and/or monitoring.

[0079] In addition, although the above example embodiments utilize two motion sensors there could alternatively be just one sensor or more than two sensors depending on the level of information required, the condition of the patient, and the cardiac assist device that is being used. For example, the system could include several of the first motion sensor 10 at the right ventricle, the second motion sensor 12 at the left ventricle, the pump sensor 14 at the pump 6 and/or a sensor at the graft 16, or just one of those sensors.

[0080] The proposed system has been tested in vitro and in vivo. The experimental set up used for in vitro testing of the proposed sensor system is shown in FIG. 7. This used a simulated circulatory system with flow direction as indicated by the arrow D. A reservoir 20 supplied fluid to a VAD 24 of conventional type. This was equipped with an accelerometer 28 for detecting motion of the VAD 24. The experiment used an injection port 28 for injection of thrombus/emboli into the system. A sample port 34 was also present, for sampling of the fluid. A pressure regulator 32 between the VAD and reservoir was used for preload adjustment and the preload produced by this regulator was measured using a pressure sensor 30 between the VAD 24 and the regulator 32. Also present is a spectrum analyzer 26 and Doppler sensor 36 for monitoring the circulation in the system.

[0081] Various tests were carried out to demonstrate the capabilities of the accelerometer and the results are shown in FIGS. 8 to 13.

[0082] FIG. 8 is a fast Fourier transform of data from the accelerometer of FIG. 7 showing detection of VAD RPM. The VAD RPM was changed between 1800, 2000 and 2200 as shown, and the accelerometer is able to easily detect this. As discussed above in relation to FIG. 3 the use of a motion sensor like an accelerometer is ideal for detecting the VAD RPM.

[0083] The frequency data can also be used to detect a simulated thromboembolism as shown in FIG. 9. A third harmonic in the acceleration signal is indicative of a possible thromboembolism.

[0084] FIGS. 10 and 11 illustrates changes in the acceleration signal when the afterload or preload changes. In FIG. 10 the afterload is increased with time and this results in an increase in the amplitude of the acceleration signal. In FIG. 11 the preload is decreased with time and again this results in an increase in amplitude of the acceleration signal. The motion sensor can hence be used to detect changes in preload or afterload. As discussed previously this can be important in detecting potential problems.

[0085] In one test various thrombi were injected into the system. FIG. 12 shows the effects of the injection of thrombus on the acceleration signal. The arrows indicate the approximate time that the thrombus was injected. From left to right, the thrombi were: 0.1 ml soft thrombus, solid thrombus, 0.25 ml soft thrombus and 0.5 ml soft thrombus. As might be expected, the larger the volume of the thrombus the greater the effect. The solid thrombus has a larger effect than the soft thrombi. FIG. 13 shows the effects of the solid thrombus in close up view.

[0086] Air embolisms were simulated in a similar fashion and FIG. 14 shows the effects of the simulated air embolisms on the acceleration signal. The arrows indicate the approximate timing for the injection of air. The volume of air injected, for the arrows from left to right, was 0.1 ml, 0.25 ml, 0.5 ml, 1 ml and 2 ml. It will be seen that it is possible to differentiate between air emboli, solid thrombi and soft thrombi.

[0087] The in vivo testing used a sensor implanted in a pig. The pig was equipped with a LVAD with motion sensor at the LVAD for monitoring motion of the VAD. FIGS. 15 and 16 show the data from the accelerometer when viscous material was infused into the pig's left ventricle and the LVAD. The effects of this infusion on the acceleration signal can be seen and it will be understood that they are similar to the in vitro testing.