Real-time detection of r-waves and methods therefor

Abstract

To provide a drive unit system for operating a cardiac assist device.SOLUTION: The drive unit system may include an R-wave detection module that makes a speculative detection of an R-wave associated with a received electrocardiogram (ECG) signal of a patient's heart and a conservative detection of the R-wave. The drive unit system includes a drive unit, which may be coupled to the cardiac assist device and may operate the cardiac assist device based on the speculative detection and the conservative detection. The cardiac assist device may be a counterpulsation device, and may be an intra-aortic balloon pump.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates generally to medical devices, and more particularly to systems and methods for the operation of cardiac assist devices (CADs) based on physiological measurements. Approximately 5.7 million adults in the United States have heart failure, according to the Centers for Disease Control and Prevention website. Approximately 100,000 people in the country are diagnosed with advanced heart failure each year, and require some mechanical support, for example, via CAD. In some cases, CADs may undergo ventricular assist, counterpulsation therapy, or copulsation therapy. In general terms, counterpulsation is a ventricular assist technique that assists circulation and reduces the work of the heart by increasing diastolic aortic pressure (i.e., diastolic aortic pressure) relative to unassisted (i.e., no counterpulsation) diastolic aortic pressure, and decreasing systolic aortic pressure (i.e., systolic aortic pressure) relative to unassisted (i.e., no counterpulsation) systolic aortic pressure. For its part, copulsation is a ventricular assist technique that increases systolic aortic blood pressure compared to unassisted systolic aortic blood pressure. [Background technology]

[0002] There are a number of CADs that can deliver different levels of assistance at different levels of invasiveness to treat cardiac dysfunction. Exemplary CADs include counterpulsation devices (i.e., devices that are specific to the counterpulsation modality and are implanted in the vasculature) and ventricular assist devices (VADs), which are at least partially implanted in the ventricles.

[0003] Counterpulsation devices include intra-aortic balloon pumps, patch devices, cuff devices, and other pump devices anastomosed to the great vessels. An intra-aortic balloon pump (IABP) may be positioned inside the aorta, typically in the proximal descending aorta. The balloon pump (typically 25-50 milliliters in volume) may be inflated and deflated in synchrony with the contraction of the left ventricle, thereby performing counterpulsation therapy. The IABP may be inflated during diastole, thereby increasing the diastolic aortic pressure relative to the unassisted diastolic aortic pressure, driving blood in the ascending aorta and aortic arch into the coronary arteries to supply oxygen to the myocardium. The IABP may be deflated during systole as the left ventricle contracts, thereby decreasing the systolic aortic pressure relative to the unassisted systolic aortic pressure, reducing the afterload (i.e., the effort required by the heart to push blood into the aorta).

[0004] Other counterpulsation devices operate similarly. For example, a patch device is generally implanted on the descending thoracic aorta. A longitudinal incision may be made and the patch device may be sutured to the outer surface of the descending thoracic aorta. The patch device may include an elongated inflatable valveless polyurethane patch powered by an external compressor. The patch may be inflated and deflated simultaneously with the contraction of the left ventricle, thereby performing counterpulsation therapy in a manner similar to that described above with respect to the balloon pump. An example of a patch device is the Kantrowitz CardioVad.

[0005] Similarly, a cuff device may be wrapped around the ascending aorta and may have a membrane (e.g., a balloon) positioned against the outer wall of the vessel. Positive and negative pressure on the membrane promotes counterpulsation. An example of a cuff device is the Sunshine C-Pulse® cardiac assist system.

[0006] The blood pump device may be implanted in a pocket under the pectoral muscle in the anterior chest and attached to a vascular graft anastomosed to the subclavian artery. The blood pump device may include a reservoir that fills with blood during systole (thereby lowering the systolic aortic blood pressure compared to the unassisted systolic aortic blood pressure) and empties during diastole (thereby raising the diastolic aortic blood pressure compared to the unassisted diastolic aortic blood pressure). An example of such a blood pump device is the Symphony Counter Pulsation Device (CPD). Other blood pump devices exist, including so-called para-aortic blood pumps for life support during cardiac surgery and disposable blood pumps (e.g. PulseCath iVAC or PUCA pumps). Both operate similarly to the Symphony CPD. The para-aortic blood pump is anastomosed to the descending aorta, and the PulseCath iVAC or PUCA pump consists of a catheter containing a bi-directional valve connected to a pump that operates to draw blood from the left ventricle during systole (lowering systolic aortic pressure relative to unassisted systolic aortic pressure) and push blood back into the aorta during diastole (increasing diastolic aortic pressure relative to unassisted diastolic aortic pressure).

[0007] In particular, some counterpulsation devices, when implanted in the ventricle, may be configured or reconfigured to function as copulsation devices. As noted above, in copulsation, the device operates to increase aortic blood pressure (relative to unassisted aortic blood pressure) during systole.

[0008] A VAD is a CAD that assists the cardiac circulation by pumping blood. Generally, blood is drawn from the left ventricle into a pump, then into the aorta, up the body, and / or out of the right ventricle, into a pump, then into the pulmonary artery, and up the lungs to pick up oxygen. The VAD pump can be external or paracortical, or can be implanted (e.g., connected to the left and / or right ventricle via open-heart surgery, or placed in place using a catheter). VADs are generally characterized as pulsatile or continuous flow. Continuous flow VADs are sometimes called CF-VADs, and are usually driven by an impeller or rotor. Some VADs can operate in a counterpulsation or copulsation modality, where the VAD is synchronized to the cardiac cycle. If the VAD is not synchronized to the cardiac cycle, the VAD can be said to operate asynchronously.

[0009] Current VADs include the CorWave left ventricular assist device (LVAD), Abbot HeartMate 3 LVAD, Abbot HeartMate 2 LVAD, Medtronic HeartWare HVAD, Abiomed Impella® pump, and EVAHEART®2 LVAD. The CorWave LVAD uses a wave membrane, while other identified VADs are CF-VADs. Some CF-VADs use a valveless axial pump with a variable magnetic field (i.e., so-called axial flow pumps) to rotate an impeller that generates a continuous outflow that flows parallel to the axis of rotation. Other VADs use a centrifugal pump with magnetic or hydrodynamic levitation of an impeller with non-contact bearings and an outflow directed perpendicular to the axis of rotation (i.e., so-called centrifugal pumps). The Abiomed Impella pump is an example of a VAD with a catheter-based micro-axial rotary pump. It is inserted into the patient's femoral artery and then guided into the left ventricle. The impeller draws blood from the left ventricle into the ascending aorta. Taking the Abiomed Impella as an example, this VAD can be deployed with copulsation by driving the rotor at a higher rpm during systole and a lower rpm during diastole. As a result, the systolic aortic blood pressure is increased compared to the unassisted systolic aortic blood pressure. The Abiomed Impella may also be deployed with counterpulsation by driving the rotor at a higher rpm during diastole and a lower rpm during systole. As with counterpulsation devices, this results in an increase in the diastolic aortic blood pressure (compared to the unassisted diastolic aortic blood pressure) and a decrease in the systolic aortic blood pressure (compared to the unassisted systolic aortic blood pressure). Summary of the Invention

[0010] CADs operating in copulsation or counterpulsation modalities require timing that is synchronized with the native heart. There is a need to provide timing algorithms and devices that can provide such synchronization. [Brief description of the drawings]

[0011] The accompanying drawings illustrate several embodiments and, together with the description, serve to explain the disclosed principles. However, those skilled in the art will understand that the embodiments may be practiced without all of the specific details of the illustrated examples. Similarly, those skilled in the art will understand that the present technology may include well-known structures or functions that are not specifically illustrated in order to avoid unnecessarily obscuring the relevant description of the various embodiments.

[0012] [Figure 1] 1 illustrates an exemplary intravascular circulatory support system implanted within a patient's vasculature and configured in accordance with an embodiment of the present technology. [Diagram 2] 1 illustrates an exemplary QRS complex in an electrocardiogram (ECG) signal in accordance with some embodiments of the present technology. [Diagram 3] 1 shows an exemplary Wiggers diagram in accordance with some embodiments of the present technology. [Figure 4] 1 shows an exemplary schematic diagram of a drive unit system according to some embodiments of the present technology. [Figure 5A] 1 shows an exemplary schematic diagram of a wavelet filter in accordance with some embodiments of the present technique; [Figure 5B] 1 illustrates an example mother Haar wavelet function in accordance with some embodiments of the present technique. [Figure 5C] 1 illustrates an example mother symlet4 wavelet function in accordance with some embodiments of the present technique. [Figure 5D] 5 illustrates an exemplary ECG signal received by the drive unit system of FIG. 4 in accordance with certain embodiments of the present technology. [Figure 5E] 5B shows a plot of an exemplary multiple discrete frequency signal obtained by decomposing the ECG signal of FIG. 5D using the Haar mother wavelet function of FIG. 5B in accordance with certain embodiments of the present technology. [Figure 5F] 1 shows an example filtered ECG signal and its corresponding plot in accordance with some embodiments of the present technology; [Figure 6A] 1 shows a plot of an exemplary ECG signal including a PVC in accordance with some embodiments of the present technology. [Figure 6B] 6B shows a plot of an example filtered ECG signal related to the ECG signal of FIG. 6A in accordance with some embodiments of the present technology. [Figure 7] 1 shows a plot of a portion of an example filtered ECG signal crossing multiple R-wave detection thresholds in accordance with certain embodiments of the present technology. [Figure 8A] 1 shows a plot of an exemplary ECG signal in accordance with some embodiments of the present technology. [Figure 8B] 8B shows a plot of an exemplary filtered ECG signal related to the ECG signal of FIG. 8A, the exemplary plot further depicting R-wave detection thresholds and crossings in accordance with certain embodiments of the present technology. [Figure 9] An example filtered ECG signal in FIG. 6B is shown along with an example windowing technique capable of detecting a filtered ECG maximum value within each window that may be used to determine an example predicted filtered ECG maximum value in accordance with some embodiments of the present technology. [Figure 10A] 5 illustrates an exemplary noisy ECG signal received by the drive unit system of FIG. 4 in accordance with some embodiments of the present technology. [Figure 10B] 10B shows an example filtered ECG signal corresponding to the noisy ECG signal of FIG. 10A in accordance with some embodiments of the present technology. [Figure 10C] 10B shows an expanded segment of the example filtered ECG signal of FIG. 10B and an example windowing technique for ascertaining noise bias in accordance with some embodiments of the present technology. [Figure 10D]10C is a plot of an expanded segment of the exemplary filtered ECG signal of FIG. 10B along with an exemplary windowing technique capable of detecting a filtered ECG maximum within each window that may be used to determine an exemplary predictive filtered ECG maximum in accordance with certain embodiments of the present technology. [Figure 10E] 10C is a plot of an expanded segment of the example filtered ECG signal of FIG. 10B, the plot showing an ambient noise value, a predictive filtered ECG maximum, and a single a priori and conservative R-wave detection threshold, in accordance with some embodiments of the present invention. [Figure 11] 5 illustrates an exemplary method for using the drive unit system of FIG. 4 in accordance with some embodiments of the present technology. [Figure 12] 5 shows a state diagram of the drive unit system of FIG. 4 in accordance with some embodiments of the present technique. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] CADs operating in copulsation or counterpulsation modalities are able to process electrocardiogram (ECG) signals for efficient and optimal operation. One of the most important parts of ECG signal processing and CAD operation is the interpretation of the QRS complex and retrieving its characteristics. The QRS complex is the name for the combination of three graphical deflections seen in a typical ECG. In adults, the QRS complex usually lasts 0.06-0.10 seconds, in children, and during physical activity, the QRS complex may be shorter. The three deflections are the Q wave, the R wave, and the S wave. These waves occur in rapid succession, do not all appear in all leads, reflect a single event, and are therefore usually considered together. In a typical ECG signal, the Q wave is any downward deflection immediately following the P wave, the R wave follows as an upward deflection, and the S wave is any downward deflection after the Q wave. The T wave follows the S wave, and in some cases, an additional U wave follows the T wave. In some ECG signals, the Q wave is an upward deflection, the R wave follows as a downward deflection, and the S wave is an upward deflection.

[0014] The R wave is one of the most important sections of this group and has a key role in both diagnosing cardiac rhythm abnormalities and in determining heart rate variability (HRV). For counterpulsation CAD (e.g., counterpulsation devices and CADs operating in counterpulsation modality) and copulsation CAD (e.g., CADs operating in copulsation modality), the R wave may be used as an important marker to determine / detect systole and / or diastole. For IABP, the R wave may precede the ventricular contraction by 80-150 ms and is an important marker to initiate the rapid deflation of the balloon so that it does not occlude in the aorta when the ventricle starts to eject blood. Traditional systems for detecting the QRS complex include differential methods, digital filters, neural networks, filter banks, hidden Markov models, genetic algorithms, and maximum a posteriori (MAP) estimators. These methods are very sensitive to noise and generally cause errors in detecting the correct timing of the R wave in the ECG signal. As a result, conventional systems and methods for intra-aortic balloon pumps suffer from the disadvantage of being unable to predict the inflation and / or deflation times of the balloon pump. Similar problems exist for other CADs, including those that operate in copulsation and counterpulsation modalities.

[0015] Given the shortcomings of conventional approaches for accurately detecting the occurrence and timing of R-waves, there is a practical need for techniques for R-wave detection that are more accurate, reliable, and efficient, and that improve the safety and effectiveness of CADs, including CADs operating in copulsation or counterpulsation modalities.

[0016] In the following, the techniques described herein are described in the context of IABP and counterpulsation. The disclosed techniques may be particularly useful for counterpulsation, but may have application to a variety of other techniques and devices, including other CADs (e.g., CADs for operation in copulsation modalities, CADs for operation in counterpulsation modalities), pacemakers, cardiac monitors, defibrillators, heart rate monitors, smart watches, athletic accessories, and the like. References to counterpulsation include references to copulsation and other therapeutic modalities for the CADs and devices identified in this paragraph.

[0017] Some embodiments are discussed in more detail below with reference to the drawings. In addition to those described herein, other embodiments are within the scope of the present technology. Moreover, those skilled in the art will understand that the embodiments of the present technology may have configurations, components, and / or procedures in addition to those shown or described herein, and that these and other embodiments may be implemented without some of the configurations, components, and / or procedures shown or described herein without departing from the present technology. Throughout this description, reference to "one embodiment," "an embodiment," "one or more embodiments," "an nth embodiment," or "some embodiments" means that a particular feature, supporting structure, or characteristic described in connection with an embodiment is included in at least one embodiment of the present disclosure. Thus, use of such terms does not necessarily refer to the same embodiment. For example, it is expressly contemplated that features described herein may be combined in any suitable manner in one or more embodiments.

[0018] (Intravascular Circulation Support System Components) FIG. 1 illustrates an exemplary intravascular circulatory support system 100 implanted within a patient's vasculature. The system 100 may include an intra-aortic balloon pump 110 that is implantable in the aorta of a patient suffering from heart failure. The system 100 may further include a first drive line 120 (also referred to as an "internal drive line"), an arterial interface device or stopper device 130, a second drive line 140, a skin or patient interface device 190, a tube 172, a drive unit 150, and a sensor 160. Once implanted, the system 100 may provide counterpulsation therapy to a patient suffering from heart failure. In some embodiments, the intravascular circulatory support system 100 may have certain features generally similar to those described in U.S. Patent Application Serial No. 16 / 876,110, the disclosure of which is incorporated herein by reference in its entirety.

[0019] The balloon pump 110 may be an expandable member, a balloon, or other element that may change size and / or shape in response to being filled with a gas or liquid. For example, in some embodiments, the balloon pump 110 is a balloon constructed of a biocompatible, non-thrombogenic elastomeric material (e.g., Biospan®-S). The balloon pump 110 may also be made of other suitable materials. The balloon pump 110 may transition between at least a first state that is generally deflated and a second state that is generally inflated. The balloon pump 110 may have a first volume when in the first (e.g., deflated) state and a second volume when in the second (e.g., inflated) state that is greater than the first volume. Thus, the balloon pump 110 may provide counterpulsation therapy by repeatedly transitioning between the first state and the second state. To transition the balloon pump 110 between the first and second states, the drive unit 150 may direct a fluid (gas or liquid, e.g., air) to an interior volume of the balloon pump 110 via the first drive line 120, the second drive line 140, and the tube 172. The balloon pump 110 may be sized and / or shaped to reduce and / or prevent blockage of arteries branching from the aorta, such as the renal arteries. In some embodiments, the balloon pump 110 includes an expandable end effector other than or in addition to a balloon. In some embodiments, the balloon pump 110 may have certain features generally similar to those described in U.S. Pat. Nos. 8,066,628, 8,323,174, and 8,326,421 and U.S. Patent Application Nos. 17 / 944,130, 17 / 944,127, and 17 / 944,125, the disclosures of which are incorporated herein by reference in their entireties.

[0020] The first drive line 120 may be an elongated structure having a lumen therethrough for delivering inflation gas to the balloon pump 110. The first drive line 120 may be at least partially disposed within the patient's vasculature (e.g., between the aorta and the axillary or subclavian arteries). After the system 100 is implanted, the first drive line 120 may have a first end (e.g., a distal end portion) coupled to the balloon pump 110 and disposed within the patient's vasculature (e.g., within the descending aorta) and a second end (e.g., a proximal end portion) coupled to the second drive line 140 and disposed outside the patient's vasculature. The first drive line 120 may exit the patient's vasculature, for example, at an arteriotomy in the axillary artery, subclavian artery, or another suitable vessel.

[0021] The second drive line 140 may also be an elongated structure having a lumen therethrough. The second drive line 140 may be disposed at least partially subcutaneously but external to the patient's vasculature. After the system 100 is implanted, the second drive line 140 may have a first end (e.g., a distal end portion) coupled to the first drive line 120 and a second end (e.g., a proximal end portion) coupled to the patient interface device 190. The first drive line 120 and the second drive line 140 may be made of the same or different materials and may have the same or different dimensions (e.g., length, outer diameter, inner diameter, etc.). In some embodiments, the second drive line 140 may include an absorption mechanism to reduce longitudinal strain that may be transferred to the first drive line 120 as the patient moves. For example, in some embodiments, the absorption feature is a curved region (e.g., an "S" shape or other serpentine curve) that can compress in response to a longitudinal force, thereby reducing and / or preventing the longitudinal force from being transmitted to the intravascular portion of the system 100 (e.g., the first drive line 120 and / or the balloon pump 110).

[0022] The first drive line 120 may be coupled to the second drive line 140 using any suitable technique. In some embodiments, the first drive line 120 is at least partially inserted into the second drive line 140 (or vice versa) and secured thereto using a compression ring. In some embodiments, the first drive line 120 is sewn, glued, stitched, or otherwise secured to the second drive line 140. In some embodiments, the first drive line 120 is attached to the second drive line 140 via a combination of the aforementioned techniques and / or via other suitable attachment techniques. Regardless of the connection mechanism, when the first drive line 120 is coupled to the second drive line 140, the lumen of the first drive line 120 may be fluidly connected to the lumen of the second drive line 140 such that gas flowing through the lumen of the second drive line 140 also flows through the lumen of the first drive line 120. Although described as first and second drive lines, in some embodiments, first drive line 120 and second drive line 140 may be a single integral component (i.e., a single drive line).

[0023] As described above, the first drive line 120 may be coupled to the balloon pump 110 to deliver inflation gas to the balloon pump 110. For example, a first end of the first drive line 120 may be connected to the balloon pump 110 such that a lumen extending through the first drive line 120 is in fluid communication with the interior of the balloon pump 110. Thus, gas flowing through the lumen of the first drive line 120 toward the balloon pump 110 may flow into the balloon pump 110 and transition the balloon pump 110 from a first state to a second state (e.g., inflating the balloon pump 110). The first drive line 120 may also receive gas from the balloon pump 110 when the balloon pump 110 transitions between the second state and the first state (e.g., when the balloon pump 110 is deflated).

[0024] The arterial interface device or stop device 130 may provide hemostasis over an extended period of time (e.g., greater than about 3 months, greater than about 6 months, greater than about 12 months, etc.) at the arteriotomy where the first drive line 120 exits the vasculature (e.g., at the subclavian or axillary artery). The stop device 130 may include multiple fixation elements that may be used to fix the stop device 130 in a desired orientation or position. In some embodiments, the stop device 130 may include certain features generally similar to those described in U.S. Pat. No. 7,892,162, the disclosure of which is incorporated herein by reference in its entirety. For example, the stopper device 130 may include a sewing ring (e.g., a polyester velour patch sutured to the artery to provide mechanical support), a graft (e.g., a polyester fabric sutured to the sewing ring and the artery to define a lumen and provide arterial access), and a stopper element (e.g., a silicone plug inserted into the lumen of the graft to provide hemostasis having a lumen to receive the first drive line 120).

[0025] The patient interface device 190 may be a percutaneous device that allows the external drive unit 150 to drive the operation of the implanted balloon pump 110. For example, in some embodiments, the patient interface device 190 provides a stable and / or secure exit site for the second drive line 140, allowing the second drive line 140 to be directly connected to the drive unit 150. In other embodiments, the second drive line 140 may be coupled to an interior facing portion of the patient interface device 190, and the drive unit 150 may be coupled to an exterior facing portion of the patient interface device 190 via a tube 172. The tube 172 may be a tube, hose, and / or other conduit having an elongated structure with a lumen extending therethrough. For example, the tube 172 may comprise a pneumatic tube with a lumen extending therethrough and one or more electrical conduits for serial communication between the patient interface device 190 and the drive unit 150. In such embodiments, the patient interface device 190 may direct gas received from the drive unit 150 (e.g., via tube 172) to the second drive line 140 for delivery to the expandable member 110. In some embodiments, the patient interface device 190 may have certain features generally similar to those described in U.S. Patent No. 10,137,230, the disclosure of which is incorporated herein by reference in its entirety.

[0026] The drive unit 150 can generate gas flow into and out of the balloon pump 110 via the first drive line 120, the second drive line 140, and the tube 172. For example, the drive unit 150 can generate a positive pressure to accelerate gas into the balloon pump 110 via the first drive line 120, the second drive line 140, and the tube 172, thereby inflating the balloon pump 110. The drive unit 150 can also induce a negative pressure to draw gas out of the balloon pump 110 via the first drive line 120, the second drive line 140, and the tube 172, thereby deflating the balloon pump 110. The drive unit 150 can induce gas flow into and out of the balloon pump 110 through a number of different mechanisms. For example, the drive unit 150 can utilize a bellows, a blower, a compressor, an accelerator, or other similar features to cause the gas flow into and out of the balloon pump 110. In some embodiments, the drive unit 150 can control the volume of air forced into the balloon pump 110 to avoid over-inflation of the balloon pump 110. For example, in embodiments utilizing a bellows to generate the air flow, the volume of the air flow generated by the bellows (e.g., the volume of the bellows) can be matched to the internal volume of the balloon pump 110. In some embodiments, the drive unit 150 uses ambient air (e.g., "room air") from the environment surrounding the drive unit 150 to drive the operation of the system 100. Without being bound by theory, the use of ambient air is expected to reduce the size, weight, and / or cost of the drive unit 150 relative to drive units that rely on an internal gas or fluid supply (e.g., a helium tank). For example, in some embodiments, the drive unit 150 can weigh approximately 2.2 kg or less. Thus, in some embodiments, the drive unit 150 can be portable / ambulatory. In other embodiments, the drive unit 150 can be operatively coupled to or otherwise include a gas or fluid supply, such as a helium tank (not shown).The drive line 140 or tube 172 may be disconnected near the patient interface device 190 when the system 100 is not in active use. In some embodiments, the drive unit 150 may have certain features generally similar to those described in U.S. patent application Ser. No. 17 / 878,632, the disclosure of which is incorporated herein by reference in its entirety.

[0027] The sensor 160 may sense one or more physiological parameters associated with the patient's natural heart rhythm to synchronize operation of the system 100 with the cardiac cycle. In particular, the one or more sensed physiological parameters may be used to automatically synchronize operation of the drive unit 150 with the patient's natural heartbeat to ensure that the balloon pump 110 is inflated and deflated at the appropriate times during the cardiac cycle. In some embodiments, the sensor 160 may sense one or more physiological parameters in real time. As shown in FIG. 1, the sensor 160 may be coupled to the patient interface device 190 via a wire 162. The wire 162 may be intracardiac, subcutaneous, epicutaneous, or any combination of the above. The patient interface device 190 may relay sensor data (e.g., data received from the sensor 160) to the drive unit 150 via a wired or wireless connection. In other embodiments, the sensor 160 may be connected to the drive unit 150 via a wired or wireless connection and may transmit sensor data directly to the drive unit 150 without the use of the patient interface device 190. The sensor 160 may be an implanted sensor, an external sensor, or any combination of the foregoing. For example, in some embodiments, the sensor 160 may be an implanted bipolar electrode positioned in and / or adjacent to the heart or other suitable tissue to determine, for example, when the left ventricle is contracting or relaxing. The sensor 160 may deliver to the drive unit 150 one or more of the following: electrical signals of the heart (e.g., signals that may be processed as ECG signals and displayed on a monitor or printer), pressure information related to pressure adjacent to the balloon pump 110 (e.g., adjacent to the proximal end of the balloon pump 100), ambient noise information (e.g., noise associated with the pectoral muscles, electronic noise from the system 100, or electronic noise from other devices (not shown) associated with or adjacent to the patient, etc.), noise generated by the heart during the cardiac cycle (e.g., information that may be processed as a phonocardiogram signal and displayed on a monitor or printer), other pressure signals, etc.In some embodiments, the sensor 160 is an electronic sound transducer, a pressure transducer, a piezoelectric transducer, or other transducer or sensor suitable for making physiological parameter measurements. Although Figure 1 shows three sensors 160, the system 100 may incorporate fewer or more than three sensors 160, depending on the nature of what is being sensed in any particular embodiment.

[0028] System 100 may provide chronic support of cardiac function and blood flow while still allowing the patient to remain ambulatory. Typical ventricular assist devices require femoral access for drive lines and / or connection to a large stationary external control unit, thus constraining the patient to a hospital bed in a supine position during the treatment period. In contrast, system 100 provides axillary / subclavian access, thus allowing the patient to move about relatively unhindered. Additionally, the level of therapy provided by system 100 may be adjusted to suit the needs of the patient. For example, the volume displacement and support ratio (e.g., 1:1, 1:2, 1:3 support) provided by balloon pump 110 may be adjusted to vary the support provided. By gradually decreasing the volume displacement over time, the load on the heart may be controlled, which may be beneficial in some cases for cardiac recovery. Additionally, unlike conventional circulatory assist systems, system 100 may be turned off to assess the patient's ability to cope with the demands of the heart without support, for example, before removing system 100 or for another suitable reason. In some embodiments, for example, the balloon pump 110 is designed to remain in the aorta in a contracted state for relatively long periods of time (e.g., 23 hours per day), as opposed to conventional devices that often must be removed if turned off for more than about 15 minutes, as reactivation can cause a large ejection of clots formed in the non-activated state.

[0029] (cardiac cycle signal) 2 illustrates an example ECG signal 202. The ECG signal 202 may be sensed by the sensor 160. The ECG signal 202 may include P waves 204, Q waves 206, R waves 208, S waves 210, and T waves 212. The Q waves 206, R waves 208, and S waves 210 may make up a QRS complex 214.

[0030] 3 illustrates an exemplary Wiggers diagram 300 of an exemplary cardiac cycle. The x-axis of the diagram plots time, subdivided into cardiac phases (left ventricular systole 302 and left ventricular diastole 304), and the y-axis of the diagram plots blood pressure (aortic pressure 306, atrial pressure 308, and ventricular pressure 310), ventricular volume 312, ECG signal 202, and phonocardiogram signal 314. Wiggers diagram 300 also illustrates, among other things, the cardiac timing for closure of mitral valve 316, opening of aortic valve 318, closure of aortic valve 320 (i.e., dicrotic notch), and opening of mitral valve 322, in addition to S1 heart sound 324, S2 heart sound 326, and S3 heart sound 328. The R wave 208 represents the start of left ventricular systole 302 and the dicrotic notch 320 or S2 heart sound 326 represents the start of left ventricular diastole 304.

[0031] (Drive unit system) FIG. 4 shows an exemplary schematic diagram of a drive unit system 400 according to some embodiments of the present technology. The drive unit system 400 may include an R-wave detection module 402 and a drive unit 404. The R-wave detection module 402 may detect one or more R-waves associated with a received ECG signal (measured by the sensor 160 and received as sensor data 401). In operation, the drive unit 404 may be coupled to an intra-aortic balloon pump 110 via the tube 172, the second drive line 140, and the first drive line 120, the balloon pump 110 being placed near the patient's heart as generally depicted in FIG. 1. The patient may be a human or other animal. In some embodiments, the drive unit system 400 may include the balloon pump 110. The drive unit 404 may inflate and deflate the balloon pump 110 based on the detection of the R-wave. Thus, the balloon pump 110 may perform counterpulsation.

[0032] (Filtered ECG signal) In some embodiments, the drive unit system 400 includes a filter module 406. The filter module 406 may convert the received ECG signal into a filtered ECG signal that emphasizes the R waves contained in the ECG signal and / or attenuates all or some of the other frequencies and waves that are not part of the R waves. For example, the filter module 406 may filter out all frequencies in the ECG signal other than those frequencies expected to define or otherwise contribute to the R waves 208. Such expected frequencies may be empirically determined and used as settings (e.g., by a clinician). In some embodiments, the filter module 406 filters out frequencies other than those in the 7.8125 Hz to 15.625 Hz band. In other embodiments, the filter module 406 filters out frequencies other than those in the 15.625 Hz to 31.25 Hz band. In yet other embodiments, the filter module 406 filters out frequencies other than those in the 7.8125 Hz to 31.25 Hz band. In some embodiments, the output of the filter module approximates the absolute value of the derivative of a smoothed version of the received ECG signal. In some embodiments, the filter module 406 may be implemented as a bandpass filter, a weighting filter, or a wavelet filter.

[0033] FIG 5A shows a schematic diagram of the wavelet filter 502. FIG 5B shows an example mother Haar wavelet function 512 that may be employed by the wavelet filter 502. FIG 5C shows an example mother symlet4 wavelet function 516 that may be employed by the wavelet filter 502. FIG 5D shows an example ECG signal that may be received by the drive unit system 400, and FIG 5E shows a plot 520 of an example multiple discrete frequency signal obtained by the wavelet filter 502 decomposing an ECG signal 518 using a Haar wavelet function (e.g., function 512). FIG 5F shows a plot of an example filtered ECG signal 524 and its corresponding ECG signal 522.

[0034] 5A-5F, the filter module 406 may be implemented as a wavelet filter 502. The wavelet filter 502 may include a discrete wavelet transform module 504, a frequency selector module 506, an aggregator module 508, and an absolute value module 510. The discrete wavelet transform module 504 may decompose at least a portion of a received ECG signal (e.g., ECG signal 518) into a plurality of discrete frequency signals (e.g., discrete frequency signals shown in plot 520 corresponding to coefficients generated from the decomposition), each of which is associated with a different frequency band using one or more wavelet transform functions, such as, for example, a mother Haar wavelet (Psi) function 512 or a mother Symlet4 wavelet (Psi) function 516.

[0035] The decomposition may be a multi-stage decomposition performed at multiple levels, generating wavelet approximation (a) and detail (d) coefficients at each level. For example, referring to FIG. 5B, the detail (d) coefficients 514 associated with level 1 (d1) may be in the 250-500 Hz subband, the detail coefficients 514 associated with level 2 (d2) may be associated with the 125-250 Hz subband, the detail coefficients 514 associated with level 3 (d3) may be associated with the 62.5-125 Hz subband, the detail coefficients 514 associated with level 4 (d4) may be associated with the 31.25-62.5 Hz subband, the detail coefficients 514 associated with level 5 (d5) may be associated with the 15.625-31.25 Hz subband, and the detail coefficients 514 associated with level 6 (d6) may be associated with the 7.8125-15.625 Hz subband. In some embodiments, the decomposition may produce only discrete frequency signals associated with frequencies or frequency bands that define or are otherwise expected to contribute to R waves.

[0036] The aggregator module 508 may aggregate one or more of the decomposed discrete frequency signals. In some embodiments, the frequency selector module 506 may be configured to identify frequencies or frequency bands expected to define or otherwise contribute to R-waves (e.g., frequencies in the 7.8125 Hz-15.625 Hz and 15.625 Hz-31.25 Hz bands associated with the d6 and d5 coefficients). For example, in embodiments in which multiple discrete frequency signals are generated, the frequency selector module 506 may identify and select one or more discrete frequency signals that may be aggregated by the aggregator module 508. In other embodiments, the frequency selector module 506 may tune the wavelet filter 502 to generate only discrete frequency signals associated with frequencies or frequency bands expected to define or otherwise contribute to R-waves, and the aggregator module 508 may aggregate all of the generated decomposed discrete frequency signals. The aggregated signal may be a filtered ECG signal. In some embodiments, an absolute value module 510 may be used to ensure that the filtered ECG signal is positive.

[0037] In some embodiments, the wavelet filter 502 includes an inverse discrete wavelet module that constructs or reconstructs a signal using one or more coefficients or decomposed discrete frequency signals. In such embodiments, the output of the inverse discrete wavelet module may be the filtered ECG signal. In some embodiments, an absolute value module 510 may be used to ensure that the filtered ECG signal is positive.

[0038] 5F, an example ECG signal 522 is shown alongside its filtered ECG signal 524. The filtered ECG signal 524 may have two peaks 524A and 524B corresponding to the rise from Q to R and the fall from R to S in the QRS complex 212. The peaks 524A and 524B may both be positive due to the operation of the absolute value module 510. In an embodiment in which the absolute value module 510 is omitted, the peaks 524A and 524B are of opposite signs and the shape of the filtered signal may be similar to the shape of the decomposed signal in the plot 520.

[0039] Figure 6A shows a plot 600 of an example ECG signal 602 including multiple premature ventricular contractions (PVCs) 604 in accordance with some embodiments of the present technology. Figure 6B shows a plot 606 of an example filtered ECG signal 608 associated with the ECG signal 602 of Figure 6A. The larger peak 608A may be associated with the QRS complex 212 in the ECG signal 602, and the smaller peak 608B may be associated with the PVCs 604.

[0040] (R wave detection) In some embodiments, the R-wave detection module 402 may be configured to detect R-waves associated with the received ECG signal by detecting a "crossing" of one or more R-wave detection thresholds by the filtered ECG signal. In some embodiments that do not use the absolute value module 510, the embodiment may employ positive and negative R-wave detection thresholds, and a "crossing" occurs when the current value of the filtered ECG signal (e.g., signal 524) crosses either R-wave detection threshold (e.g., when the current value of the filtered ECG signal changes to be greater than the positive R-wave detection threshold or changes to be less than the negative R-wave detection threshold). In embodiments in which the absolute value module 510 is employed, the filtered ECG signal may be positive, and a "crossing" may occur when the current value of the filtered ECG signal changes to be greater than the R-wave detection threshold.

[0041] FIG. 7 illustrates a plot 700 of a portion of an exemplary filtered ECG signal 702 highlighting a single QRS complex, multiple R-wave detection thresholds 704, 706, 708, and multiple crossings 710, 712, and 714 of the filtered ECG signal 702 relative to such R-wave detection thresholds 704, 706, 708. The R-wave detection thresholds 704, 706, 708 may be any suitable values ​​that support counterpulsation. In some embodiments, the R-wave detection thresholds may be based on a predicted filtered ECG maximum. In some embodiments, the R-wave detection thresholds may represent a percentage of the predicted filtered ECG maximum. For example, the R-wave detection threshold 704 may be approximately 40% of the predicted filtered ECG maximum, the R-wave detection threshold 706 may be approximately 10% of the predicted filtered ECG maximum, and the R-wave detection threshold 708 may be approximately 60% of the predicted filtered ECG maximum.

[0042] In some embodiments, a single R-wave detection threshold, e.g., a nominal R-wave detection threshold 704, is used to detect R-waves. The nominal R-wave detection threshold 704 may be set to about 40% of the predicted filtered ECG maximum. In such embodiments, a crossing 710 of the single nominal R-wave detection threshold 704 by the filtered ECG signal 702 may be considered a "nominal detection" of an R-wave, and deflation of the balloon pump 101 may be based on such a nominal detection. Specifically, the deflation may be based on a deflation trigger, as discussed below. Once the crossing 710 is detected (or nominal detection), the drive unit 404 may instantly cause the deflation of the balloon pump 101. In some embodiments, the drive unit 404 may cause the balloon pump 110 to deflate at least 50% in about 100-120 ms of the single crossing 710 and to fully deflate in about 140-250 ms of the single crossing 710. The dilation may also be based on the same intersection 710 (as described below).

[0043] Using such a single nominal R-wave detection threshold, e.g., nominal R-wave detection threshold 704, to detect a single crossing 710 (i.e., nominal detection) can create challenges in supporting arrhythmic cardiac patterns such as atrial fibrillation (a-fib) and PVCs, or when the ECG is noisy (e.g., due to pectoral muscle activity or other noise interference). For example, with reference to FIGs. 6A and 6B, if the single nominal R-wave detection threshold 704 is set too high, it can delay detection of an R-wave or "miss" or not detect an R-wave (e.g., R-wave 608B associated with a PVC 604), which can prevent ejection during the systole associated with such a missed R-wave. In such cases, it may be better to risk a false detection than to risk missing an R-wave. Similarly, if the single nominal R-wave detection threshold 704 is set too low, false positives caused by, for example, a noisy ECG signal (e.g., the noisy ECG signal 1002 of FIG. 10A, discussed below) may result in inflation being triggered too early and ejection may also be prevented. In such cases, it may be better to miss an R-wave than to falsely detect an R-wave.

[0044] Also, two R-wave detection thresholds may be used to detect R-waves. For example, a first or relatively low R-wave detection threshold and a second or relatively high R-wave detection threshold may be employed, such values ​​being relative to a single nominal R-wave detection threshold. Crossing of the first or lower R-wave detection threshold by the filtered ECG signal may be considered a "speculative detection" of an R-wave, and thus the first or lower R-wave detection threshold may be considered a "speculative R-wave detection threshold." Similarly, crossing of the second or higher R-wave detection threshold by the filtered ECG signal may be considered a "conservative detection" of an R-wave, and thus the second or higher R-wave detection threshold may be considered a "conservative R-wave detection threshold."

[0045] In such embodiments where a dual R-wave detection threshold is used to perform R-wave detection, speculative detection may be associated with deflation only. Upon speculative detection, the drive unit 404 may rapidly cause deflation of the balloon pump 101. In some embodiments, the drive unit 404 may cause the balloon pump 110 to deflate at least 50% at approximately 100-120 ms of speculative detection and fully deflate at approximately 140-180 ms of speculative detection. Conservative detection may be associated with inflation, as described below.

[0046] For example, in FIG. 7, the first or speculative R-wave detection threshold 706 may be less than the single nominal R-wave detection threshold 704. In the example of FIG. 7, the first R-wave detection threshold 706 is about 10% of the predicted filtered ECG maximum. The speculative detection may occur at crossing 712, i.e., when the filtered ECG signal 702 crosses the first R-wave detection threshold 706. Using a lower R-wave detection threshold 706 for the same ECG signal compared to the single nominal R-wave detection threshold 704 may reduce the risk of not detecting (i.e., missing) an R-wave because the speculative detection (i.e., crossing 712 compared to crossing 710) is "early" (e.g., 10-25 ms earlier). This early speculative detection reduces the likelihood that the balloon pump 110 will remain inflated for too long (i.e., during the early stages of LV systole 302).

[0047] The second or conservative R-wave detection threshold 708 may be greater than the single nominal R-wave detection threshold 704 to reduce the risk of false detection of an R-wave. In the example of FIG. 7, the second R-wave detection threshold 708 is about 60% of the predicted filtered ECG maximum. The conservative detection may occur at crossing 714, i.e., when the filtered ECG signal 702 crosses the second R-wave detection threshold 708. Using a higher (or larger) R-wave detection threshold 708 for the same ECG signal compared to the single R-wave detection threshold 704 may reduce the risk of false detection of an R-wave, and the conservative detection (i.e., crossing 714 compared to crossing 710) may occur later in time. This later conservative detection reduces the chance of early inflation (e.g., during the later stages of LV systole).

[0048] Cumulatively, using dual R-wave detection thresholds such as values ​​706 and 708 instead of a single nominal R-wave detection threshold 704 means that in the majority of cases, an R-wave will be detected due to the relatively low a priori R-wave detection threshold 704 and the balloon pump 110 will deflate in time. However, if the relatively large conservative R-wave threshold 708 is not crossed, perhaps the only drawback is that a cardiac cycle initiated by an undetected R-wave will not be supported by an inflated blood pump 110. An advantage of the dual threshold embodiment is that the detection of an R-wave can occur 10-20 ms earlier than the nominal detection, which allows the balloon to be deflated sooner, minimizing the possibility of its occlusion during ventricular ejection.

[0049] FIG. 8A illustrates a plot 800 of an exemplary ECG signal 802 in accordance with some embodiments of the present technology. FIG. 8B illustrates a plot 804 of an exemplary filtered ECG signal 806 associated with the ECG signal 802. The plot 804 further depicts multiple R-wave detection thresholds 808, 810, 812, and associated crossings of such values ​​(814A-D, 815A-D, and 816A-D) by the filtered ECG signal 806. The R-wave detection threshold 808 may be the so-called "single nominal R-wave detection threshold" as described above and associated with single threshold embodiments. The single nominal R-wave detection threshold 808 may be set at approximately 40% of the predicted filtered ECG maximum. The R-wave detection thresholds 810 and 812 may be the so-called "first and second R-wave detection thresholds" or "a priori and conservative R-wave thresholds" as described above and associated with dual threshold embodiments. The first and second R-wave detection thresholds may be set at approximately 10% and 60% of the predictive filtered ECG maximum, respectively. In some embodiments, the use of a single nominal R-wave detection threshold (compared to using dual R-wave detection thresholds) may result in delayed detection. For example, in FIG. 8A, detection may be delayed for the third R-wave because crossing 814C occurs well after crossing 815C.

[0050] (Predictively filtered ECG maximum value) 4 and 9, the drive unit system 400 may include a filtered ECG maximum module 408 that generates a predicted filtered ECG maximum based on one or more actual maxima in the filtered ECG signal or one or more portions thereof. In some embodiments, the filtered ECG maximum module 408 applies a windowing technique to generate the predicted filtered ECG maximum. FIG. 9 shows a plot 606 (i.e., an example filtered ECG signal 608 from FIG. 6) along with an example windowing technique that may detect filtered ECG maxima 904A-E in accordance with some embodiments of the present technology. The filtered ECG maximum module 408 may receive at least one portion 902 of the filtered ECG signal 608, which is divided into multiple sub-portions or windows 902A-E. In other embodiments, the filtered ECG maximum module 408 may generate the portion 902 and the windows 902A-E. In some embodiments, the portion 902 is 7.5 seconds long and each window 902A-E is 1.5 seconds long. Other sizes and more or fewer windows are contemplated. For example, as described in connection with FIG. 10D, a portion may be 3 seconds long and have two 1.5 second windows. In such embodiments, using fewer windows allows the filtered ECG maximum value module 408 to be more responsive to changes in the patient's ECG. In some embodiments, the size of the portion and the number and size of the windows may be selected to ensure that at least some of the windows include peaks associated with R waves. In some embodiments, the size of the portion and the number and size of the windows may be selected to ensure that at least half of the windows or a majority (i.e., more than half) of the windows include peaks associated with R waves.

[0051] Returning to Figure 9, the filtered ECG maximum module 408 may detect a maximum value 904A-E of the filtered ECG signal 608 in each window 902A-E. The filtered ECG maximum module 408 may then generate a predicted filtered ECG maximum value based on one or more of the detected filtered ECG maxima 904A-E. For example, the predicted filtered ECG maximum value may be the average, median, minimum, or maximum of one or more of the detected filtered ECG maxima 904A-E.

[0052] The filtered ECG maximum value module 408 may generate the predicted filtered ECG maximum value continuously, periodically (e.g., during a learning state), or any combination of the above. The filtered ECG signal 904 may be current, or historical, or some combination of both current and historical filtered ECG signals 904.

[0053] The R-wave detection thresholds (single / nominal, first / speculative, and second / conservative) mentioned above may be based on the predicted filtered ECG maximum. For example, the R-wave detection threshold may be a percentage of the predicted filtered ECG maximum. The first or speculative R-wave detection threshold may be a first percentage of the predicted filtered ECG maximum, and the second or conservative R-wave detection threshold may be a second percentage of the predicted filtered ECG maximum. The second percentage may be greater than the first percentage. In one embodiment, the first R-wave detection threshold is between about 7% and about 30% of the predicted filtered ECG maximum, and the second R-wave detection threshold is between about 30% and about 60% of the predicted filtered ECG maximum. For purposes of this disclosure, the term "about" means that the percentages may vary by up to 2 percentage points with respect to other number percentages or number ranges. Thus, in the preceding sentence, the first R-wave detection threshold may be between 5-9% and 28-34% of the predictively filtered ECG maximum, and similarly, the second R-wave detection threshold may be between 28-32% and 58-62% of the predictively filtered ECG maximum.

[0054] In some embodiments, the drive unit system 400 includes a GUI 410 that allows a clinician to adjust the R-wave detection threshold as a device setting to customize and make counterpulsation more efficient for each patient. In some embodiments, the clinician may select the desired setting based on observation of a current or past ECG signal associated with the patient. The R-wave detection module 402 may generate the first and second R-wave detection thresholds or a single R-wave threshold based on default settings and / or based on input from the GUI 410. For example: (a) for patients with low PVCs (e.g., where 5% or less of the R waves in the ECG are PVCs) and for patients with no history of atrial fibrillation and normal sinus rhythm, the clinician can adjust the system settings so that the first R-wave detection threshold is about 15% of the predictive filtered ECG maximum and the second R-wave detection threshold is about 50% of the predictive filtered ECG maximum; (b) for patients with a-fib or frequent PVCs (e.g., at least 20% of the R waves in the ECG are PVCs), the clinician can adjust the system settings so that the first R-wave detection threshold is about 7% of the predictive filtered ECG maximum and the second R-wave detection threshold is about 60% of the predictive filtered ECG maximum; (c) for patients in which the amplitude of the P or T waves is similar to the R waves (e.g., the amplitude of one or more P or T waves is about 90% of the maximum amplitude of an R wave within a predetermined time window, such as 5 seconds), and who are free of arrhythmia, the clinician can adjust the system settings so that the first R wave detection threshold is about 30% of the predictive filtered ECG maximum and the second R wave detection threshold is about 60% of the predictive filtered ECG maximum; (d) For patients with large R-wave amplitude variability (e.g., a 50% or greater change (increase or decrease) in maximum amplitude between R-waves within a given time window, such as 5 seconds), the clinician can adjust the system settings so that the first R-wave detection threshold is approximately 10% of the predictive filtered ECG maximum and the second R-wave detection threshold is approximately 30% of the predictive filtered ECG maximum.

[0055] In certain embodiments, the drive unit system 400 includes a noise estimator module 412 operable to generate noise data indicative of the presence of noise (e.g., noise that may interfere with R-wave detection as described above). Sources of such noise may be pectoral muscle activity and / or electrical circuitry in the vicinity of the patient. The noise estimator module 412 may generate the noise data based on information, data, or signals from the sensor 160 (e.g., an electronic sound transducer sensor). The R-wave detection module 402 may adjust the first / predictive, second / conservative, or single / nominal R-wave detection thresholds based on the generated noise data. For example, the R-wave detection module 402 may adjust (e.g., increase) any or all of the R-wave detection thresholds for any duration of noise. In other embodiments, the R-wave detection module 402 may generate the first, second, or single R-wave detection thresholds based on the predictive filtered ECG maximum and noise data.

[0056] In other embodiments, the noise estimator module 412 may generate noise data based on the filtered ECG signal. For example, the noise estimator module 412 may employ a windowing technique to generate an ambient noise value based on the filtered ECG signal. The windowing technique may be similar to the windowing technique described above with respect to the filtered ECG maximum value module 408.

[0057] FIG. 10A shows a plot 1000 of an exemplary "noisy" ECG signal 1002. FIG. 10B shows a plot 1003 of a corresponding filtered ECG signal 1004. The filtered ECG signal 1004 may be generated using the filter module 406, as described above. FIG. 10C is a zoomed-in view of a first segment 1006 of the filtered ECG signal 1004, and FIGs. 10D and 10E are zoomed-in views of a second segment 1007 of the filtered ECG signal 1004.

[0058] 10A-10E, the noise estimator module 412 may obtain a portion 1008 of a filtered ECG signal 1002 having multiple subportions or windows 1008A-G. Alternatively, the noise estimator module 412 may generate the portion 1008 and associated windows 1008A-G. The portion 1008 may be 420 ms long and the multiple windows 1008A-G may include seven windows, each of length 60 ms. The size of the portion and the number and size of the windows may vary. In some embodiments, the size of the portion and the number and size of the windows may be selected to ensure that at least some of the windows do not include peaks associated with R-waves. In some embodiments, the size of the portion and the number and size of the windows may be selected to ensure that at least half of the windows or a majority (i.e., more than half) of the windows include peaks that are not associated with R-waves. For example, by selecting a portion size of 420 ms and using seven windows, each 60 ms long, it is expected that for heart rates up to 200 bpm, no more than three subportions or windows will contain peaks associated with the R wave.

[0059] 10C, for each window 1008A-D, the noise estimator module 412 may detect a filtered ECG maximum 1010A-G. The noise estimator module 412 may generate an ambient noise value as a function of one or more of the detected maxima 1010A-G. For example, the ambient noise value may be the average, median, or minimum of such detected maxima 1010A-G. In the example shown in FIG. 10C, the noise estimator module 412 is configured such that the ambient noise value is the median of the seven detected maxima 1010A-G. Using the data in this figure, the ambient noise value in FIG. 10C corresponds to the maximum 101D.

[0060] The noise estimator module 412 may generate the ambient noise value continuously, periodically (e.g., during a learning state), or any combination of the foregoing. The filtered ECG signal used to generate the ambient noise value may be current or historical, or some combination of both current and historical filtered ECG signals.

[0061] The filtered ECG maximum value module 408 may perform its own windowing technique on the same or another portion of the filtered ECG signal 1004 to generate a predicted filtered ECG maximum value. For example, with respect to FIG. 10D showing an expanded plot of a segment 1007 of the filtered ECG signal 1004, the filtered ECG maximum value module 408 may generate a predicted filtered ECG maximum value based on the portion 1008 and the windows 1008A-B. In one embodiment, the portion 1008 is 3 seconds long and each of the windows 1008A-B is 1.5 seconds long. The predicted filtered ECG maximum value may be based on one or more of the maxima within the windows 1008A-B. In some embodiments, the predicted filtered ECG maximum value may be an average, a minimum, or a maximum of the detected maxima. In the example shown in FIG. 10D, the filtered ECG maximum value module 408 is configured such that the predicted filtered ECG maximum value is the minimum of the detected maxima. Using the data in this figure, the predictive filtered ECG maximum in FIG. 10C corresponds to the detected maximum 1114B.

[0062] In some embodiments, the R-wave detection module 402 may generate one or more R-wave detection thresholds based on the predicted filtered ECG maximum and the ambient noise value 1012. In some embodiments, the ambient noise value acts as a bias. In some embodiments, the R-wave detection threshold may be the sum of (1) the ambient noise value and (2) the product of a weighted difference between the predicted filtered ECG maximum 1016 and the ambient noise threshold 1012. In equation form, the R-wave detection threshold is set as in Equation 1.0, where "RDTV" is the R-wave detection threshold, "ANV" is the ambient noise value, "n" is a weighting factor, and "PMV" is the predicted filtered ECG maximum. RDTV=ANV+{n·(PMV-ANV)}(Formula 1.0)

[0063] The weighting factor (i.e., "n" in Equation 1.0) may be about 40% in embodiments using a single / nominal R wave detection threshold to detect R waves, or in embodiments using dual R wave detection thresholds, the weighting factor for the first / speculative R wave detection threshold may be between about 7-30% and the weighting factor for the second / conservative R wave detection threshold may be between about 30-60%.

[0064] In embodiments in which the drive unit system 400 includes a GUI 410, a clinician can adjust the R-wave detection threshold as a device setting to customize and make counterpulsation more efficient for each patient. In some embodiments, the clinician may select the desired setting based on observation of a current or past ECG signal associated with the patient. The R-wave detection module 402 may generate the first and second R-wave detection thresholds or a single R-wave threshold based on default settings and / or based on input from the GUI 410. For example: (a) for patients with low PVCs (e.g., where 5% or less of the R waves on the ECG are PVCs) and for patients with no history of atrial fibrillation and normal sinus rhythm, the clinician may adjust the system settings so that the weighting factor for the first R wave detection threshold is about 15% and the weighting factor for the second R wave detection threshold is about 50%; (b) for patients with a-fib or frequent PVCs (e.g., at least 20% of the R waves in the ECG are PVCs), the clinician may adjust the system settings so that the weighting factor for the first R wave detection threshold is about 7% and the weighting factor for the second R wave detection threshold is about 60%; (c) for patients whose P-wave or T-wave amplitudes are similar to R-waves (e.g., the amplitude of one or more P-waves or T-waves is about 90% of the maximum amplitude of an R-wave within a predetermined time window, such as 5 seconds), and who are free of arrhythmia, the clinician may adjust the system settings such that the weighting factor for the first R-wave detection threshold is about 30% and the weighting factor for the second R-wave detection threshold is about 60%; (d) For patients with large R-wave amplitude variability (e.g., a 50% or greater change (increase or decrease) in maximum amplitude between R-waves within a given time window, such as 5 seconds), the clinician may adjust the system settings so that the weighting factor for the first R-wave detection threshold is approximately 10% and the weighting factor for the second R-wave detection threshold is approximately 30%.

[0065] Figure 10E shows an enlarged segment 1007 of the filtered ECG signal 1004 for an exemplary ambient noise value 1012 (determined in Figure 10C), a predicted filtered ECG maximum value 1016 (determined in Figure 10D), and exemplary first, single and second R-wave detection thresholds 1018, 1020 and 1022, respectively. The single R-wave detection threshold 1020 is based on a weighting factor of 40%, the first R-wave detection threshold 1018 is based on a weighting factor of 7% and the second R-wave detection threshold 1022 is based on a weighting factor of 60%.

[0066] (Contraction trigger) As discussed above, the deflation of the balloon may be based on a deflation trigger. The deflation trigger may be based on either a single / nominal detection (e.g., in embodiments with a single R-wave detection threshold) or an a priori detection (e.g., in embodiments with a dual R-wave detection threshold). In some embodiments, the deflation trigger is a single detection or an a priori detection. In other embodiments, the deflation trigger is the first of (1) a single detection or an a priori detection (based on the embodiment) and (2) the expiration of a predicted RR time interval. The predicted RR time interval may represent the time at which an R-wave is "expected" or "predicted to be detected." In embodiments with a single R-wave detection threshold, the predicted RR time interval may be measured from the last (e.g., most recent) single / nominal detection, e.g., detection 710. In embodiments with two R-wave detection thresholds, the predicted RR time interval may be measured from the last (e.g., most recent) conservative detection, e.g., conservative detection 714.

[0067] In some embodiments, the drive unit system 400 includes an RR interval module 414 that generates a predicted RR time interval based on a filtered ECG signal (e.g., from the filter module 406). For example, referring to FIG. 8B, the predicted RR time interval may be based on the time between single / nominal detections 814A-D or the time between conservative detections 816A-D. In some embodiments, the predicted RR time interval is the average or median of the time between multiple such past detections.

[0068] In other embodiments, the RR interval module 414 operates similarly to (or is part of the same module as) the filtered ECG maximum module 408, but adds functionality related to detecting the time of each filtered ECG maximum 904A-E. The predicted RR time interval may be a function of such detected times. For example, the predicted RR time interval may be the average or median of such detected times.

[0069] In embodiments where the deflation trigger is the first of (1) detection (single / nominal or conservative) and (2) expiration of the predicted RR time interval, the latter component ensures that the balloon pump 110 is deflated if the technique misses an R wave.

[0070] (expansion) The drive unit 404 may inflate the balloon pump 110 based on a single / nominal detection (e.g., either detection 710 or detections 814A-D) or based on a conservative detection (e.g., either detection 714 or detections 816A-D). In some embodiments, the drive unit 404 inflates the balloon pump 110 based on the time of the applicable detection and a predicted QS2 timing interval. As shown in FIG. 3, the QS2 timing interval 330 is the time interval from the onset of the Q wave 206 to the S2 heart sound 326. The predicted QS2 timing interval may approximate the actual QS timing interval 330. In some embodiments, the drive unit 404 inflates the balloon pump 110 at a time equal to or approximately equal to the detection (e.g., single detection or conservative detection) plus the predicted QS2 time interval.

[0071] In some embodiments, the predicted QS2 timing interval is empirically derived based on scientific studies. In other embodiments, the predicted QS2 timing interval is patient-specific. For example, the predicted QS2 time interval may be obtained using a phonocardiograph system (not shown) capable of detecting the S2 heart sound 326 and the time of each such detection. In other embodiments, referring to FIG. 4, the drive unit system 400 may include a QS2 time interval module 416 that generates the predicted QS2 timing interval based on pressure data sensed by the sensor 160. In some embodiments, the drive unit system 400 pauses counterpulsation and the drive unit 404 partially inflates the balloon pump 110 for one or more cardiac cycles. In certain embodiments, the balloon pump 110 is inflated to about 50% of its volume. During LV systole (i.e., ejection) 302, the balloon pump 110 remains partially inflated and a pressure sensor located within the drive unit 404 senses the change in pressure caused by compression of the balloon pump 110 during systole. The drive unit 404 may communicate pressure information to the QS2 time interval module 416, which may indirectly detect the dicrotic notch 320 from such pressure information. Based on the time of detection of one or more dicrotic notches 320, the QS2 time interval module 416 may generate a predicted QS2 timing interval that represents the time from detection of an R-wave 208 to the predicted dicrotic notch. In certain embodiments, the predicted QS2 timing interval is based on the time of one or more R-wave detections (e.g., as determined by the R-wave detection module 402) and one or more dicrotic notch detections (as determined by the QS2 time interval module 416). For example, the predicted QS2 timing interval may be the average or median of the time difference between the occurrence of one or more R-wave detections and the occurrence of one or more dicrotic notch detections.

[0072] FIG. 11 illustrates an exemplary method 1100 for using the drive unit system 400 of FIG. 4 in accordance with some embodiments of the present technology. The method 1100 may include block 1102, where a received filtered ECG signal may be generated. In some embodiments, the filtered ECG signal may be generated by the filter module 406. In some embodiments, the method 1100 may include block 1101, where an absolute value of the filtered ECG signal may be obtained. In some embodiments, the absolute value of the filtered ECG signal may be obtained by the absolute value module 510. The method 1100 may also include block 1103, where a predicted RR time interval is generated or obtained. In some embodiments, the predicted RR time interval may be generated or obtained by the RR time interval module 414. The method 1100 may also include block 1105, where a predicted filtered ECG maximum value is generated or obtained. In some embodiments, the predicted filtered ECG maximum value is generated or obtained by the filtered ECG maximum value module 408. Method 1100 may also include block 1107, where an ambient noise value is generated or obtained. In some embodiments, the ambient noise value is generated or obtained by noise estimator module 412. Method 1100 may also include block 1109, where a predicted QS2 time interval is generated or obtained. In some embodiments, the predicted QS2 time interval is generated or obtained by QS2 time interval module 416. Method 1100 may also include blocks 1111 and 1113, where a first and second R wave detection value is generated or obtained, respectively. In some embodiments, the first and second R wave detection value is generated or obtained by R wave detection module 402. In some embodiments, the first and second R wave detection value is based on a predicted filtered ECG maximum value. In some embodiments, the first and second R wave detection value is based on a predicted filtered ECG maximum value and an ambient noise value.

[0073] Method 1100 may include block 1104 where a priori detection of R waves is performed, and 1106 where conservative detection of R waves is performed. In some embodiments, the a priori and conservative detection may be performed by R wave detection module 402. Method 1100 may include block 1115, where it is determined whether the filtered ECG signal exceeds or crosses a first R wave threshold and / or a second R wave threshold. In some embodiments, this determination may be performed by R wave detection module 402.

[0074] The method 1110 may include block 1108, where the balloon pump 110 may be deflated based on a deflation trigger. In some embodiments, the deflation is triggered by the drive unit 404. The method 1110 may include block 1107, where the deflation is based on the first occurring of a priori detection and expiration of a predicted RT time interval. In some embodiments, such deflation is triggered by the drive unit 404.

[0075] Finally, the method 1100 may include block 1110, where the balloon pump 110 may be inflated based on the conservative detection. In some embodiments, the inflation is triggered by the drive unit 404. The method 1100 may include block 1119, where the balloon pump 110 is inflated based on the conservative detection and the predicted QS2 timing interval. In some embodiments, such inflation is triggered by the drive unit 404.

[0076] 12 shows an example state diagram 1200 of the drive unit system 400 in accordance with some embodiments of the present technology. The state diagram 1200 may include three states: an initialization state 1202, a learning state 1204, and a detection state 1206. During the initialization state 1202, an ECG signal and sensor data may be initialized and received. The learning state 1204 may be associated with method blocks 1102, 1101, 1103, 1105, 1107, 1109, 1111, and 1113. The detection state 1206 may be associated with method blocks 1104, 1106, 1115, 1108, 1117, 1110, and 1119. In some embodiments, the detection state 1206 may be associated with all method blocks (e.g., where parameters and values ​​may be continuously updated).

[0077] The techniques and techniques described herein provide efficient and effective detection of R-waves for use by CADs, including but not limited to counterpulsation devices. Employing dual R-wave detection thresholds, the techniques minimize both missed R-waves (thereby reducing the likelihood that the balloon pump 110 will remain inflated for too long) and false positives (thereby reducing the likelihood that the balloon pump 110 will inflate too early and prevent systolic drainage). Although the techniques described above have been described with particular reference to counterpulsation devices, the techniques may be readily applied to other CADs operating in copulsation or counterpulsation modalities. For example, while the above refers to the inflation of the IABP during diastole and the contraction of the IABP during systole, one skilled in the art will readily understand that the inflation of the IABP increases the aortic blood pressure and the contraction decreases the aortic blood pressure. Thus, the techniques described above may be applied to any modality of CAD, including CADs operating in copulsation and counterpulsation modalities that operate to increase or decrease the aortic blood pressure in synchrony with the heart. For example, the above-described techniques may use a VAD operating in copulsation and modalities to determine when to increase aortic blood pressure (i.e., during systole) and when to decrease aortic blood pressure (i.e., during diastole).

[0078] As used herein, a "module" may refer to any single or collection of circuits, integrated circuits, hardware processors, processing devices, transistors, non-transitory memories, storage devices, non-transitory computer-readable media, combinatorial logic circuits, or any combination of the above that can provide a desired operation or functionality. For example, a "module" may take the form of a hardware processor that executes instructions from one or more non-transitory memories, storage devices, or non-transitory computer-readable media, or may take the form of a dedicated integrated circuit. "Non-transitory memory," "non-transitory computer-readable media," and "storage device" may refer to any suitable internal or external non-transitory, volatile or non-volatile memory device, memory chip, or storage device or chip, such as, but not limited to, system memory, frame buffer memory, flash memory, random access memory (RAM), read-only memory (ROM), registers, latches, etc., or any combination of the above. A "hardware processor" may mean one or more dedicated or non-dedicated hardware microprocessors, hardware microcontrollers, hardware sequencers, hardware microsequencers, digital signal hardware processors, hardware processing engines, hardware accelerators, application specific circuits (ASICs), hardware state machines, programmable logic arrays, any integrated circuits capable of processing data or information, discrete circuits, etc., or any suitable combination thereof. A "processing device" may mean any number of physical devices capable of processing (e.g., performing various operations on) information (e.g., information in the form of binary data, or information carried / represented by any suitable media signal, etc.). For example, a processing device may be a hardware processor capable of executing executable instructions, a desktop computer, a laptop computer, a mobile device, a handheld device, a server (e.g., a file server, a web server, a program server, or any other server), any other computer, etc., or any combination of the above.An example of a processing device may be a device that includes one or more integrated circuits with transistors programmed or configured to perform particular tasks. "Executable instructions" may mean software, firmware, programs, instructions, or any other suitable instructions or commands capable of being processed by a suitable hardware processor.

[0079] Although exemplary embodiments have been described herein, the scope includes any and all embodiments having equivalent elements, modifications, omissions, combinations (e.g., combinations of aspects across various embodiments), adaptations and / or alterations as would be understood by one of ordinary skill in the art based on this disclosure. For example, the number and orientation of components shown in the exemplary systems may be modified.

[0080] Those skilled in the art will appreciate that the method illustrated in Figure 11 may be modified in various ways, such as by rearranging the order of acts, performing some acts in parallel, omitting or including other acts, dividing an illustrated act into sub-acts, combining multiple illustrated acts into a single act, etc.

[0081] Those skilled in the art will appreciate that the functionality described above can be readily adapted or extended in various ways. While the foregoing description has been made with reference to specific embodiments, the scope of the present invention is defined only by the following claims and the elements recited therein.

Claims

1. 1. A drive unit system for operating a cardiac assist device, the drive unit system comprising: An R wave detection module, a priori detection of an R wave associated with a received electrocardiogram (ECG) signal of the patient's heart; Performing a conservative detection of the R-waves associated with the ECG signal. an R-wave detection module operable to: a drive unit operably coupled to the cardiac assist device, the drive unit operable to operate the cardiac assist device based on the a priori detection and the conservative detection; A drive unit system comprising:

2. The drive unit system of claim 1 , wherein the cardiac assist device is a counterpulsation device.

3. 3. The drive unit system of claim 2, wherein the drive unit is operable to reduce an aortic blood pressure relative to an unassisted aortic blood pressure based on the a priori detection and to increase an aortic blood pressure relative to an unassisted aortic blood pressure based on the conservative detection.

4. The drive unit system of claim 2 , wherein the drive unit is operable to reduce aortic blood pressure based on a deflation trigger, the deflation trigger being based on the a priori detection.

5. The drive unit system of claim 3 , wherein the cardiac assist device is an intra-aortic balloon pump.

6. 6. The drive unit system of claim 5, wherein the drive unit is operable to deflate the intra-aortic balloon pump based on the a priori detection and to inflate the intra-aortic balloon pump based on the conservative detection.

7. 5. The drive unit system of claim 4, wherein the deflation trigger is the first occurrence of the a priori detection of the R-wave and expiration of a predicted R-R time interval.

8. The drive unit system of claim 7, wherein the expiration of the predicted R-R time interval is measured from the time of the most recent conservative detection of a previous R-wave.

9. The drive unit system of claim 7 , further comprising an R-R time interval module operable to generate the predicted R-R time interval based on the filtered ECG signal.

10. The drive unit system of claim 9 , wherein the R-R time interval module is operable to generate the predicted R-R time interval based on multiple conservative detections of R-waves.

11. The predicted R-R time interval is: an average of the time between a plurality of said conservative detections; a median time between a plurality of said conservative detections; 11. The drive unit system of claim 10, wherein the drive unit system is one of:

12. The R-R time interval module includes: receiving at least a portion of the filtered ECG signal divided into a plurality of windows; For each window, find the time associated with a maximum value of the filtered ECG signal within each window. It is possible to operate as follows: The drive unit system of claim 9, wherein the predicted R-R time interval is based on one or more of the detected times associated with a plurality of the maximum values.

13. The predicted R-R time interval is: an average of one or more of the detected times; a median value of one or more of the detected times; 13. The drive unit system of claim 12, wherein the drive unit system is one of:

14. The drive unit system of claim 4 , further comprising a filter module operable to convert the received ECG signal into a filtered ECG signal.

15. The drive unit system of claim 14 , wherein the filter module filters out frequencies not associated with R-waves.

16. The drive unit system of claim 14 , wherein the filtered ECG signal approximates the absolute value of the derivative of a smoothed version of the ECG signal.

17. The drive unit system of claim 14 , wherein the filter module comprises a bandpass filter.

18. The drive unit system of claim 14 , wherein the filter module comprises a weighting filter.

19. The drive unit system of claim 14 , wherein the filter module comprises a wavelet filter.

20. The wavelet filter is decomposing at least a portion of the ECG signal into a plurality of discrete frequency signals; aggregating signals based on one or more of the decomposed signals 20. The drive unit system of claim 19, operable to:

21. The drive unit system of claim 20 , wherein the one or more of the decomposed signals have a frequency associated with an R-wave.

22. The drive unit system of claim 19 , wherein the wavelet filter comprises a discrete wavelet transform module and an aggregator module.

23. the discrete wavelet transform module is operable to perform a discrete wavelet multi-stage decomposition of the ECG signal using a wavelet transform function at a plurality of levels, thereby generating wavelet approximation (a) and detail (d) coefficients at each such level; The drive unit system of claim 22, wherein the aggregator module is operable to generate an aggregated signal by aggregating one or more coefficients.

24. 24. The drive unit system of claim 23, wherein at least one or more of the levels is associated with a frequency associated with an R-wave.

25. 21. The drive unit system of claim 20, wherein the filter module comprises an absolute value module operable to generate the filtered ECG signal by taking an absolute value of the one or more of the aggregated and decomposed signals.

26. The drive unit system of claim 18 , wherein the wavelet filter is configured to apply a Haar wavelet function.

27. The drive unit system of claim 18 , wherein the wavelet filter is configured to apply a symlet4 wavelet function.

28. The R wave detection module includes: performing the a priori detection based on the filtered ECG signal; Performing the conservative detection based on the filtered ECG signal.

15. The drive unit system of claim 14, operable to:

29. The R-wave detection module is further operable to determine a current value of the filtered ECG signal; the R-wave detection module is operable to perform the a priori detection by determining that a current value of the filtered ECG signal exceeds a first R-wave detection threshold; the R-wave detection module is operable to perform the conservative detection by determining that a current value of the filtered ECG signal exceeds a second R-wave detection threshold; The drive unit system of claim 18 , wherein the second R wave detection threshold is greater than the first R wave detection threshold.

30. 30. The drive unit system of claim 29, wherein the first and second R-wave detection thresholds are based on a predictive filtered ECG maximum value.

31. 31. The drive unit system of claim 30, wherein the system further comprises a maximum filtered ECG value module operable to generate the predicted filtered ECG maximum value based on a filtered ECG maximum value in the filtered ECG signal.

32. The maximum filtered ECG value module converts the predicted filtered ECG maximum value into receiving at least a portion of the filtered ECG signal divided into a plurality of windows; for each window, by finding the maximum value of the filtered ECG signal within each window; 31. The drive unit system of claim 30, operable to generate.

33. The predictive filtered ECG maximum value is an average of one or more of said detected maxima; and a median value of one or more of said detected maxima; and a minimum of one or more of said detected maxima; and the maximum of one or more of said detected maximum values; 33. The drive unit system of claim 32, wherein:

34. the first R-wave detection threshold is a first percentage of the prediction filtered ECG maximum; the second R-wave detection threshold is a second percentage of the prediction filtered ECG maximum; The drive unit system of claim 30, wherein the second percentage is greater than the first percentage.

35. 31. The drive unit system of claim 30, wherein the first and second R-wave detection thresholds are based on the predictive filtered ECG maximum value and an ambient noise value.

36. 36. The drive unit system of claim 35, wherein the system further comprises a noise estimator module operable to generate the ambient noise value based on the filtered ECG signal.

37. The noise estimator module: receiving at least a portion of the filtered ECG signal divided into a plurality of windows; For each window, generate the ambient noise value by finding the maximum value of the filtered ECG signal within each window.

37. The drive unit system of claim 36, operable to:

38. The ambient noise value is an average of one or more of said detected maxima; and a median value of one or more of said detected maxima; and a minimum of one or more of said detected maxima; and 38. The drive unit system of claim 37, wherein:

39. The drive unit system of claim 4 , wherein the drive unit is operable to inflate a pre-inflation balloon pump based on a predicted QS2 timing interval.

40. The drive unit system of claim 39, wherein the predicted QS2 time interval is obtained empirically.

41. 40. The drive unit system of claim 39, wherein the predicted QS2 time interval is specific to the patient.

42. 42. The drive unit system of claim 41, further comprising a QS2 time interval module operable to generate a predicted QS2 time interval based on pressure data obtained via a pressure sensor associated with the drive unit.

43. 43. The drive unit system of claim 42, wherein the predicted QS2 time interval is based on pressure changes caused by the ejection of blood during systole of a partially inflated balloon pump.

44. 40. The drive unit system of claim 39, wherein the drive unit is operable to inflate the balloon based on the conservative detection and the predicted QS2 time interval.

45. 5. The drive unit system of claim 4, wherein the drive unit is operable to deflate the balloon pump 50% at about 100-120 ms of the deflate trigger and to fully deflate the balloon pump at about 140-180 ms of the deflate trigger.

46. The drive unit system of claim 1 , wherein the drive unit comprises a bellows.

47. The drive unit system of claim 5 further comprising an intra-aortic balloon pump.

48. The drive unit system of claim 1 , wherein the cardiac assist device is a ventricular assist device (VAD) operated in a counterpulsation modality.

49. 49. The drive unit system of claim 48, wherein the drive unit is operable to decrease an aortic blood pressure based on the a priori detection and to increase an aortic blood pressure based on the conservative detection.

50. The drive unit according to claim 1 , wherein the cardiac assist device is a ventricular assist device (VAD) operated in a copulsation modality.

51. 51. The drive unit system of claim 50, wherein the drive unit is operable to increase aortic blood pressure relative to an unassisted aortic blood pressure based on the conservative detection.

Images