Medical device and method for detecting change in cardiac pacing evoked response signal

Abstract

A medical device system includes a sensing circuit configured to sense at least one cardiac electrical signal and a therapy delivery circuit configured to deliver pacing pulses. The medical device system includes control circuitry configured to determine evoked response signal features from the at least one cardiac electrical signal sensed following pacing pulses delivered by the therapy delivery circuit. The control circuitry is configured to determine an adaptive threshold based on a center metric of a first portion of the evoked response signal features. The control circuitry may be further configured to detect an evoked response morphology change in response to a second portion of the evoked response signal features being outside the adaptive threshold from the center metric.

Description

MEDICAL DEVICE AND METHOD FOR DETECTING CHANGE IN CARDIACPACING EVOKED RESPONSE SIGNAL

[0001] This application claims the benefit of U.S. Provisional Patent Application Serial No. 63 / 676,195, filed July 26, 2024, the entire content of which is incorporated herein by reference.TECHNICAL FIELD

[0002] This disclosure relates to a medical device and method for detecting a change in a cardiac pacing evoked response signal.BACKGROUND

[0003] During normal sinus rhythm (NSR), the heartbeat is regulated by electrical signals produced by the sino-atrial (SA) node located in the right atrial wall. Each depolarization signal produced by the SA node spreads across the atria, causing the depolarization and contraction of the atria, and arrives at the atrioventricular (AV) node. The AV node responds by propagating a depolarization signal through the bundle of His of the atrioventricular septum and thereafter to the left and right bundle branches and the Purkinje fibers of the right and left ventricles, sometimes referred to as the “His-Purkinje system” and referred to herein as the “conduction system.”

[0004] Patients with poor SA node function, poor AV node conduction (sometimes referred to as AV block), or conduction system abnormalities of the His bundle or left and / or right bundle branches (sometimes referred to as bundle branch block) or other conduction system abnormalities may receive a pacemaker to restore a more normal heart rhythm and heart chamber synchrony. Atrial pacing may be performed to provide a regular atrial rate in a patient having SA node dysfunction. Ventricular pacing may be performed to promote a regular ventricular rate in a patient having AV conduction abnormalities. A single chamber ventricular pacemaker may be coupled to a transvenous ventricular lead carrying electrodes placed in the right ventricle (RV), e.g., in the right ventricular apex. The pacemaker itself is generally implanted in a subcutaneous pocket with the transvenous ventricular lead tunneled to the subcutaneous pocket. Intracardiac pacemakers have been introduced or proposed for implantation entirely within a patient’s heart, eliminating the need for transvenous leads. An intracardiac pacemaker may provide sensing and pacingfrom within a chamber of the patient’s heart, e.g., from within the right ventricle in a patient having AV conduction block.

[0005] Dual chamber pacemaker systems are available which may include a transvenous atrial lead carrying electrodes which are placed in the right atrium and a transvenous ventricular lead carrying electrodes that are placed in the right ventricle via the right atrium. Some leadless dual chamber pacemaker systems have been proposed for implantation within a patient’s heart, without requiring transvenous leads. A dual chamber pacemaker system senses atrial electrical signals and ventricular electrical signals and can provide both atrial pacing and ventricular pacing as needed to promote a normal atrial and ventricular rhythm and promote AV synchrony when SA node, AV node, bundle branch block or other conduction abnormalities are present.SUMMARY

[0006] The techniques of this disclosure generally relate to a medical device and method for detecting a change in a cardiac pacing evoked response (ER) signal. A change in the cardiac pacing ER signal may indicate a change in a pacing electrode location due to lead or electrode dislodgment or shifting, a pathological change in the underlying cardiac tissue substrate, e.g., due to an infarct, or other change in the operative relation between the pacing electrodes and the cardiac tissue being stimulated. The medical device may be an implantable medical device (IMD), e.g., a pacemaker or implantable cardioverter defibrillator (ICD) or cardiac resynchronization therapy (CRT) device, configured to deliver cardiac pacing pulses and sense cardiac electrical signals. The medical device may be connected to leads carrying pacing and sensing electrodes. In other examples, the medical device may be a leadless medical device carrying electrodes for cardiac pacing and cardiac electrical signal sensing on the housing of the medical device.

[0007] A medical device operating according to the techniques disclosed herein may sense an ER signal, from a cardiac electrical signal, following a pacing pulse that captures the paced heart chamber, eliciting the ER signal. One or more features of the ER signal may be determined for updating an adaptive center metric and / or an adaptive variability metric. An adaptive threshold value, range or region can be determined by the medical device from the adaptive center metric and / or the adaptive variability metric. The medical device may compare the adaptive threshold value, range or region at a given time point to the ERsignal feature determined at that time point for detecting a change in the ER signal morphology. The medical device may perform an ER morphology change response when the change in the ER signal morphology is detected, which may include generating an alert or notification transmitted as a communication signal to another medical device and / or adjusting a pacing control parameter, as examples.

[0008] In one example, the disclosure provides a medical device system including a sensing circuit configured to sense a cardiac electrical signal, a therapy delivery circuit configured to deliver pacing pulses and a control circuit configured to, for each of a plurality of morphology monitoring time points, determine an evoked response signal feature from one or more evoked response signals of the cardiac electrical signal sensed after a respective pacing pulse delivered by the therapy delivery circuit. The control circuit may be further configured to determine a center metric of a first portion of the evoked response signal features determined for the plurality of morphology monitoring time points and determine an adaptive threshold based on the center metric. The control circuit may compare the adaptive threshold to a second portion of the evoked response signal features, the second portion of the evoked response signal features different than the first portion of the evoked response signal features, the second portion of the evoked response signal features comprising multiple morphology monitoring time points of the plurality of morphology monitoring time points. The control circuit may detect an evoked response morphology change in response to at least a threshold percentage of the second portion of the evoked response signal features falling outside the adaptive threshold from the center metric. The medical device system may include a communication circuit configured to transmit a notification signal in response to the control circuit detecting the evoked response morphology change.

[0009] In another example, the disclosure provides a method including sensing a cardiac electrical signal, delivering pacing pulses and, for each of a plurality of morphology monitoring time points, determining an evoked response signal feature from one or more evoked response signals of the cardiac electrical signal sensed after a respective delivered pacing pulse. The method may include determining a center metric of a first portion of the evoked response signal features determined for the plurality of morphology monitoring time points, determining an adaptive threshold based on the center metric, and comparing the adaptive threshold to a second portion of the evoked response signal features where thesecond portion of the evoked response signal features is different than the first portion of the evoked response signal features and the second portion of the evoked response signal features includes multiple morphology monitoring time points of the plurality of morphology monitoring time points. The method may further include detecting an evoked response morphology change in response to at least a threshold percentage of the second portion of the evoked response signal features falling outside the adaptive threshold from the center metric. The method may include transmitting a notification signal in response to detecting the evoked response morphology change.

[0010] In another example, the disclosure provides a non-transitory computer readable medium storing instructions that, when executed by control circuitry of a medical device system, cause the medical device system to sense a cardiac electrical signal, deliver pacing pulses and, for each of a plurality of morphology monitoring time points, determine an evoked response signal feature from one or more evoked response signals of the cardiac electrical signal sensed after a respective delivered pacing pulse. The instructions may further cause the system to determine a center metric of a first portion of the evoked response signal features determined for the plurality of morphology monitoring time points, determine an adaptive threshold based on the center metric and compare the adaptive threshold to a second portion of the evoked response signal features, where the second portion of the evoked response signal features is different than the first portion of the evoked response signal features, and the second portion of the evoked response signal features includes multiple morphology monitoring time points of the plurality of morphology monitoring time points. The instructions may further cause the system to detect an evoked response morphology change in response to at least a threshold percentage of the second portion of the evoked response signal features falling outside the adaptive threshold from the center metric and transmit a notification signal in response to detecting the evoked response morphology change.

[0011] The details of one or more aspects of the disclosure are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the techniques described in this disclosure will be apparent from the description and drawings, and from the claims.BRIEF DESCRIPTION OF DRAWINGS

[0012] FIG. l is a conceptual diagram of a medical device system for sensing cardiac electrical signals, delivering cardiac pacing and monitoring for a cardiac pacing ER signal change over time according to some examples.

[0013] FIG. 2 is a conceptual diagram of a leadless IMD that may operate according to the methods disclosed herein in some examples.

[0014] FIG. 3 is a conceptual diagram of a medical device system including leadless IMD implanted at a different cardiac site than the position shown in FIG. 2.

[0015] FIG. 4 is a diagram of a medical device system including two or more coimplanted devices that may perform the methods disclosed herein for detecting a change in the ER signal morphology.

[0016] FIG. 5 is a conceptual diagram of circuitry that may be enclosed within a medical device configured to sense cardiac electrical signals, deliver cardiac pacing, and detect a change in the ER signal morphology according to some examples.

[0017] FIG. 6 is a flow chart of a method for detecting an ER signal morphology change by a medical device system according to some examples.

[0018] FIG. 7 is a flow chart of a method for determining an adaptive threshold applied to ER signal features for detecting ER morphology change according to some examples.

[0019] FIG. 8 is a graph of ER signal morphology match scores (y-axis) that may be determined over time (x-axis) for updating an adaptive threshold and detecting an ER morphology change according to some examples.

[0020] FIG. 9 is a graph of ER signal morphology match scores (y-axis) that may be determined over time (x-axis) for updating an adaptive threshold and detecting an ER morphology change according to another example.

[0021] FIG. 10 is a graph of ER signal morphology features (y-axis) that may be determined over time (x-axis) by a medical device system according to yet another example.

[0022] FIG. 11 is a diagram of plotted ER signal features and an adaptive threshold region that may be determined for detecting an ER morphology change according to another example.

[0023] FIG. 12 is a flow chart of a method for determining an adaptive threshold for detecting an ER morphology change according to yet another example.

[0024] FIG. 13 is a diagram of ER morphology match scores (y-axis) plotted over time (x- axis).

[0025] FIG. 14 is a diagram of the data set shown in FIG. 13 after medical device processing circuitry has completed a clustering method to obtain two clusters of the data points.

[0026] FIG. 15 is a diagram of ER signal morphology feature data points determined during morphology monitoring tests plotted in three dimensions.

[0027] FIG. 16 is a diagram of the projection of the data points shown in FIG. 15 onto the two dimensional y-z plane defined by the ER signal features.DETAILED DESCRIPTION

[0028] A medical device system is disclosed herein that is capable of delivering cardiac pacing, sensing cardiac electrical signals, and monitoring the cardiac pacing ER signal of a sensed cardiac electrical signal for detecting a change in ER signal morphology that may be evidence of a change in the pacing substrate. The pacing substrate refers to the tissue that is in direct or indirect contact with the cardiac pacing electrodes that is captured by the cardiac pacing pulse and / or through which the resulting pacing evoked depolarization is conducted. A change in the pacing substrate can be caused by shifting or movement of one or both pacing electrodes of a pacing electrode vector relative to the cardiac tissue.The pacing electrodes of the pacing electrode vector may be located on the medical device housing and / or carried by a cardiac pacing lead coupled to the medical device. A change in the pacing substrate could be caused by a change in the cardiac tissue, e.g., an infarct or other pathological change. Some tissue changes, such as encapsulation of the pacing electrodes following surgical implantation, may occur gradually over time and are not concerning. However, a sudden change in the morphology of the pacing ER signal may indicate shifting or partial dislodgement of the pacing electrode vector or a sudden change in the cardiac tissue itself. A relatively sudden change in the ER signal, as opposed to a gradual change, may be of greater clinical concern because it may represent partial or total dislodgment of a pacing electrode or a sudden pathological change of the tissue, as examples.

[0029] In some examples, the medical device system performing techniques disclosed herein may be configured to deliver ventricular pacing via the conduction system, referredto herein as “conduction system pacing” (CSP). In this case, at least one electrode of the ventricular pacing electrode vector may be implanted in operative proximity to the conduction system, e.g., in the area of the His bundle, the left bundle branch (LBB), right bundle branch (RBB) and / or Purkinje fibers, for pacing and capturing at least a portion of the conduction system. Myocardial ventricular pacing, via electrodes at or near the right ventricular apex for instance, has been found to be associated with increased risk of atrial fibrillation and heart failure. Delivery of ventricular pacing at sites along the His-Purkinje conduction system for capturing at least a portion of the conduction system may promote a more physiological electrical activation pattern of the heart because the pacing-evoked depolarizations can be propagated along the native conduction system. Pacing the ventricles via the His bundle, the RBB and / or the LBB for example, allows recruitment along the heart’s natural conduction system, including the Purkinje fibers. The techniques disclosed herein, however, are not necessarily limited to a medical device system configured to deliver CSP and may be employed in a medical device delivering cardiac pacing to capture the myocardium.

[0030] When CSP is being delivered, a ventricular pacing pulse that captures the cardiac tissue to cause a pacing ER may capture a portion of the conduction system, myocardial tissue or both. The term “capture” refers to the pacing-evoked depolarization of cardiac tissue propagating through the ventricles and resulting in a QRS complex in the cardiac electrical signal that is caused by a delivered pacing pulse as opposed to an intrinsic depolarization of the cardiac tissue (e.g., arising from the SA node, conducted via the AV node, or arising from an ectopic site) and the associated intrinsic QRS complex that occurs in the cardiac signal when a pacing pulse is not delivered.

[0031] A cardiac pacing ER signal, e.g., the QRS waveform following a ventricular pacing pulse, may have a morphology that can change if the type of capture achieved by the ventricular pacing pulse changes, e.g., conduction system capture without myocardial capture (sometimes referred to as “selective CSP capture”), myocardial only capture without conduction system capture (e.g., loss of conduction system capture), or a combination of both conduction system capture and myocardial capture (sometimes referred to as “non-selective CSP capture”). A change in the position of the pacing electrodes, e.g., relative to a portion of the conduction system, can alter the type of capture being achieved resulting in a change in the pacing ER signal. The techniques disclosedherein provide methods for detecting a relatively sudden change in the cardiac pacing ER signal, which may be an indication of a change in the pacing substrate, e.g., due to a shift or change in pacing electrode location or other pacing substrate change.

[0032] FIG. 1 is a conceptual diagram of a medical device system 10 for sensing and analyzing cardiac electrical signals and delivering cardiac pacing according to some examples. Medical device system 10 includes an IMD 14 connected to an atrial lead 16 and a ventricular lead 18 in this example. IMD 14 includes a housing 15, which may be hermetically sealed, to enclose internal circuitry corresponding to the various circuits and components for sensing cardiac signals from heart 8 and controlling cardiac pacing delivered to heart 8 by IMD 14. The housing 15 may be formed of a conductive material, such as titanium or titanium alloy. The housing 15 may function as an electrode (sometimes referred to as a “can” electrode). In some examples, housing 15 may be available as a return anode electrode for delivering unipolar pacing pulses and / or for use in a sensing electrode vector for sensing cardiac electrical signals in combination with electrodes carried by lead 16 and / or lead 18.

[0033] IMD 14 includes a connector assembly 13 (sometimes referred to as a “connector block” or “header”), coupled to housing 15, having connector bores configured to receive the proximal lead connectors (not shown) of atrial lead 16 and ventricular lead 18. Connector block 13 may have one or more additional connector bores for receiving one or more additional leads, e.g., for receiving a coronary sinus lead for providing pacing and sensing in the left ventricle (LV) of heart 8 from a location in a cardiac vein along the left lateral free wall in some examples.

[0034] Atrial lead 16 is shown advanced transvenously into the right atrial chamber of a patient’s heart 8 for sensing atrial signals, e.g., P-waves attendant to atrial depolarizations, and for delivering atrial pacing pulses. Atrial lead 16 includes pacing and sensing electrodes 20 and 22. Electrode 20 is shown as a screw-in, helical tip electrode at the distal end of atrial lead 16. Electrode 22 is shown as a ring electrode (e.g., circumscribing the atrial lead body 17) spaced proximally from tip electrode 20. Electrodes 20 and 22 can form a bipolar pair for sensing atrial signals and delivering atrial pacing pulses via tip electrode 20 as a cathode electrode and ring electrode 22 as the return anode electrode, for example. Atrial lead 16 includes an elongated lead body 17 through which insulated electrical conductors extend from the respective electrodes 20 and 22 to the proximal leadconnector (not shown) connected to the IMD 14 via connector assembly 13. The electrodes 20 and 22 are thereby connected to internal electronics of IMD 14 via respective electrical feedthroughs in connector assembly 13 that cross pacemaker housing 15.

[0035] Ventricular lead 18 is shown advanced transvenously into the right atrial chamber of a patient’s heart 8 and further into the right ventricle (RV) for positioning tip electrode 32 within the interventricular septum 12 in the vicinity of the heart’s conduction system, e.g., at a His bundle pacing site, an LBB area pacing (LBBAP) site or at an RBB area pacing (RBBAP) site. Ventricular lead 18 is positioned for sensing ventricular event signals, e.g., R-waves attendant to intrinsic depolarizations of the ventricular myocardium and pacing evoked ventricular depolarizations, and for delivering ventricular pacing pulses, e.g., CSP pulses.

[0036] Ventricular lead 18 includes pacing and sensing electrodes 32 and 34. Electrode 32 is shown as a screw-in, helical tip electrode at the distal end of ventricular lead 18. Electrode 34 is shown as a ring electrode spaced proximally from tip electrode 32 and circumscribing ventricular lead body 19. Electrodes 32 and 34 can form a bipolar pair for sensing ventricular signals and delivering ventricular pacing, e.g., as CSP pulses or myocardial pacing pulses, via tip electrode 32 as a cathode electrode and ring electrode 34 as the return anode electrode, for example. While the electrodes 20, 22, 32 and 34 are represented as either helical screw-in electrodes or ring electrodes in FIG. 1, it is to be understood that other electrode types may be used such as button electrodes, hook electrodes, segmented electrodes, short coil electrodes, or the like. Tip electrode 32 of ventricular lead 18 may be a tissue-piercing electrode, which may or may not have a helical shape, to facilitate advancement of tip electrode 32 into the interventricular septum 12 to a CSP site in the area of the His bundle, LBB or RBB.

[0037] In some examples, IMD 14 may be capable of delivering cardioversion / defibrillation (CV / DF) shocks for treating ventricular tachyarrhythmias. In this case, ventricular lead 18 (and / or another lead coupled to IMD 14) may carry one or more coil electrodes 36 and 38 for use in delivering high voltage CV / DF shocks. As such, while IMD 14 can be referred to as a “pacemaker,” capable of delivering cardiac pacing therapies, IMD 14 could be referred to as “implantable cardioverter defibrillator” or “ICD”when capable of delivering high voltage CV / DF shocks in addition to the pacing functionality as disclosed herein.

[0038] Ventricular lead 18 includes an elongated lead body 19 through which insulated electrical conductors extend from the respective electrodes 32 and 34 (and coil electrodes 36 and 38 if present) to a proximal lead connector (not shown) connected to IMD 14 via connector assembly 13. The electrodes 32 and 34 (and coil electrodes 36 and 38 if present) are thereby connected to internal electronics of IMD 14 via respective electrical feedthroughs in connector assembly 13 that cross housing 15.

[0039] Electrodes 20, 22, 32, 34 (and 36 and 38 if present) may be formed from titanium, platinum, iridium or alloys thereof, as examples with no limitation intended, and may include a low polarizing coating, such as titanium nitride, iridium oxide, ruthenium oxide, platinum black, among others. Lead bodies 17 and 19 may each be formed from a non- conductive material, including silicone, polyurethane, fluoropolymers, or mixtures thereof, as examples. Each lead body may be shaped to form one or more lumens within which one or more insulated electrical conductors extend between the proximal lead connectors and the distal electrodes 20, 22, 32, 34, 36 and 38 carried by the respective lead 16 or 18.

[0040] While ventricular lead 18 is shown advanced into the RV for positioning tip electrode 32 in the interventricular septum 12, e.g., in the area of the LBB or the RBB for delivering CSP, it is to be understood that the distal end of ventricular lead 18 may be positioned at other locations for delivering CSP to heart 8 for causing depolarizations of the tissue of the conduction system thereby pacing the ventricles. For instance, ventricular lead tip electrode 32 may be positioned along or in the area of the His bundle and / or the RBB or LBB in a basal portion of the interventricular septum 12. In other examples, ventricular lead 18 may be advanced into the right atrium with tip electrode 32 advanced into the interatrial septum toward the His bundle, e.g., at the inferior end of the interatrial septum. Ventricular lead 18 may be advanced toward the His bundle from a location in the right atrium beneath the AV node and near the tricuspid valve annulus, generally in the Triangle of Koch, to position ventricular lead tip electrode 32 near the His bundle for delivering CSP for pacing the ventricles from a right atrial approach.

[0041] The techniques disclosed herein are not limited to a particular pacing location, however, and may be practiced in a variety of medical device systems including cardiac pacing and sensing capabilities. Various ventricular pacing sites can include a myocardialpacing site, e.g., the RV apex, the left ventricular lateral free wall, interventricular septum, CSP site, etc. A pacing electrode positioned at a pacing site may be carried by a medical electrical lead, as shown in the example of FIG. 1, or along the housing of a leadless IMD, e.g., as described below in conjunction with FIGs. 2 and 3.

[0042] IMD 14 includes therapy delivery circuitry for generating pacing pulses delivered via the atrial lead 16 and ventricular lead 18. As described below, cardiac electrical signal sensing circuitry included in IMD 14 may receive a cardiac electrical signal sensed from electrodes carried by atrial lead 16 and a cardiac electrical signal sensed from electrodes carried by ventricular lead 18 for use in controlling the timing and delivery of atrial pacing pulses and ventricular pacing pulses. As described below, a sensed cardiac electrical signal received by IMD 14 may be analyzed by processing circuitry of IMD 14 for detecting ER signal morphology changes following delivered pacing pulses. Morphology monitoring tests may be performed for detecting a change in the pacing substrate, e.g., due to a change in the pacing electrode location relative to cardiac tissue or a pathological change in the cardiac tissue.

[0043] While IMD 14 is shown as a dual chamber pacemaker receiving both atrial lead 16 and ventricular lead 18, it is to be understood that in other examples, IMD 14 may be a single chamber device, e.g., configured to receive one lead for sensing cardiac electrical signals and delivering ventricular pacing without necessarily having atrial pacing capabilities. In still other examples, IMD 14 may be a multi -chamber pacemaker configured to sense cardiac signals and deliver atrial pacing via atrial lead 16, ventricular pacing (e.g., CSP) via ventricular lead 18, and left ventricular pacing via a third lead that may be advanced via the coronary sinus ostium of the right atrium into the coronary sinus and further into a cardiac vein to a left ventricular pacing site, e.g., for providing CRT.

[0044] Medical device system 10 is shown including an external medical device 50 for receiving data from IMD 14 and for transmitting programming commands to IMD 14, which may include various sensing and pacing control parameters used by IMD 14. External device 50 may receive data from IMD 14 including sensed cardiac electrical signal episodes, therapy delivery data logged by IMD 14 and results of ER morphology analysis performed by IMD 14 according to the techniques disclosed herein.

[0045] External device 50 may be embodied as a programmer used in a hospital, clinic or physician’s office to program IMD 14 and to acquire data from IMD 14. External device50 may alternatively be embodied as a handheld device, such as a tablet or cell phone. In some examples, external device 50 is a home monitor configured to interrogate IMD 14 to receive signals or data from IMD 14 and transmit data to IMD 14 via a wireless communication link 48. An example programmer that may be configured to program IMD 14 and included in medical device system 10 configured to perform the techniques disclosed herein is the CARELINK® Programmer, commercially available from Medtronic, Inc., Dublin, Ireland.

[0046] External device 50 may include a processor 52, memory 53, display unit 54, user interface unit 56, and telemetry unit 58. Processor 52 is coupled to the other components and units of external device 50, e.g., via a data bus, for controlling the functions attributed to external device 50 herein. Processor 52 may execute instructions stored in memory 53. Processor 52 may include any one or more of a microprocessor, a controller, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field- programmable gate array (FPGA), or equivalent discrete or analog logic circuitry. In some examples, processor 52 may include multiple components, such as any combination of one or more microprocessors, one or more controllers, one or more DSPs, one or more ASICs, or one or more FPGAs, as well as other discrete or integrated logic circuitry. The functions attributed to processor 52 herein may be embodied as software, firmware, hardware or any combination thereof.

[0047] Memory 53 may include any volatile, non-volatile, magnetic, optical, or electrical media, such as a random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), electrically-erasable programmable ROM (EEPROM), flash memory, or any other digital or analog media. Memory 53 may include non-transitory computer- readable media that may store instructions that, when executed by processor 52, cause medical device system 10 to perform various methods and functions attributed to medical device system 10 as disclosed herein.

[0048] User interface unit 56 may include a mouse, touch screen, keypad or the like to enable a user to interact with external device 50, e.g., to initiate and terminate an interrogation session for retrieving data from IMD 14, adjust settings of display unit 54, enter programming commands or selections or make other user requests. Display unit 54, which may include a liquid crystal display, light emitting diodes (LEDs) and / or other visual display components, may generate a display of cardiac electrical signals receivedfrom IMD 14 and / or data derived therefrom, including results of capture management tests performed by IMD 14. Display unit 54 may be configured to generate a graphical user interface (GUI) including various windows, icons, user selectable menus, etc. to facilitate interaction by a user with the external device 50. Display unit 54 may display various windows to a user, e.g., in a GUI, for enabling a clinician or other user to review cardiac signal and therapy delivery related data retrieved from IMD 14.

[0049] Display unit 54 may function as an input and / or output device using technologies including liquid crystal displays (LCD), quantum dot display, dot matrix displays, light emitting diode (LED) displays, organic light-emitting diode (OLED) displays, cathode ray tube displays, e-ink, or monochrome, color, or any other type of display capable of generating tactile, audio, and / or visual output. In some examples, display unit 54 is a presence-sensitive display. Display unit 54 may serve as a user interface device that operates both as one or more input devices and one or more output devices.

[0050] External device 50 may receive data, via telemetry unit 58, from IMD 14 via the wireless communication link 48. Data received from IMD 14 may include cardiac signals, sensed by IMD 14, marker channel data indicating the timing of pacing pulses delivered by IMD 14 and sensed cardiac event signals (e.g., sensed P-waves and sensed R-waves), data relating to the pacing history, and data relating to ER morphology analysis as further described below.

[0051] Telemetry unit 58 includes a transceiver and antenna configured for bidirectional communication with a communication circuit included in an implantable IMD 14. Telemetry unit 58 includes communication circuitry configured to operate in conjunction with processor 52 for sending and receiving data relating to pacemaker functions via a wireless communication link 48 with the implantable IMD 14. Communication link 48 may be established using a radio frequency (RF) link such as BLUETOOTH®, Wi-Fi, Medical Implant Communication Service (MICS) or other communication bandwidth. In some examples, external device 50 may include a programming head that is placed proximate IMD 14 to establish and maintain communication link 48, and in other examples external device 50 and IMD 14 may be configured to communicate using a distance telemetry algorithm and circuitry that does not require the use of a programming head and does not require user intervention to maintain a communication link 48.

[0052] It is contemplated that external device 50 may be in wired or wireless connection to a communications network via telemetry unit 58 that includes a transceiver and antenna or via a hardwired communication line for transferring data to a centralized database or computer to allow remote management of the patient. External device telemetry unit 58 may be coupled to a communication network / cloud (not shown in FIG. 1) for receiving and transmitting data to a remote computing device (not shown in FIG. 1), which may be a personal computer, personal mobile device or other computing device at a remote location from the patient to enable remote monitoring of data obtained from IMD 14 by a clinician or other user. The CARELINK™ network available from Medtronic, Inc., Dublin, Ireland, is an example of a remote patient monitoring system and database that may collect and display data retrieved from a patient’s pacemaker for review by a clinician or other user. Processing circuitry of the medical device system 10, e.g., any combination of one or more of external device processor 52, network / cloud based processing, a remote computing device and / or processing circuitry included in IMD 14 (see FIG. 4), may be configured to execute portions of the methods disclosed herein for detecting an ER morphology change, which may include establishing a morphology template used for determining match scores between the template and post-pace ER signal waveforms for the purposes of monitoring for a change in pacing ER signal morphology.

[0053] The ER signal morphology change can be detected in the ambulatory patient to allow recordation of cardiac signals representative of the change in the ER signal and / or other responses to be performed upon detection of the ER signal morphology change by the IMD. Long term monitoring of ER signals for detecting a sudden change in morphology provides specific improvements in the field of medical devices that have practical applications. By providing a medical device system configured to detect a sudden ER signal morphology change, the IMD or a clinician can perform a remedial response to the sudden change that may otherwise go undetected. Improved processor-based methods for detecting a change in ER signal morphology as disclosed herein reduces the likelihood of a change in the pacing substrate, e.g., due to a shift in electrode location or a pathological change, going undetected and untreated in a patient and reduces the likelihood of human error in analyzing and evaluating the cardiac electrical signals for detecting such a change.

[0054] FIG. 2 is a conceptual diagram of a leadless IMD 114 that may operate according to the methods disclosed herein in some examples. The IMD 114 may be positioned within the right atrium for providing ventricular pacing by delivering CSP in the area of the His bundle. IMD 114 may deliver ventricular pacing, via the conduction system, and sense cardiac electrical signals for monitoring for ER morphology changes over time as further described below.

[0055] IMD 114 may include a distal tip electrode 132 extending away from a distal end 112 of the IMD housing 115. Leadless IMD 114 is shown implanted in the right atrial chamber of the patient’s heart for advancing distal tip electrode 132 to a His bundle pacing site from a right atrial approach. For example, the distal tip electrode 132 may be inserted into the inferior end of the interatrial septum, beneath the AV node and near the tricuspid valve annulus to position tip electrode 132 at a His bundle pacing site. In other examples, leadless IMD 114 may be implanted within the right ventricle, e.g., high along the interventricular septum, for positioning distal tip electrode 132 in the basal portion of the interventricular septum in the vicinity of the His bundle or in another interventricular septal location along the His-Purkinje system.

[0056] Distal tip electrode 132 may be a helical electrode providing fixation to anchor the IMD 114 at the implant position. In other examples, IMD 114 may include a fixation member that includes one or more tines, hooks, barbs, helices or other fixation member(s) that anchor the distal end of the IMD 114 at the implant site. A proximal portion of the distal tip electrode 132 may be electrically insulated such that only the most distal end of tip electrode 132, furthest from housing distal end 112, is exposed to provide targeted pacing at a tissue site that includes a portion of the conduction system.

[0057] One or more housing-based electrodes 120 and 134 may be carried on the surface of the housing 115 of IMD 114, on or proximal to distal end 112. Electrodes 120 and 134 are shown as ring electrodes circumscribing the longitudinal sidewall of housing 115 that extends from the housing distal end 112 to housing proximal end 110. In other examples, a return anode electrode used in sensing and pacing with distal tip electrode 132 may be positioned on housing proximal end 110. Ventricular pacing via the His-Purkinje system may be achieved using the distal tip electrode 132 as the cathode electrode and either of the housing-based electrodes 120 or 134 as the return anode. In some examples, pacing of atrial tissue may be achieved by delivering atrial pacing pulses via the distal ring electrode120 using proximal ring electrode 134 as the return anode electrode. In other examples, distal ring electrode 120 shown circumscribing the lateral sidewall of housing 115 for the sake of clarity in FIG. 2 may be located on the pacemaker distal end 112 for providing atrial pacing and atrial sensing, e.g., in combination with proximal ring electrode 134. CSP pulses may be delivered in the area of the His bundle via tip electrode 132 with proximal ring electrode 134 as the return anode. In this way, dual chamber pacing of the atria and the ventricles may be delivered by leadless IMD 114. In other examples more than two ring electrodes (or other types of electrodes) may be provided on housing 115, e.g., to provide two distinct atrial pacing and CSP electrode vectors.

[0058] Intrinsic and pacing evoked cardiac electrical signals may be sensed by IMD 114 using one or more sensing electrode pairs selected from electrodes 120, 132 and 134. For example, a ventricular electrical signal may be sensed using distal tip electrode 132 and distal ring electrode 120 or proximal ring electrode 134. Intrinsic R-waves may be sensed by sensing circuitry of IMD 114 via the ventricular electrical signal sensing electrode pair for use in inhibiting a scheduled CSP pulse and scheduling the next CSP pulse. Post-pace ER signal waveforms of the ventricular electrical signal may be acquired by IMD 114 for monitoring for changes in the ER signal morphology as further described below.

[0059] An atrial electrical signal may be sensed using electrodes 120 and 134. Intrinsic P- waves may be sensed by the atrial electrical signal sensing electrode pair for use in inhibiting and scheduling atrial pacing pulses and / or for scheduling atrial synchronous CSP pulses. The cardiac electrical signals sensed by IMD 114 may be used for determining the atrial rate, ventricular rate, and / or for detecting atrial and / or ventricular tachyarrhythmias.

[0060] While external device 50 is not shown in FIG. 2, it is to be understood that IMD 114 can be included in a medical device system including external device 50 configured to communicate with IMD 114 and in some cases, as described above, with a network / cloud based patient database and / or a remote computing device. IMD 114 may be configured to communicate via a communication circuit with external device 50 for receiving programming commands and transmitting data to external device 50 as generally described above in conjunction with FIG. 1. Leadless IMD 114 may be configured to transmit data to external device 50, which may be further analyzed by external device processor 52 and / or used in generating a graphical user interface of data acquired by IMD114 for review by a clinician, including data and ER morphology monitoring test results obtained by performing post-pace ER morphology analysis according to the techniques described below.

[0061] FIG. 3 is a conceptual diagram of a medical device system 200 including leadless IMD 114 implanted at a different CSP site than the position shown in FIG. 2. In this example, tip electrode 132 may be advanced into the interventricular septum 12 from a right ventricular approach for delivering ventricular pacing as CSP pulses in the area of the LBB or the area of the RBB. In some cases, the medical device system 200 may include a second leadless IMD 214 implanted in the right atrium for delivering atrial pacing pulses and sensing atrial electrical signals. In this example, atrial IMD 214 includes a tip electrode 232 that may be paired with a proximal ring electrode 234 circumscribing the lateral sidewall of cylindrical housing 215 of atrial IMD 214. Tip electrode 232 is shown as a non-tissue piercing button electrode in this example but may be provided as a tissue piercing or non-tissue piercing electrode and may be any of the types of example electrodes listed herein. Atrial IMD 214 may include a fixation member 213, e.g., as one or more fixation tines, extending from distal end 212 of IMD housing 215 to provide fixation of atrial IMD 214 at an atrial pacing site.

[0062] In some examples, atrial IMD 214 can provide atrial pacing and sensing and ventricular IMD 114 can provide ventricular pacing (via the conduction system and / or ventricular myocardium) and sensing in a dual chamber leadless pacemaker system 200. Atrial IMD 214 and ventricular IMD 114 may communicate wirelessly, as shown by arrow 218, to coordinate dual chamber pacing delivery. For example, atrial IMD 214 may transmit a signal to ventricular IMD 114 when an atrial pacing pulse is delivered or an atrial P-wave is sensed so that ventricular IMD 114 can deliver an atrial synchronous CSP pulse at a desired atrioventricular pacing interval. In other examples, ventricular IMD 114 may sense atrial systolic event signals, e.g., from an electrical signal sensed by IMD 114 or from an acceleration signal sensed by an accelerometer included in IMD 114, for use in synchronizing CSP pulses to the atrial event signals. While not shown in FIG. 3, it is to be understood that external device 50 (shown in FIG. 1), which may be in communication with a network / cloud and / or remote computing device, may be included in the medical device system 200 and may be configured to send data to and receive data from ventricular IMD 114 and atrial IMD 214, if present.

[0063] While several examples of medical device systems are shown and described in conjunction with FIGs. 1 — 3, it is to be understood that the methods disclosed herein for performing ER morphology monitoring as described below are not limited to a particular medical device system having cardiac signal sensing and pacing capabilities. The methods disclosed herein may be practiced in any medical device system that includes or is in communication with a medical device that is capable of sensing or receiving cardiac electrical signals sensed following delivered pacing pulses for analyzing post-pace ER signals, e.g., pacing evoked QRS waveforms.

[0064] The methods for analyzing ER signals and detecting a change in an ER signal morphology may be performed by the medical device that is delivering the cardiac pacing that causes cardiac capture and the pacing-evoked depolarizations observed as the ER signal, e.g., IMD 14 or IMD 114. In other examples, a medical device that is not delivering the cardiac pacing that elicits the ER signals could sense a cardiac electrical signal and analyze the post-pace ER signals for detecting a change in the ER signal morphology using the techniques disclosed herein. For instance, in the example of FIG. 3, atrial IMD 214 could sense an EGM signal that includes far field R- waves and may analyze the postpace QRS signals for detecting a change in the pacing ER signals according to the techniques disclosed herein. Other examples of medical devices systems that include one device delivering cardiac pacing pulses and a second, different device performing the ER morphology analysis methods disclosed herein for detecting a change in the pacing substrate, e.g., due to a change in pacing electrode location or a pathological cardiac tissue change, are described in conjunction with FIG. 4.

[0065] FIG. 4 is a diagram of a medical device system 250 including two or more coimplanted devices that may perform the methods disclosed herein for detecting a change in the ER signal morphology. Medical device system 250 is shown including IMD 114 for providing cardiac pacing to the patient’s heart 8. IMD 114 may be implanted in or on the patient’s heart 8, in any of the example locations described above. IMD 114 may deliver ventricular pacing pulses at a selected pacing site, e.g., a myocardial pacing site or a CSP site, for capturing and pacing the ventricles.

[0066] Medical device system 250 may include a cardiac monitoring device 252 coimplanted with IMD 114 in patient 6 for monitoring cardiac signals. Cardiac monitoring device 252 may include a housing 253 and two or more electrodes 254 and 256 carried bythe housing 253 for sensing electrocardiogram (ECG) signals. Cardiac monitoring device 252 may generally correspond to the LINQ ™ Implantable Cardiac Monitor available from Medtronic, Inc., Dublin, Ireland, as an example. Cardiac monitoring device 252 may be configured to perform the methods disclosed herein for analyzing ER signal features of the ECG signal to monitor for changes in the ER signal morphology following pacing pulses delivered by IMD 114.

[0067] Cardiac monitoring device 252 may sense an ECG signal during pacing delivered by IMD 114 and detect a change in the ER signal morphology as further described below. In response to detecting an ER signal morphology change, cardiac monitoring device 252 may transmit a communication signal to external device 50, via a communication link 255 (as generally described above in conjunction with FIG. 1). The transmitted communication signal may be a notification or alert of the detected change and may include an ECG signal episode including the ER signal detected as having a change in morphology.

[0068] Cardiac monitoring device 252 may be configured to sense an ECG signal via electrodes 254 and 256 at a specified time of day. IMD 114 may be configured to deliver pacing pulses at the same specified time of day in a manner that captures the heart 8 to elicit pacing ER signals that can be sensed and analyzed by cardiac monitoring device 252. In some cases, cardiac monitoring device 252 may include pulse detection circuitry for detecting the cardiac pacing pulses delivered by IMD 114 and may sense the post-pace ECG signals for analyzing the ER signals following detected pacing pulses.

[0069] In other examples, cardiac monitoring device 252 and IMD 114 may be configured to communicate via the wireless communication link 258, e.g., via any of the RF communication methods described above in conjunction with FIG. 1 or via tissue conduction communication (TCC) or another communication method. Cardiac monitoring device 252 and IMD 114 may be configured to cooperatively perform morphology monitoring tests. Cardiac monitoring device 252 may receive a signal from IMD 114 indicating that one or more pacing pulses are being (or scheduled to be) delivered for capturing the heart so that ER signal sensing and analysis can be performed by cardiac monitoring device 252.

[0070] In response to detecting the change in the ER signal morphology, cardiac monitoring device 252 may transmit a notification signal to IMD 114 via communication link 258. IMD 114 may respond to the notification signal by adjusting a pacing controlparameter. As further described below, a pacing electrode vector, pacing pulse output, pacing mode or other pacing control parameter may be adjusted if an ER signal morphology change is detected by the medical device system.

[0071] Medical device system 250 is further shown to include an ICD 260 coupled to lead 270. ICD 260 includes a housing 265 enclosing internal circuitry for sensing cardiac electrical signals and generating electrical stimulation pulses, e.g., CV / DF shock pulses, anti -tachycardia pacing pulses, bradycardia pacing pulses etc. For these purposes, lead 270 may carry one or more electrodes that are electrically coupled to internal circuitry of ICD 260 via connector block 267. A lead connector (not shown) at proximal lead end 281 may be received by connector block 267 for making electrical connection between contacts of the lead connector that are electrically coupled to insulated conductors extending through one or more lumens of lead body 280 to the electrodes carried by lead 270.

[0072] Lead 270 may be advanced subcutaneously, submuscularly or substemally to a location over heart 8 for sensing cardiac signals and delivering electrical stimulation pulses. In some examples, lead 270 may be advanced transvenously to a venous location outside of heart 8 for sensing cardiac signals and delivering electrical stimulation pulses. For instance, distal portion 275 of lead 270 may be advanced to a location beneath rib cage 9 and / or sternum 7, e.g., generally in the anterior mediastinum in front of heart 8 and beneath sternum 7.

[0073] In the example shown, a distal portion 275 of lead body 280 carries two ring electrodes 272 and 274 and two coil electrodes 276 and 278 though other electrode arrangements and combinations are possible. The ring electrodes 272 and 274 may be used in a sensing electrode vector and / or for delivering relatively low voltage electrical stimulation pulses to heart 8. The coil electrodes 276 and 278 can be used for delivering high voltage CV / DF shock pulses and can be used in sensing electrode vectors. ICD housing 265 may be an active can electrode used in delivering CV / DF shock pulses in combination with one or both of coil electrodes 276 and 278.

[0074] Various sensing electrode vectors may be switchably coupled to sensing circuitry in housing 265 for sensing an ECG signal that may be analyzed by processing circuitry of ICD 260 according to the techniques disclosed herein for detecting a change in the ER signal morphology. In some examples, IMD 114 may be delivering ventricular pacingpulses causing ER signals that are sensed and analyzed by ICD 260 for detecting a change in morphology as described below.

[0075] ICD 260 may be configured to sense an ECG signal at a specified time of day when IMD 114 is scheduled to deliver one or more pacing pulses in a manner that captures the ventricles causing ER signals that can be analyzed by ICD 260 for detecting a change in ER morphology. In some examples, ICD 260 and IMD 114 may be configured to communicate via communication link 292. ICD 260 may transmit a command to IMD 114 to deliver pacing pulses for a morphology monitoring test. Alternatively, IMD 114 may transmit a communication signal to ICD 260 indicating that pacing pulses are being (or scheduled to be) delivered for performing the morphology monitoring test. In still other examples, ICD 260 may include pulse detection circuitry for detecting pacing pulses delivered by IMD 114. ICD 260 (or cardiac monitoring device 2502) may detect pacing pulses being delivered at a specified rate that is an indication that the morphology monitoring test is in progress and an ER signal morphology analysis is to be performed.

[0076] Accordingly, a variety of methods may be implemented for coordinating cardiac electrical signal sensing and ER signal analysis by a second device that is co-implanted with an IMD that is delivering the pacing pulses causing the ER signals. While patient 6 is shown having both a co-implanted cardiac monitoring device 252 and co-implanted ICD 260 with IMD 114, all three devices are shown for the sake of providing illustrative examples of a combination of two co-implanted devices cooperatively performing the techniques disclosed herein. The patient need not have all three devices and other combinations of co-implanted HMDs may be configured to cooperatively perform the ER signal morphology monitoring techniques disclosed herein. Furthermore, while ICD 260 is described as performing signal sensing and analysis for detecting an ER signal morphology change when IMD 114 is delivering pacing pulses, in some examples, ICD 260 could be delivering ventricular pacing pulses via lead 270 and sensing the resulting ER signals for detecting an ER signal morphology change or another device, e.g., cardiac monitoring device 252 may sense and analyze the ER signals.

[0077] ICD 260 can be configured for bi-directional communication with external device 50, via communication link 290, as generally described above in conjunction with FIG. 1. When ICD 260 is configured to detect an ER signal morphology change, ICD 260 may transmit a notification to external device 50, which may include an ECG signal episode.Additionally or alternatively, ICD 260 may transmit a notification to IMD 114 via communication link 292. IMD 114 may adjust a pacing control parameter in response to the notification that a change in the ER signal morphology has been detected.

[0078] FIG. 5 is a conceptual diagram of circuitry that may be enclosed within a medical device configured to sense cardiac electrical signals, deliver cardiac pacing, and detect a change in the ER signal morphology according to some examples. FIG. 5 is described with reference to IMD 14 connected to atrial lead 16 carrying electrodes 20 and 22 and ventricular lead 18 carrying electrodes 32 and 34 of FIG. 1 for the sake of convenience. Housing 15 is depicted in FIG. 5 as an electrode coupled to IMD circuitry, e.g., for use in a unipolar pacing and / or sensing electrode vector. It is to be understood, however, that the functionality attributed to the various circuits and components shown in FIG. 5 may correspond to circuitry of a medical device system configured to perform techniques disclosed herein, which may include one IMD configured to pace and sense cardiac electrical signals for detecting a change in ER signal morphology or two or more coimplanted devices and / or an external device configured to cooperatively pace and sense cardiac electrical signals for detecting a change in ER signal morphology.

[0079] The electronic circuitry enclosed within housing 15 (depicted as an electrode in FIG. 5) includes software, firmware and hardware that cooperatively monitor cardiac electrical signals, determine when a pacing pulse is necessary, and deliver electrical pacing pulses to the patient’s heart as needed according to a programmed pacing mode and other pacing control parameters. The electronic circuitry may include a control circuit 80, memory 82, therapy delivery circuit 84, sensing circuit 86, communication circuit 88 and power source 98. In some examples, IMD 14 may include one or more other sensors 90 for sensing physiological signals. For instance, IMD 14 may include a motion sensor such as an accelerometer for sensing patient physical activity, monitoring patient posture, and / or sensing cardiac motion (e.g., when IMD 114 is implanted within the heart such as in the examples of FIGs. 2-4). Other examples of physiological sensors that may be included in IMD 14 include heart sound sensors, pressure sensors, temperature sensors, oxygen sensors, or the like.

[0080] Power source 98 provides power to the circuitry of IMD 14 including each of the components 80, 82, 84, 86, 88 and 90 as needed. Power source 98 may include one or more energy storage devices, such as one or more rechargeable or non-rechargeablebatteries. The connections between power source 98 and each of the other components 80, 82, 84, 86, 88 and 90 are to be understood from the general block diagram of FIG. 5 but are not shown for the sake of clarity. For example, power source 98 may be coupled to one or more charging circuits included in therapy delivery circuit 84 for providing the power needed to charge holding capacitors included in therapy delivery circuit 84 that are discharged at appropriate times under the control of control circuit 80 for delivering pacing pulses. Power source 98 is also coupled to components of sensing circuit 86 (such as sense amplifiers, analog-to-digital converters, switching circuitry, etc.), communication circuit 88, sensors 90 and memory 82 to provide power to the various components and circuits as needed.

[0081] The components shown in FIG. 5 represent functionality included in IMD 14 and may include any discrete and / or integrated electronic circuit components that implement analog and / or digital circuits capable of producing the functions attributed to IMD 14 herein. The various components may include an application specific integrated circuit (ASIC), an electronic circuit, a processor (shared, dedicated, or group) and memory that execute one or more software or firmware programs, a combinational logic circuit, state machine, or other suitable components or combinations of components that provide the described functionality. Providing software, hardware, and / or firmware to accomplish the described functionality in the context of any modern medical device system, given the disclosure herein, is within the abilities of one of skill in the art.

[0082] Control circuit 80 communicates, e.g., via a data bus, with therapy delivery circuit 84 and sensing circuit 86 for cooperatively sensing cardiac electrical signals and controlling delivery of cardiac electrical stimulation pulses in response to sensed cardiac event signals (or absence thereof), e.g., R-waves attendant to ventricular depolarization and / or P-waves attendant to atrial depolarization, in accordance with a pacing mode. Electrodes 20, 22, 32 and 34 and housing 15 may be electrically coupled to therapy delivery circuit 84 for delivering electrical stimulation pulses generated by therapy delivery circuit 84. Electrodes 20, 22, 32 and 34 and housing 15 may be electrically coupled to sensing circuit 86 for sensing cardiac electrical signals produced by the heart. Sensing circuit 86 may sense intrinsic signals (such as intrinsic P-waves and intrinsic R- waves) produced by the heart in the absence of a pacing pulse that captures the heart.Sensing circuit 86 may sense ER signals, e.g., pacing-evoked P-waves and pacing evoked R-waves, following a delivered pacing pulse of sufficient energy to cause cardiac capture.

[0083] As further described below, post-pace ER signals may be analyzed for detecting a change in the ER signal morphology. The pacing pulses may be delivered having a pulse energy that is equal to or greater than the capture threshold, which is the lowest pacing pulse energy that captures the cardiac tissue, in order to promote capture of the cardiac tissue and a pacing evoked QRS waveform that can be analyzed to detect an ER morphology change. While examples presented herein refer to analysis of the ventricular ER signal morphology, sensed following delivery of a ventricular pacing pulse, it is contemplated that the methods disclosed herein may be applied to an analysis of the ER signal morphology following delivery of an atrial pacing pulse for detecting a change in pacing-evoked P-wave morphology due to a change in the pacing substrate, e.g., a shift or dislodgment of an atrial pacing electrode or a pathological change of the atrial tissue.

[0084] Sensing circuit 86 may include an input pre-filter and amplifier 140 for receiving a cardiac electrical signal from a bipolar pair of sensing electrodes, e.g., atrial lead electrodes 20 and 22 and / or ventricular lead electrodes 32 and 34. Sensing circuit 86 may receive a unipolar or far field cardiac electrical signal, e.g., between tip electrode 32 and housing 15, between a coil electrode 36 or coil electrode 38 (shown in FIG. 1) and housing 15, or between tip electrode 32 or ring electrode 34 and either of coil electrodes 36 and 38 (shown in FIG. 1), as examples. As such, sensing circuit 86 may include multiple sensing channels and may include switching circuitry for switchably connecting sensing electrode vector pairs selected from the available electrodes coupled to IMD 14 (and housing 15).

[0085] The filtered and amplified signal output by input pre-filter and amplifier 140 may be passed to an analog-to-digital converter and wide bandpass filter (ADC) 141 for producing a multi-bit digital cardiac electrical signal that may be passed to control circuit 80 and is referred to herein as a cardiac electrogram or “EGM” signal when the raw signal is sensed from electrodes on or within a heart chamber. The EGM signal sensed post-pace (e.g., after delivering a ventricular pacing pulse) may be received by processor 148 for analysis in detecting ER signal morphology changes. When the sensing electrodes are outside the heart, for example when sensing circuitry of cardiac monitoring device 252 or ICD 260 is sensing the cardiac electrical signal, the digitized, amplified and widebandfiltered cardiac electrical signal passed to processing circuitry of the medical device may be referred to as an “ECG” signal.

[0086] In some examples, the EGM signal passed to control circuit 80 may be sensed by sensing circuit 86 by selecting a sensing electrode vector, via switching circuitry included in sensing circuit 86, that may be a relatively far field signal in some examples. For instance, sensing circuit 86 may receive a unipolar signal which may be sensed using the tip electrode 32 and housing 15 or the ring electrode 34 and housing 15. When a defibrillation electrode is available, the defibrillation electrode may be used in a sensing electrode vector for sensing a relatively far field signal that includes relatively more global information of the ventricular electrical activation than a near-field EGM signal, e.g., a bipolar signal, sensed to detect a local depolarization of the cardiac tissue by cardiac event detector circuit 143 as described below.

[0087] The ER signal morphology analysis described herein is not limited to being performed on a far field or unipolar cardiac electrical signal, however. In some examples, a near field or bipolar EGM signal may be sensed, e.g., using electrodes 32 and 34, for sensing ER signals. In some examples, sensing circuit 86 may include a dedicated morphology sensing channel configured to receive a far-field signal or a near-field signal from a selected sensing electrode pair. The morphology sensing channel may include an input pre-filter / amplifier, ADC and wideband filter for passing an EGM signal to control circuit 80 for performing ER signal morphology analysis for detecting changes in the ER signal according to techniques described herein.

[0088] Processor 148 may establish a morphology template of an ER signal of the received EGM signal. The morphology template is representative of the QRS waveform that can be sensed after a pacing pulse captures the ventricles at an initial time point, e.g., when morphology monitoring begins. The morphology template may be established during pacing at a relatively high pacing pulse output, e.g., 5.0 volts pulse amplitude, that is well above the pacing capture threshold to capture the adjacent cardiac tissue (myocardium and / or conduction system). The morphology template may be stored in memory 82 for use by control circuit 80 during ER morphology monitoring as further described below. For example, a post-pace waveform may be acquired by processor 148 from the EGM signal sensed during a template window following a capturing pacing pulse for comparison to the established morphology template stored in memory 82 fordetermining a morphology match score. The morphology match scores determined over time may be used by processor 148 for determining an adaptive threshold for detecting an ER signal morphology change. In some examples, in addition to or alternatively to determining a morphology match score, features of the ER signal may be determined by processor 148 of control circuit 80 for determining an adaptive threshold for detecting an ER signal morphology change.

[0089] Sensing circuit 86 may further include a rectifier and narrowband filter / amplifier 142 for receiving the ADC signal and passing a rectified, filtered signal to cardiac event detector circuit 143. Cardiac event detector circuit 143 may produce a cardiac sensed event signal, e.g., a sensed ventricular event (Vsense) signal or a sensed atrial event (Asense) signal, in response to the rectified signal crossing a sensing threshold amplitude, e.g., an R-wave sensing threshold amplitude or a P-wave sensing threshold amplitude, respectively.

[0090] Sensing circuit 86 may include multiple sensing channels, e.g., a ventricular sensing channel and an atrial event sensing channel, and in some examples, a morphology sensing channel as described above for passing an EGM signal to control circuit 80. For example, sensing circuit 86 may include an atrial sensing channel configured to receive an atrial signal from atrial electrodes 20 and 22, which may be filtered, amplified, rectified and passed to an atrial event detector circuit included in event detector circuit 143. Cardiac event detector circuit 143 may generate an Asense signal in response to the atrial signal crossing a P-wave sensing threshold. A ventricular sensing channel of sensing circuit 86 may include the same or a different narrowband filter than the atrial sensing channel. A ventricular event detector circuit of event detector circuit 143 can be configured to generate a Vsense signal in response to the narrowband filtered ventricular signal crossing an R-wave sensing threshold. Each atrial sensing channel and ventricular sensing channel may include separate pre-filter / amplifiers 140, ADC 141, rectifier and narrowband filter / amplifier 142 and event detector circuit 143 or some components may be shared between the different sensing channels. Cardiac event detector circuit 143 may include one or more sense amplifiers, comparators and / or other components configured to receive the filtered and amplified atrial and ventricular signals, compare the signals to respective P- wave and R-wave sensing thresholds, and generate respective Asense and Vsense signals passed to control circuit 80.

[0091] The sensed event signals can passed to control circuit 80 for use in controlling pacing pulses. For example, in response to receiving an Asense signal, pace timing and control circuit 147 included in control circuit 80 may set a pacing escape interval timer for scheduling a ventricular pacing pulse (Vpace), which may be a CSP pulse, at an AV pacing interval. Control circuit 80 may inhibit an atrial pacing pulse in response to the Asense signal and schedule a subsequent atrial pacing pulse by starting an atrial lower rate interval (LRI) to provide bradycardia pacing of the atria when the LRI expires without an Asense signal being received from sensing circuit 86. The LRI may correspond to a programmed base pacing rate and is used by control circuit 80 to control the minimum heart rate of the patient to be at least the base pacing rate. At other times, the LRI may be a temporary rate response pacing interval based on a patient activity metric determined by control circuit 80 from a patient activity signal received from sensors 90.

[0092] In response to receiving a Vsense signal, a scheduled Vpace may be inhibited and a ventricular pacing interval, e.g., a ventricular LRI, may be started for scheduling a Vpace. If the ventricular LRI expires before a Vsense signal is received and before an Asense signal is received or an atrial pacing pulse is delivered for triggering an atrial synchronous Vpace at the AV pacing interval, therapy delivery circuit 84 may deliver the scheduled Vpace to pace the ventricles.

[0093] Pace timing and control circuit 147 may include various timers or counters for performing timing related functions of control circuit 80 including, but not limited to, counting down various pacing escape intervals set according to a permanent or temporary pacing mode. The pacing modes of IMD 14 may include single chamber atrial pacing, dual chamber atrial synchronous ventricular pacing, dual chamber atrial asynchronous ventricular pacing, and single chamber ventricular pacing, as non-limiting examples, and may include bradycardia pacing with rate response pacing enabled in some examples. A sensed event signal may cause pace timing and control circuit 147 to trigger or inhibit a pacing pulse depending on the particular pacing mode in effect. Control circuit 80 may be configured to control therapy delivery circuit 84 to deliver atrial and ventricular pacing pulses according to a programmed or automatically selected pacing mode and programmed or automatically adjusted pacing control parameters.

[0094] Therapy delivery circuit 84 is configured to generate pacing pulses and includes a charging circuit 144 including one or more charge storage devices such as one or moreholding capacitors, an output circuit 146, and switching circuitry 145. Switching circuitry145 can be controlled by control signals from control circuit 80 to control when the holding capacitor(s) of charging circuit 144 are charged and when the charged holding capacitor(s) are discharged through the output circuit 146 to deliver pacing pulses via a selected pacing electrode vector for pacing the atria and / or the ventricles (which may be via the His-Purkinje conduction system) according to the programmed pacing mode.

[0095] Output circuit 146 may include switching circuitry for selecting the pacing electrode vector(s) and associated pacing electrode polarities coupled to a holding capacitor of charging circuit 144 via switching circuitry 145. For instance, output circuit146 may include switching circuitry for selecting the ventricular lead tip electrode 32 as a pacing cathode electrode with return anode ring electrode 34 for bipolar pacing in the area of the His bundle, LBB or RBB or a selected myocardial pacing site. Alternatively, ring electrode 34 may be selected as a cathode electrode with tip electrode 32 selected as the return anode in a bipolar pacing electrode vector for delivering ventricular pacing, e.g., CSP in interventricular septum. The RA lead electrodes 20 and 22 may be selected by switching circuitry included in output circuit 146 in an atrial pacing electrode vector for delivering atrial pacing pulses.

[0096] Charging of a holding capacitor to a programmed pacing voltage amplitude and discharging of the capacitor for a programmed pacing pulse width may be performed by therapy delivery circuit 84 according to control signals received from control circuit 80. For example, pace timing and control circuit 147 included in control circuit 80 may include programmable digital counters set by processor 148 for controlling the basic pacing time intervals, which can also be referred to as “escape intervals,” associated with various single chamber and dual chamber pacing modes as described above. Control circuit 80 may also set the amplitude, pulse width, polarity, and or other characteristics of the cardiac pacing pulses, which may be based on programmed values stored in memory 82.

[0097] Therapy delivery circuit 84 may include multiple pacing channels for delivering pacing pulses to the atria and to the area of the LBB, RBB, His bundle or other locations along the conduction system and / or myocardial pacing locations. Each pacing channel may be coupled to selected electrodes via switching circuitry included in output circuit 146 for selecting various unipolar or bipolar pacing electrode combinations for deliveringpacing pulses. In some examples, multiple pacing channels may include an additional left ventricular pacing channel, e.g., when IMD 14 is coupled to a coronary sinus lead to facilitate delivery of cardiac resynchronization pacing therapy.

[0098] When IMD 14 is configured to deliver high voltage CV / DF shocks, therapy delivery circuit 84 may include a high voltage therapy circuit for generating high voltage CV / DF shock pulses in addition to a low voltage therapy circuit that generates cardiac pacing pulses. Therapy delivery circuit 84 of IMD 14 may operate under the control of control circuit 80 to deliver a variety of cardiac electrical stimulation pulses, which may include bradycardia pacing pulses, anti-tachycardia pacing pulses, induction pulses for defibrillation testing, CV / DF shock pulses, post-shock pacing pulses, impedance measurement drive signals, capture test pulses, and ER signal monitoring test pulses, as examples.

[0099] Sensor(s) 90 may include a patient activity sensor provided for sensing a signal correlated to patient physical activity for use by control circuit 80 in controlling rate response pacing. In one example, the activity sensor is an accelerometer, e.g., a single or multi-axis piezoelectric sensor or MEMS device, for sensing an acceleration signal. The accelerometer may produce an electrical signal correlated to motion or vibration of the accelerometer, e.g., when subjected to patient body motion. The activity sensor may include one or more filter, amplifier, rectifier, analog-to-digital converter (ADC) and / or other components for producing an acceleration signal that may be passed to control circuit 80 for use in determining a patient physical activity metric for controlling rate response pacing.

[0100] Communication circuit 88 may include a transceiver and antenna for communicating with external device 50 (shown in FIG. 1) and / or co-implanted medical devices using radio frequency communication or other communication protocols as described above. Control parameters utilized by control circuit 80 for sensing cardiac event signals (e.g., intrinsic depolarization signals and ER signals), analyzing EGM signals, and controlling cardiac pacing may be programmed into memory 82 via communication circuit 88 for retrieval and execution by processor 148 of control circuit 80. Under the control of control circuit 80, communication circuit 88 may receive downlink telemetry from and send uplink telemetry to the external device 50. As described above in conjunction with IMD 114 in FIG. 3, in some examples, communication circuit88 may be configured to communicate with another implanted device, e.g., the atrial pacemaker 214 shown in FIG. 3, the cardiac monitoring device 252 shown in FIG. 4, or the ICD 260 shown in FIG. 4, for coordinating ER signal monitoring for detecting a change in the ER signal morphology. In some examples, intra-body communication performed by communication circuit 88 with another medical device may include tissue conductance communication where electrical communication signals are emitted by the electrodes of the transmitting device, conducted via the patient’s body fluids and tissues, and received by the electrodes of the receiving device.

[0101] FIG. 6 is a flow chart 300 of a method for detecting an ER signal morphology change by a medical device system according to some examples. For the sake of illustration, flow chart 300 and other flow charts and diagrams presented herein are described with reference to IMD 14 of FIG. 1 and FIG. 5. It is to be understood from the descriptions and examples presented herein, however, that other medical devices or medical device systems including two or more medical devices implanted and / or external to the patient may be configured to perform the methods disclosed herein.

[0102] At block 302, control circuit 80 may start an initialization phase during which ER signal features are determined (block 304) according to a morphology monitoring protocol. Detection of a change in the ER signal morphology may not be detected during the initialization phase. During the initialization phase, one or more ER signal features may be determined from one or more post-pace ER signals sensed during one or more pacing episodes. As an example, control circuit 80 may determine an ER signal feature at block 304 as a morphology match score between a morphology template stored in IMD memory 82 and at least one ER signal sensed following a pacing pulse delivered by therapy delivery circuit 84. Other ER signal features that may be determined at block 304, in addition to or alternatively to a morphology match score, may include a ventricular activation time (VAT), a QRS width, QRS area, maximum slope, maximum peak, maximum peak time, and peak polarity, as examples with no limitation intended.

[0103] The ER signal feature(s) determined at block 304 may be determined during a morphology monitoring test performed at one or more scheduled times of day or specified time intervals. In an illustrative example, the ER signal feature(s) may be determined during a morphology monitoring test performed once per day. In various examples, themorphology monitoring test may be performed once per hour, every four hours, every eight hours, every twelve hours, daily, weekly, bi-weekly, or monthly.

[0104] During the morphology monitoring test, control circuit 80 may control therapy delivery circuit 84 to deliver one or more ventricular pacing pulses at a pulse output and a pacing interval that is expected to capture the ventricles with a high degree of confidence prior to an intrinsic depolarization and without resulting in a fusion pacing pulse that occurs when an intrinsic ventricular depolarization and a pacing evoked depolarization occur concomitantly resulting in fusion of the intrinsic and pacing evoked depolarization wavefronts. During atrial synchronous ventricular pacing, for example, morphology monitoring test Vpaces may be delivered at a short atrioventricular (AV) interval to promote an early pacing evoked depolarization, prior to an intrinsic ventricular depolarization being conducted from the atria. In other instances, one or more Vpaces may be delivered at an overdrive pacing rate that is faster than the intrinsic ventricular rate to promote pacing evoked depolarizations and corresponding non-fusion pacing ER signals.

[0105] Because the morphology monitoring test is performed to detect a change in the ER signal during a subsequent monitoring phase, the ventricular pacing pulses are delivered at a pacing pulse output (pulse amplitude and pulse width) that is expected to capture the ventricles. If loss of capture occurs, an ER signal is not detected and not analyzed for detecting a change in the ER signal morphology. For example, the ER signal features determined at block 304 may be determined from the EGM signal sensed during an ER window that begins and ends at specified times following the delivery of a Vpace. If the Vpace does not capture the ventricles, an ER signal is not expected to occur during the ER window (an intrinsic QRS waveform may occur later than the ER signal and not in timed relation to the delivered Vpace). The Vpaces delivered during the morphology monitoring test may be delivered at a specified, relatively high pacing pulse output, e.g., 5 volts pulse amplitude or higher or other pacing output selected for a given patient that is well above a pacing capture threshold. In some examples, the Vpaces may be delivered during the morphology monitoring test at a specified, relatively large safety margin greater than a recently determined pacing capture threshold or at the same pacing pulse output used to establish the morphology template stored in memory 82.

[0106] The Vpaces delivered during the morphology monitoring test may be delivered at the same pacing pulse output used to deliver Vpaces when the morphology monitoring testis not being performed or at a higher pacing pulse output. When the test Vpaces are delivered at a higher pacing pulse output during the morphology monitoring test, the ER signal morphology may be different than the ER signal morphology that occurs during “normal” pacing, outside of the morphology monitoring tests. However, a change in the ER signal morphology that occurs at the higher test pacing pulse output during the morphology monitoring test can be an indication of a change in the underlying pacing substrate, e.g., due to electrode shifting or a sudden change in the cardiac tissue.

[0107] At block 306, control circuit 80 may determine an adaptive threshold from the determined ER signal features. The adaptive threshold can be a value, range or region determined from ER signal features by control circuit 80 against which new or more recently determined ER signal features can be compared to for detecting a change in the ER signal morphology. The adaptive threshold can be determined by control circuit 80 from multiple, historical ER signal features. The adaptive threshold is a benchmark against which recently determined ER signal features can be compared to for detecting an ER signal change. The adaptive threshold may be determined by control circuit 80 by determining an adaptive center metric and / or an adaptive variability metric, from which the adaptive threshold can be updated using the results of new morphology monitoring tests. Various examples of methods for determining and updating an adaptive threshold as a value, range or region at block 306 are further described below.

[0108] At block 308, control circuit 80 may determine if the initialization period is complete. In some examples, the initialization period is one morphology monitoring test. Based on the ER signal feature(s) determined from one or more ER signals during a ventricular pacing episode of the morphology monitoring test, an adaptive threshold may be set to an initial value at block 306, and control circuit 80 may begin the monitoring phase at block 310. In other examples, the initialization phase may include multiple morphology monitoring tests, which may be performed over the course of multiple minutes, hours or days. When the morphology monitoring tests are performed once daily, for instance, the initialization phase may be performed over 2, 3, 7, 10, 15 or 30 days as examples. In other examples, the morphology monitoring tests may be performed at a higher frequency during the initialization phase than during the monitoring phase. For instance, during an initialization phase the morphology monitoring tests may be performed once every hour or once every day followed by daily or weekly morphology monitoringtests during the monitoring phase. In still other examples, the initialization phase is optional. A nominal starting center metric and variability metric may be used to apply a starting threshold to the first morphology monitoring test result and determine a new center metric, variability metric and adaptive threshold that can be subsequently updated with each morphology monitoring test.

[0109] The adaptive threshold may be updated at block 306 as ER signal features are determined from each of the morphology monitoring tests. If the initialization phase is not complete (“no” branch of block 308), control circuit 80 may return to block 304 to wait for the next morphology monitoring test to be performed for obtaining new ER signal features. In some examples, control circuit 80 may obtain ER signal features from multiple morphology monitoring tests performed over an initialization phase and determine the adaptive threshold at the end of the initialization phase instead of updating the adaptive threshold after each morphology monitoring test during the initialization phase as shown in the flow chart 300.

[0110] If the initialization phase is complete, control circuit 80 may begin the monitoring phase at block 310. When the monitoring phase begins, control circuit 80 continues to update the adaptive threshold and uses the adaptive threshold to detect a change in the ER signal morphology. At block 312, control circuit 80 may determine the ER signal features during the next morphology monitoring test, e.g., at the next scheduled time according to the morphology monitoring protocol. The ER signal features determined at block 312 are analogous to the signal features determined during the initialization phase, e.g., morphology match score, VAT, or any of the examples listed above or combinations thereof.[oni] At block 318, the adaptive threshold and the most recently determined ER signal features are assessed by control circuit 80 for determining if ER morphology change criteria are met. Example methods for determining when the ER morphology change criteria, also referred to herein as “change criteria,” are met as further described below. Briefly, control circuit 80 may determine that the change criteria are met at block 318 when the ER signal features from a sufficient number of most recent time points of morphology monitoring tests are outside the adaptive threshold. For example, if the ER signal features determined from a threshold number of morphology monitoring tests fall outside the adaptive threshold value, range or region relative to an adaptive center metricused to determine the adaptive threshold, control circuit 80 may detect an ER morphology change.

[0112] If the change criteria are not met, control circuit 80 may return to block 312 to wait for the next morphology monitoring test for determining new ER signal features. When the change criteria are unmet, the most recently determined ER signal features may be used to update the adaptive threshold at block 316 in some examples. As further described below, however, the adaptive threshold may not be updated using the most recent ER signal features if the most recent ER signal features fall outside the adaptive threshold relative to an adaptive center metric, even when the change criteria are not yet met.

[0113] When control circuit 80 determines that the change criteria are met at block 318, control circuit 80 may perform a change detection response at block 320. The change detection response can include storing a pacing signal episode, including one or more ER signals, in memory 82 with a flag indicating the ER morphology change detection. The change detection response can include storing one or more ER signal features and the adaptive threshold, adaptive center metric, and / or adaptive variability metric at the time of the ER morphology change detection. The change detection response can include adjusting a pacing control parameter, such as adjusting the pacing pulse output (e.g., by adjusting the pulse amplitude and / or the pulse width), changing the pacing electrode vector (e.g., selecting a different cathode electrode, selecting a different anode electrode and / or changing from a bipolar pacing electrode vector to a unipolar pacing electrode vector), changing a pacing mode, adjusting a pacing interval or other pacing adjustment. The pacing mode could be adjusted to a sensing only pacing mode (e.g., OVO or ODO) or to a single chamber pacing mode (e.g., to atrial chamber pacing only), as examples, to avoid delivering pacing pulses that are not capturing a targeted pacing site.

[0114] For example, when the ER signal morphology is changed, control circuit 80 may select a different pacing electrode vector, if available, to regain the ER signal morphology present prior to the detected change or to deliver pacing at a different pacing site in a manner that may be more energy efficient and / or promote heart chamber synchrony. Control circuit 80 may increase the pacing pulse output using the same or a different pacing electrode vector in an attempt to regain the ER signal morphology. In other examples, control circuit 80 may decrease the pacing pulse output to be a safety margin greater than a capture threshold that results in the new ER signal morphology to promoteconsistent ventricular capture while conserving the power source of IMD 14, instead of attempting to re-establish the ER signal morphology that matches a stored template for instance. The response performed at block 320 may include performing a pacing capture threshold search for selecting a different pacing pulse output, which may include testing multiple pacing electrode vectors if available. Control circuit 80 may perform the response at block 320 by decreasing or increasing the AV pacing interval used for controlling the time of Vpace signals to promote synchrony between the atria and ventricles and / or between the right and left ventricles.

[0115] The response to the change detection may include controlling communication circuit 88 to transmit a signal to another medical device. The transmitted signal may be a notification or alert signal transmitted to the external device 50 to make a clinician or other caregiver aware of the detected change. The signal transmission may include other data relating to the detected change, e.g., the adaptive threshold at the time of change detection, the adaptive threshold stored over time, a history of the adaptive center metrics and / or adaptive variability metrics used to determine the adaptive threshold, the ER signal features at the time of change detection, and / or the ER signal features determined over time prior to change detection. When a second medical device is performing the ER signal analysis (e.g., external device 50 as shown in FIG. 1, cardiac monitoring device 252 or ICD 260 as shown in FIG. 4), the transmitted signal may be to the pacing device, e.g., IMD 114, to cause the pacing device to modify the delivery of pacing pulses, e.g., according to any of the pacing control parameter adjustments listed above.

[0116] In some examples, a response to detecting a change in ER signal morphology may include restarting the initialization phase. The initialization phase may be restarted after detecting an ER morphology change according to the techniques disclosed herein or after detecting an absolute threshold change in an ER signal feature to re-establish a new adaptive threshold. The initialization phase may be restarted upon a user-entered command received from external device 50.

[0117] FIG. 7 is a flow chart 400 of a method for determining an adaptive threshold applied to ER signal features for detecting ER morphology change according to some examples. At block 402, control circuit 80 may determine an ER signal feature. The ER signal feature may be determined during a scheduled morphology monitoring test (during the initialization phase or the monitoring phase). The ER signal feature may be themorphology match score, the VAT, or any of the other signal features listed herein as examples. The ER signal feature may be stored in memory 82 with previously determined ER signal features, e.g., determined during successive morphology monitoring tests. At block 404, control circuit 80 may determine an updated center metric using the ER signal feature. The center metric may be the mean, median, mode, a trimmed mean, or a specified percentile of stored values of the ER signal feature. In an example, a daily morphology match score may be determined from an ER signal and a stored morphology template. The daily morphology match score may be used to update a center metric determined as the average of morphology match scores stored in memory for each of the preceding, daily morphology monitoring tests. In some examples, as described below, control circuit 80 may determine two different ER signal features for use in determining an updated center metric, such as a centroid of a two-dimensional plot of the two different ER signal features.

[0118] In some examples, the new center metric may be determined as a weighted combination of the previous center metric and the new ER signal feature determined from the current morphology monitoring test. For example, the center metric (CM) may be determined as CM(i) = W1 * CM (i-l) + W2 * ERfeature(i), where CM(i) is the updated center metric at the current time point, W 1 and W2 are two different weighting factors (though W1 and W2 could be equal in some examples), CM(i-l) is the most recent preceding center metric and ERfeature(i) is the currently determined ER signal feature resulting from the current morphology monitoring test. In some examples, the weighting factors W 1 and W2 applied to the previous center metric and the new ER signal feature, respectively, may be different during the initialization phase than during the monitoring phase. For instance, W1 may be 0.75 and W2 may be 0.25 during the initialization phase so that each new ER signal feature has a relatively high influence on updating the center metric to an expected ER signal feature value during the initialization phase. During the initialization phase, W1 may be between 0.5 and 0.9 and W2 may be between 0.5 and 0.1 as examples.

[0119] During the monitoring phase a relatively higher weighting factor may be applied to the previous center metric combined with a lower weighting factor applied to the new ER signal feature. For instance, when the monitoring phase begins, control circuit 80 may determine the center metric using the above equation where W1 is 0.90 or higher and W2is 0.10 or lower. In other examples, the same weighting factors may be used during the initialization and monitoring phases. Generally a higher weighting factor W1 is applied to the previous center metric and a lower weighting factor is applied to the new ER signal feature each time the center metric is updated.

[0120] At block 406, control circuit 80 may determine an updated variability metric using the ER signal feature. Control circuit 80 may determine an absolute difference between the ER signal feature and the center metric and store the difference in memory 82. Control circuit 82 may determine the updated variability metric as the average of absolute differences between the updated center metric and each of the ER signal features stored in memory 82. Control circuit 80 may determine the updated variability metric as the standard deviation or variance of stored ER signal features.

[0121] In some examples, the updated variability metric may be determined as a weighted combination of the previous variability metric and the new absolute difference between the new ER signal feature and the previous (or updated) center metric. For example, the updated variability metric may be determined as VM(i) = W1 * VM(i-l) + W2 * distance^), where VM(i) is the updated variability metric for the current time point, W 1 and W2 are weighting factors, VM(i-l) is the previous variability metric, and distance(i) is the distance from the previous (or updated) center metric and the new ER signal feature. A relatively higher weighting factor may be applied to the previous variability metric and a relatively lower weighting factor may be applied to the distance between the new ER signal feature and the previous (or updated) center metric. During the initialization phase the weighting factors used by control circuit 80 for determining the updated variability metric may be different than the weighting factors used during the monitoring phase. The weighting factors used by control circuit 80 for updating the variability metric may be the same or different than the weighting factors used in updating the center metric during each of the initialization phase and monitoring phase.

[0122] At block 410, control circuit 80 may determine the adaptive threshold using the updated center metric and the updated variability metric. When the ER signal feature is the morphology match score, for instance, the adaptive threshold may be determined as the updated center metric less the updated variability metric or a multiple thereof. In other examples, depending on the ER signal feature being assessed, the adaptive threshold may be determined as the updated center metric plus the updated variability metric or amultiple thereof. In still other examples, the adaptive threshold may be determined as a range defined by the updated center metric plus or minus a multiple of the updated variability metric. The adaptive threshold may be applied as a threshold value or range to subsequently determined ER signal features. When an ER signal feature falls outside the adaptive threshold relative to the center metric, for example, control circuit 80 may detect an ER morphology change. In some examples, as described below, control circuit 80 may determine the adaptive threshold as a two-dimensional region (e.g., rectangular, elliptical, etc.) using an updated center metric of the two-dimensional points defined by two different ER signal features (at each morphology monitoring test time point) and updated variability metrics in each dimension. When more than two ER signal features are being determined and evaluated, a multi-dimensional region may be determined as the adaptive threshold using multi-dimensional points defined by the multiple ER signal features.

[0123] FIG. 8 is a graph 450 of ER signal morphology match scores (y-axis) that may be determined over time (x-axis) for updating an adaptive threshold and detecting an ER morphology change according to some examples. IMD control circuit 80 (or external device processor 52) may establish a morphology template prior to initiating the morphology monitoring tests. Control circuit 80 may receive multiple post-pace waveforms from a cardiac electrical signal sensed by sensing circuit 84 during an ER window. The ER window may be started at a specified time point after a delivered pacing pulse and may have a specified duration to encompass the post-pace ER signal and may exclude early post-pace variation due to pacing artifact. The ER window may be started, for example, at 20 to 100 ms after a Vpace. The ER window duration may be between 100 and 300 ms, as examples. In an example, the ER window begins at about 50 ms after the Vpace and ends at about 235 ms after the Vpace.

[0124] The post-pace ER signal waveforms collected in multiple ER windows may be aligned and ensemble averaged to establish the morphology template stored in memory 82. The post-pace ER signal waveforms may be acquired during ventricular pacing using a pacing pulse output known or expected to cause ventricular capture. The morphology template established by control circuit 80 may be representative of a baseline morphology of the ER signal, which may include conduction system capture.

[0125] During each morphology monitoring test, e.g., during and after the initialization phase, control circuit 80 may determine a morphology match score between themorphology template and an ER signal using waveform correlation methods, wavelet transform or other morphology matching techniques. In some examples, a Haar wavelet transform is employed to determine wavelet coefficients stored in memory 82 to represent the morphology template. The wavelet coefficients may be weighted to increase contributions of certain time-scales of the wavelet transform coefficients, e.g., to emphasize wider scale wavelet transform coefficients relative to narrower scale wavelet transform coefficients. In this way, the contribution of noise or insignificant EGM waveform information in the morphology template can be reduced in the resulting wavelet transform. Example methods of performing a Haar wavelet transform for determining a morphology match score are generally disclosed in U.S. Patent No. 6,393,316 (Gillberg, et al., filed May 8, 2000) and in U.S. Patent No. 8,521,268 (Zhang, et al., filed May 10, 2011), the entire content of both incorporated herein by reference.

[0126] The morphology match scores shown in FIG. 8 may be determined by summing the errors (differences) between the morphology template and wavelet transform coefficients of an ER signal obtained during a morphology monitoring test. The wavelet transform coefficients may be filtered and normalized in some examples. The absolute differences between the filtered and normalized wavelet coefficients of the ER signal sensed during a morphology monitoring test and the morphology template may be summed to obtain a “distance” between the ER signal and the morphology template. The ratio of the distance to a wavelet representation of the morphology template area (e.g., summation of the filtered and normalized wavelet transform coefficients of the template) may be determined. This ratio may be weighted and subtracted from 100 to determine the morphology match score between the ER signal morphology and the morphology template. In an example, the distance to area ratio can be multiplied by a weighting factor of 3, 4 or other selected weighting factor (where distance is the sum of absolute differences between the morphology template and ER signal wavelet transform coefficients, and area is the sum of absolute values of all morphology template wavelet coefficients).

[0127] With continued reference to HMD 14 of FIG. 5, control circuit 80 may begin determining morphology match scores during an initialization phase 452 according to a morphology monitoring protocol. For the sake of example, daily morphology monitoring tests may be performed such that a morphology match score is determined between an ERsignal and the stored morphology template. The morphology match score may be determined using the ER signal from a single paced ventricular cycle. In other examples, multiple ER signals may be ensemble averaged for comparison to the morphology template. In still other examples, multiple morphology match scores may be determined for multiple ER signals (sensed during the ER window following a respective number of Vpaces) and a representative morphology match score (e.g., a mean or median match score) may be stored in memory 82 for the morphology monitoring test time point.

[0128] Control circuit 80 may determine an adaptive threshold 460 that subsequent morphology match scores are compared against during the monitoring phase 462. The adaptive threshold 460 can be initialized to a default morphology match threshold value 466 at 0 days, e.g., which may be at the time of implantation of IMD 14. However, the initialization phase 452 may begin (and be repeated) at any time after implantation, e.g., after detecting an ER morphology change or in response to receiving a programming command from external device 50.

[0129] On day 1, the morphology match score may be averaged with the nominal threshold value 466 to obtain an updated center metric 455, in this case an updated mean morphology match score, which may be determined as a weighted combination of the previous center metric and the new morphology match score as described above, starting with the nominal threshold value 466 as the first center metric. An updated variability metric 457 may be determined as the average of a starting nominal variability metric and the difference between the first center metric (nominal threshold value 466) and the morphology match score determined on day 1. As described above, the updated variability metric 457 may be determined as a weighted combination of the previous variability metric and the new variability measurement, e.g., the new absolute difference between the new ER signal feature and the previous center metric. The adaptive threshold 460 may be determined as the updated mean morphology match score (center metric 455) less a multiple of the updated variability metric 457.

[0130] In some examples, instead of a specified nominal starting threshold 466, a starting center metric may be determined by control circuit 80 as the mean (or a minimum, median or other representative value) of multiple morphology match scores, which may be determined at the start of the initialization phase (and optionally more frequently than the morphology monitoring tests performed during the remainder of the initialization phase452 and / or the monitoring phase 462). For example, multiple morphology match scores may be determined beat by beat, once per minute, once per hour, once per day or other frequency for determining a mean morphology match score as the starting centric metric and determining a starting variability metric at the onset of the initialization phase 452. The starting center metric may be used in a weighted combination with the next morphology match score determined at the first morphology monitoring test time point during the initialization phase 452 after establishing the starting center metric. The starting variability metric may be used in a weighted combination with the difference between the next variability measurement and the starting center metric.

[0131] This process of updating the center metric and the variability metric using new morphology match scores may be repeated on each day of the initialization phase 452. An updated center metric, e.g., the mean morphology match score determined as a weighted combination of the previous center metric and the new morphology match score, may be determined from the daily morphology match score and all previous daily morphology match scores. An updated variability metric 457 may be determined as the average difference between the new mean morphology match score and each of the daily morphology match scores (e.g., stored in memory 82). As described above, updated variability metric 457 may be determined as the weighted combination of the previous variability metric determined from all previous variability measurements and the new variability measurement, e.g., the new absolute difference between the new morphology match score and the previous center metric. The adaptive threshold 460 may be determined by control circuit 80 as the updated mean morphology match score less a multiple of the updated variability metric 457. The adaptive threshold 460 may be the updated morphology match score center metric less 1.25 to 3 times the updated variability metric or twice the updated variability metric as examples.

[0132] During the initialization phase 452, control circuit 80 may not compare a morphology match score to the adaptive threshold 460 for detecting an ER morphology change. When the initialization phase 452 is complete and the monitoring phase 462 begins, control circuit 80 may begin comparing each daily morphology match score to the adaptive threshold 460 determined from the day before (the previous morphology monitoring test, before updating the adaptive threshold 460 based on the current day morphology match score). While the initialization phase 452 is shown as one week in theexample of FIG. 8, the initialization phase 452 may be a different number of days and may be as short as one day (or one morphology monitoring test) for establishing the first value of the adaptive threshold 460 using one morphology match score and a nominal threshold value 466, for example.

[0133] As observed by the graph 450 of FIG. 8, from day 1 to approximately day 14, the daily morphology match scores are relatively stable, resulting in a decreasing adaptive variability metric that is subtracted from the mean morphology match score for determining the adaptive threshold 460. As such, the adaptive threshold 460 approaches the mean morphology match score more closely as the variability in the match scores narrows.

[0134] From approximately day 15 to approximately day 30, the variability of the daily morphology match scores increases as indicated by arrow 454. As the range of morphology match scores increases, the updated variability metric that is subtracted from the updated center metric (e.g., mean morphology match score) for determining the adaptive threshold 460 increases. As such, the adaptive threshold 460 may move further away from the updated mean morphology match score as the variability in the match scores increases. In this way, increased variability of the morphology match score does not lead to a detection of an ER morphology change when the center metric remains relatively stable.

[0135] Furthermore, it is noted that a single morphology match score 456 from a daily morphology monitoring test that is less than (outside) the adaptive threshold 460 (relative to the center metric of the morphology match scores) is not detected as an ER morphology change in this example. The morphology match score 456 outside the adaptive threshold 460 at a single time point may be an outlier that could result in a false ER morphology change detection if one single morphology monitoring test result is compared to the adaptive threshold 460 for detecting an ER morphology change. Instead, control circuit 80 may determine that the morphology match score 456 is less than the adaptive threshold 460 but does not detect a change in the ER signal morphology based on a single outlier and does not use the morphology match score 456 to update the adaptive threshold 460 because it is outside the adaptive threshold 460.

[0136] Beginning after day 30 in this illustrative example, a sudden change in the daily ER morphology match scores 458 is observed. Each of the morphology match scores 458determined to be less than the adaptive threshold 460 are not used by control circuit 80 to update the adaptive threshold 460 in this example. As such the adaptive threshold 460 remains constant when daily morphology monitoring test results in morphology match scores 458 that are less than the adaptive threshold 460. Control circuit 80 may detect an ER morphology change at arrow 464, e.g., after a specified threshold number of morphology monitoring tests result in a morphology match score that is less than the adaptive threshold 460. In the example shown, when six daily morphology match scores are less than the adaptive threshold 460, an ER morphology change is detected. In other examples, control circuit 80 may detect an ER morphology change when two to ten consecutive morphology monitoring tests or three to eight consecutive morphology monitoring tests result in an ER signal feature falling outside the adaptive threshold. In some examples, control circuit 80 may detect an ER morphology change when at least a threshold percentage of a specified number of consecutive morphology monitoring tests result in an ER signal feature falling outside the adaptive threshold.

[0137] As observed in FIG. 8, when a non-adaptive threshold is applied to the morphology match scores, such as the nominal threshold value 466, a sudden change in ER morphology may go undetected. Morphology match scores 458 all remain greater than the nominal threshold value 466. By updating the adaptive threshold 460 using the updated center metric (e.g.,. mean morphology match score) and variability metric, a sudden change in ER morphology can be reliably detected without falsely detecting a change due to increased variability in the ER morphology.

[0138] FIG. 9 is a graph 470 of ER signal morphology match scores (y-axis) that may be determined over time (x-axis) for updating an adaptive threshold 480 and detecting an ER morphology change according to another example. With continued reference to IMD 14 of FIG. 5, control circuit 80 may begin determining morphology match scores during an initialization phase 472 according to an ER signal monitoring protocol as described above. Continuing with the example of daily morphology monitoring tests, a morphology match score is determined by control circuit 80 between one or more ER signals and the stored morphology template each day. The daily ER morphology match score, which may be a representative morphology match score determined from multiple morphology match scores determined on a given day, is buffered in memory 82. The adaptive threshold 480 may be updated during the initialization phase 472. The adaptive threshold 480 can beinitialized to a nominal morphology match threshold 486 at 0 days, e.g., at the time of implantation of IMD 14 or at any subsequent time at which the initialization phase 472 is started.

[0139] As described above, during the initialization phase 472 the daily morphology match score can be used by control circuit 80 to determine an updated center metric 475, e.g., mean morphology match score determined as a weighted combination of the previous center metric and the new morphology match score, and an updated variability metric 476, e.g., determined as a weighted combination of the previous variability metric and the new variability measurement. The adaptive threshold 480 may be determined by the control circuit 80 as the updated center metric 475 less a multiple of the updated variability metric 476. By the end of the initialization phase 472, the adaptive threshold 480 is tracking the mean morphology match score (updated center metric 475) but is a multiple of the updated variability metric less than the mean morphology match score. When variability is large, the adaptive threshold 480 is further from the mean morphology match score. When the variability is small, the adaptive threshold 480 more closely approaches the mean morphology match score. In this way, when the monitoring phase 482 begins, expected variation in the ER signal based on past morphology match scores can fall above (inside) the adaptive threshold 480.

[0140] During the initialization phase 472, control circuit 80 may not compare the daily morphology match score to the adaptive threshold 480 for detecting an ER morphology change. All of the daily morphology match scores may be used in updating the center metric, variability metric and adaptive threshold 480 during the initialization phase 472 in some examples. When the monitoring phase 482 begins, control circuit 80 may begin comparing the daily morphology match score to the adaptive threshold 480 for ER morphology change detection. When the daily morphology match score is not less than the adaptive threshold 480, it can be used by control circuit 80 to determine an updated center metric 475 and an updated variability metric for updating the adaptive threshold 480. When the daily morphology match score is less than (outside) the adaptive threshold 480, control circuit 80 may not use the morphology match score for updating the center metric 475, the variability metric 476 or the adaptive threshold 480.

[0141] As observed by the graph 450 of FIG. 9, daily morphology match scores are gradually decreasing with relatively stable variability. As such, the adaptive threshold 480decreases over time, but, in this example, the daily morphology match scores remain greater than the adaptive threshold 480 such that an ER morphology change is not detected by control circuit 80. A gradual change in ER signal morphology, resulting in a gradual change in the center metric 475, in contrast to a more sudden change as depicted by the drop in morphology match scores 458 shown in FIG. 8, may not be a condition warranting an alert or notification of the change or another morphology change response. A change in ER morphology due to a shift in electrode location or a sudden pathological change of the pacing substrate is expected to occur more suddenly such that daily morphology match scores (or other ER signal features) will fall outside the expected variability metric 476 (or multiple thereof) from the center metric 475, e.g., below the adaptive threshold 480.

[0142] As observed in FIG. 9, if a non-adaptive nominal threshold value 486 is applied to the morphology match scores, the gradual decrease in morphology match scores that crosses the non-adaptive threshold 486 may be detected as an ER morphology change resulting in a response being performed by the medical device system, e.g., an alert or notification signal transmission and / or pacing control parameter adjustment. The detection of the ER morphology change may be a “false” detection because, even though the ER signal morphology has gradually changed over time, the cause for the gradual change may not be of clinical concern compared to a more sudden ER morphology change. For example, a gradual change in the ER signal morphology may occur with gradual disease progression or other cardiac structural or physiological changes that occur gradually over time but are not a cause for an alert or notification. By updating the adaptive threshold as a benchmark that takes into account both the center and variability of the ER signal morphology features, in this case morphology match scores, a “false” detection of a nonconcerning change in ER signal morphology that occurs gradually over time can be avoided. In some examples, however, an absolute threshold such as non-adaptive threshold 486, may be applied to the ER signal features in addition to the adaptive threshold 480. When the ER signal feature crosses a non-adaptive threshold 486, for example, control circuit 80 may restart the initialization phase 472 to re-establish the adaptive threshold 480 because the older ER signal features contributing to the center metric and variability metric may be less relevant in monitoring for a sudden change in ER signal morphology over time.

[0143] FIG. 10 is a graph 500 of ER signal morphology features (y-axis) that may be determined over time (x-axis) by IMD 14 according to yet another example. In the examples of FIGs. 8 and 9, control circuit 80 determines the morphology match score between a stored ER morphology template and one or more ER signals sensed during the morphology monitoring test performed at scheduled time points according to the morphology monitoring protocol. Additionally or alternatively, control circuit 80 may determine other features of the ER signal, such as a ventricular activation time (VAT). The VAT, which may be determined using a time point of a feature extracted from the ER signal, may represent a time interval from a delivered Vpace to an electrical activation time of one or both ventricles. During CSP, for example, it can be desired to shorten the left ventricular activation time (LVAT) to promote ventricular electrical synchrony between the right and left ventricles.

[0144] In 12-lead surface ECG signals, the LVAT can be measured from a delivered pacing pulse to a maximum peak or other fiducial point of the V5 or V6 ECG signal, for example. The QRS signal of the V5 or V6 ECG signal tends to be more representative of the timing of the left ventricle electrical activation compared to the VI or V2 ECG signal that tends be to more representative of the timing of the right ventricular electrical activation. A VAT metric that can be determined as surrogate for the LVAT determined from surface ECG signals can be determined from an EGM or ECG signal sensed by an IMD. A VAT metric that is correlated to the LVAT determined from a surface ECG signal can be determined by control circuit 80 by determining a center of area from the ER signal.

[0145] Control circuit 80 may determine a center of area of a segment of the ER signal waveform and the corresponding time of the center of area. Control circuit 80 may determine a center of area time by calculating the geometric center under a segment of the ER signal extending from i=l to i=N, where i is the sample point number starting from 1 at a specified start time (e.g., an ER window) after the delivered Vpace to N, the sample point of the maximum peak amplitude of the ER signal, as an example. The geometric center of this segment of the ER signal may be computed by control circuit 80 as the arithmetic mean position, e.g., given by the summation of the products {i * A(i)} where i = 1 to N divided by the summation of A(i) where i = 1 to N. A(i) represents the amplitude of the ith sample point of the ER signal, e.g., beginning from the ER window to themaximum peak amplitude of the ER signal during the ER window. Control circuit 80 may determine the VAT metric as the difference between the maximum peak amplitude time and the center of area time as an ER signal feature during morphology monitoring tests.

[0146] In the example of FIG. 10, the VAT is shown on the y-axis as an ER signal feature. Control circuit 80 may determine the VAT during multiple morphology monitoring tests over an initialization phase 502 for establishing an adaptive threshold against which VATs are compared for detecting an ER morphology change during the monitoring phase 512. The VAT may increase or decrease with changes in the ER signal morphology. As such, in some cases, an adaptive threshold determined by control circuit 80, against which VATs are being assessed during morphology monitoring tests, may be defined by an upper adaptive threshold 510b and a lower adaptive threshold 510a, collectively referred to as adaptive threshold range 510 applied as two values (high and low) at each time point instead of the single adaptive threshold value applied at each time point in the examples of FIGs. 8 and 9 pertaining to morphology match scores.

[0147] A VAT center metric 505, e.g., a mean VAT which may be determined as a weighted combination of the previous VAT center metric and a new VAT from the current morphology monitoring test, may be updated each time a new VAT is obtained, e.g., daily or according to another morphology monitoring protocol. A nominal starting VAT center metric may be a specified value, and a starting VAT variability metric may be a specified value. In other examples, on day 1, the VAT may optionally be averaged with a nominal lower threshold starting value 504. The lower adaptive threshold 510a may be determined as the mean VAT less a variability metric, e.g., a multiple of the average (absolute) difference between the updated mean VAT and stored VATs. On day 1, the variability metric may be a multiple of the distance from the mean VAT to the first VAT or the weighted combination of a nominal starting variability metric and the distance from the updated VAT center metric to the new VAT.

[0148] On day 1, the VAT may optionally be averaged with a nominal upper threshold starting value 506. The upper adaptive threshold 510b may be determined as the mean VAT plus a variability metric. In some examples, the VAT center metric 505 at the first time point, e.g., day 1, may be determined as the mean of the first VAT determined, the nominal lower threshold starting value and the nominal upper starting threshold value for use in determining both the starting value of the lower adaptive threshold 510a and theupper adaptive threshold 510b. In some examples, the VAT center metric 505 may be updated at subsequent time points, e.g., daily, using only the determined VATs without using the nominal lower threshold starting value 504 and the nominal upper starting threshold value 506.

[0149] In still other examples, a starting center metric may be a specified value or may be determined by determining a mean of multiple VATs, which may be determined more frequently than during the remainder of the initialization phase 502 and / or the monitoring phase 512. For example, multiple VATs may be determined beat by beat, once per minute, once per hour, once per day or other frequency for determining a mean VAT as the starting VAT centric metric and determining a starting VAT variability metric. The starting VAT center metric may be used in a weighted combination with the next VAT determined at the first morphology monitoring test time point during the initialization phase after establishing the starting VAT center metric. The starting variability metric may be used in a weighted combination with the difference between the next VAT measurement and the starting VAT center metric.

[0150] During the initialization phase 502, the variability of the VATs is relatively small. The adaptive threshold range 510 narrows to define the range of values that the VAT is expected to fall in based on the VAT center metric 505 and the variability metric. Thus, the adaptive threshold range 510 defines a benchmark against which subsequent VATs can be compared for detecting a change in the ER signal morphology during the monitoring phase 512.

[0151] If the variability of the VATs increases, e.g., as indicated by arrow 508, the adaptive threshold range 510 widens. A false detection of an ER morphology change is avoided by applying a wider adaptive threshold range 510 to subsequent VATs when the variability metric increases. Control circuit 80 may not detect an ER morphology change in response to a single outlier 516. Control circuit 80 may not use the outlier 516, that is outside the adaptive threshold range 510 (relative to the center metric 505 at the given time point) for updating the adaptive threshold range 510. In some examples, outliers may not be excluded during the initialization phase 502 to allow the adaptive threshold range 510 to reflect the variation of the VAT, but outliers may be excluded after the initialization phase 502.

[0152] At arrow 514, control circuit 80 may detect an ER morphology change in response to a threshold number of VATs being outside the adaptive threshold range 510. Control circuit 80 may verify that the threshold number of VATs that are outside the adaptive threshold range 510 are all in the same direction, e.g., all less than the adaptive threshold range 510 or all greater than the adaptive threshold range 510, in order to detect the ER morphology change. The VATs falling outside the adaptive threshold range 510 may not be used by control circuit 80 to update the VAT center metric 505 or the VAT variability metric such that the adaptive threshold range 510 remains constant when VATs are consistently outside the adaptive threshold range 510, as observed in FIG. 10. In the example shown, control circuit 80 detects the ER morphology change in response to the sixth consecutive daily morphology monitoring test resulting in a VAT that is outside the adaptive threshold range 510.

[0153] It is to be understood that a different threshold number of VATs falling outside the adaptive threshold range 510 may result in ER morphology change detection. The threshold number of ER signal features outside a given adaptive threshold that is required to detect the ER morphology change may depend on the frequency of determining the ER signal features (e.g., hourly, daily or weekly), the particular ER signal feature being determined and / or other factors. The threshold number of morphology monitoring tests resulting in ER signal features falling outside a respective adaptive threshold, in this case the adaptive threshold range 510, may be 2, 4, 6, 10, 15, 20 or other selected number and may or may not be required to be results of consecutive morphology monitoring tests. For example, if a specified percentage of the most recent Y VATs, e.g., 70%, 80%, or 90%, of the most recent 5, 8, 10, 12 or 15 morphology monitoring tests are outside the adaptive threshold range 510, control circuit 80 may detect an ER morphology change.

[0154] Furthermore, it is to be understood that in some examples more than one ER signal feature may be determined during each of the ER morphology monitoring tests and compared to respective adaptive thresholds for detecting an ER morphology change. For instance, the morphology match score (as described in conjunction with FIGs. 8 and 9) and the VAT may both be determined during each morphology monitoring test and compared to respective adaptive thresholds. When at least one or both is outside a respective adaptive threshold for a threshold number of times, control circuit 80 may detect an ER morphology change. The threshold number of times applied to one ER signalfeature may be the same or different than the threshold number of times applied to a different ER signal feature. The threshold number of times applied to one ER signal feature may be lowered by control circuit 80 when a second ER signal feature has fallen outside its respective adaptive threshold a specified threshold number of times. To illustrate, if the VAT is outside the adaptive threshold range 510 for at least six consecutive daily morphology monitoring tests, and the ER morphology match score is less than its respective adaptive threshold for at least three out of those six consecutive morphology monitoring tests, control circuit 80 may detect the ER morphology change. In another illustrative example, control circuit 80 may detect the ER morphology change when at least 90% of the morphology match scores are less than the morphology match score adaptive threshold and over the same period at least 50% of the VATs are outside the VAT adaptive threshold range.

[0155] In still other examples, an adaptive threshold may be applied to one ER signal feature and a fixed threshold may be applied to a second ER signal feature. For instance, when the VAT falls outside the adaptive threshold range 510 for a threshold number of morphology monitoring tests, control circuit 80 may determine the morphology match score of the ER signal and compare the morphology match score to a specified, fixed threshold. If the morphology match score is less than the fixed threshold value, control circuit 80 may detect an ER morphology change in response to the combination of the VATs being outside an adaptive threshold and the morphology match score being less than a fixed threshold.

[0156] FIG. 11 is a diagram 550 of plotted ER signal features and an adaptive threshold region 566 that may be determined for detecting an ER morphology change according to another example. VATs are plotted along the x-axis and morphology match scores are plotted along the y-axis. As described above in conjunction with FIG. 10, control circuit 80 may determine two or more ER signal features that can be compared to respective adaptive thresholds (or a combination of adaptive and fixed thresholds). In the example of FIG. 11, an adaptive threshold is determined by control circuit 80 as a two-dimensional (2D) adaptive threshold region 566 (or elliptical 2D adaptive threshold region 576) from two different ER signal features determined at each time point of the morphology monitoring tests.

[0157] Each of the plotted data points 552 within the adaptive threshold region 566, outlier 554, and plotted data points 556 falling outside the adaptive threshold region 566 represent the VAT (x-axis) and corresponding ER morphology match score (y-axis) determined at a respective time point of a morphology monitoring test. Using the example of daily morphology monitoring tests, control circuit 80 may determine and store the VAT and the morphology match score in a buffer in memory 82 for accumulating the ER signal features for use in updating the center metrics 561 and 565, respectively, and the variability metrics 562 and 567, respectively, for the two ER signal features. The center metric and variability metric are updated at each time point and stored in memory 82 for use in determining the next updated center metric and variability metric in a weighted combination with a new ER signal feature at the next morphology monitoring time point. In this way, a large number ER signal features need not be accumulated in memory 82.

[0158] The center metrics and variability metrics may be determined for each morphology monitoring test and used to update the adaptive threshold region 566 as indicated. The data points 552, 554, and 556 represent accumulated morphology monitoring test results that may be buffered in memory 82 and used for updating the adaptive threshold region 566 (or 576) up to the current time point of the most recent morphology monitoring test. As such the adaptive threshold region 566 represents the adaptive threshold applied by control circuit 80 to the morphology monitoring test result at a current time point. In FIGs. 8 and 9, the adaptive threshold values are plotted over time, representing the changing adaptive threshold value that is applied to the results of morphology monitoring tests performed at a respective time point.

[0159] The adaptive lower threshold 560a and adaptive upper threshold 560b, referred to collectively as adaptive range 560, may be determined using each VAT result that is not outside the low and high values 560a and 560b of the adaptive threshold range 560 at the respective given time point. As such, the adaptive threshold range 560 is determined by control circuit 80 using an updated VAT center metric 561 of data points 552 and 554 (within range 560) but not using data points 556 that fall outside adaptive threshold range 560.

[0160] The adaptive threshold 564 may be determined using each ER morphology match score that is not below the adaptive threshold 564 at the respective time point. In the example of FIG. 11, outlier 554 may be within the adaptive threshold range 560 applied toVATs but is outside the adaptive threshold 564 relative to an updated morphology match score center metric 565. As such, the VAT corresponding to outlier 554 may have been used by control circuit 80 for updating the VAT center metric 561 and VAT variability metric 562 used to update adaptive threshold range 560. The morphology match score associated with outlier 554, however, may have been ignored by control circuit 80 for updating the morphology match score center metric 565 and variability metric 567 used to update adaptive threshold 564 (as the center metric 565 less the variability metric 567) because the morphology match score of outlier 554 is less than the adaptive threshold 564. As such, the adaptive threshold range 560 may have changed on the day that outlier 554 was acquired, but the value of adaptive threshold 564 may have remain unchanged.

[0161] The grouping of points 556 that are numbered 1-6 are determined on consecutive days after the outlier point 554 and all fall outside the adaptive threshold region 566. As such, in some examples, none of the VATs associated with points 1-6 are used to update the adaptive range 560. None of the morphology match scores associated with points 1-6 may be used by control circuit 80 to update the adaptive threshold 564. The adaptive threshold region 566, bounded by adaptive threshold range 560 and adaptive threshold 564 (and the maximum possible morphology match score of 100 along the upper boundary), may therefore remain constant after the time point associated with outlier 554 because points 556 may not be used by control circuit 80 in updating either adaptive threshold range 560 or adaptive threshold 564 defining adaptive threshold region 566.

[0162] The data points 556 are numbered in the chronological order in which they are determined from morphology monitoring tests. In the illustrative example, the data point numbered 6 of the grouping of data points 556 may result in an ER morphology change detection 514 by control circuit 80 because the threshold number of six data points falling outside the adaptive threshold region 566 is met. As described above, a different threshold number of data points (corresponding to a threshold number of morphology monitoring tests) may be required to detect the ER morphology change.

[0163] In other examples, control circuit 80 may determine an adaptive threshold region 576 defined as an elliptical region instead of the rectangular adaptive threshold region 566. In determining the rectangular adaptive threshold region 576, control circuit 80 may determine a VAT center metric and variability metric and a morphology match score center metric and variability metric after each morphology monitoring test (with a resultthat falls within the respective adaptive threshold range 560 and adaptive threshold value 564). In other examples, control circuit 80 may determine a center metric 570 of the 2D data points 552 obtained from each morphology monitoring test as a geometric center or centroid of the 2D data points. If a 2D data point falls outside the elliptical adaptive threshold region 576, such as outlier 554, control circuit 80 may not use the outlier 554 for updating the center metric 570 of the 2D data points.

[0164] Control circuit 80 may determine a 2D variability metric by determining the average distance from each 2D data point (that falls within the elliptical adaptive threshold region 576) to the center metric 570. Control circuit 80 may determine the elliptical adaptive threshold region 576 as having a boundary defined by a circle having a radius defined as a multiple of the variability metric. In other examples, control circuit 80 may determine an x-axis radius 572 of the elliptical adaptive threshold region 576 by determining the average horizontal distance from each 2D data point to the center metric 570. Control circuit 80 may determine the y-axis radius 574 of the elliptical adaptive threshold region 576 by determining the average vertical distance from each 2D data point to the center metric 574 as a morphology match score variability metric.

[0165] The elliptical adaptive threshold region 576 may be determined by control circuit 80 as the ellipse having a center at center metric 570, horizontal radius 572 as a multiple of the average absolute horizontal distance of the VATs from the center metric 570, and vertical radius 574 as a multiple of the average absolute vertical distance of the ER morphology match scores from the center metric 570. The center and radii of the elliptical adaptive threshold region 576 may change each time the morphology monitoring test is performed when the VAT and ER morphology match score fall within the elliptical adaptive threshold region 576 at that time point. As generally described above, the updated center metric 570 may be determined as a weighted combination of the most recent previous centroid and a new 2D point defined by the new VAT and morphology match score at a given time point. The updated variability metric may be determined as a weighted combination of the most recent previous variability metric and a new variability measure (e.g., differences between the new 2D point and the previous or updated centroid in each x- and y- direction).

[0166] When the outlier 554 or any of the data points 556 fall outside the elliptical adaptive threshold region 576, the center metric 570 and variability metric(s) may not beupdated by control circuit 80 such that the center and radii of elliptical adaptive threshold region 576 remain unchanged. As observed in FIG. 11, data point 554 is outside the elliptical adaptive threshold region 576. It is not used to update the elliptical adaptive threshold region 576 resulting in elliptical adaptive threshold region 576 being narrower than the rectangular adaptive threshold region 566, which may be determined by control circuit 80 using outlier 554 because it falls within the adaptive threshold range 560 applied to VATs.

[0167] In this example, the outlier 554 may be among the earliest 2D data points falling outside the elliptical adaptive threshold region 276. The 2D data point numbered 5 in the grouping of data points 556 may be the sixth consecutive data point falling outside the elliptical adaptive threshold region 576. When the threshold number of data points for detecting an ER morphology change is six, continuing with the example given above, control circuit 80 may detect the ER morphology change in response to the outlier 554 and subsequent data points numbered 1-5, as indicated by arrow 580, when control circuit 80 applies the elliptical adaptive threshold region 576 to the morphology monitoring test results.

[0168] FIG. 12 is a flow chart 600 of a method for determining an adaptive threshold for detecting an ER morphology change according to another example. At block 602, one or more ER signal features are determined by control circuit 80 during a morphology monitoring test. Any of the example ER signal features listed herein may be determined at block 602, from one or more cardiac electrical signals sensed during an ER window following ventricular pacing pulses delivered to capture the ventricles. The ER signal feature(s) may be buffered in memory 82 for each of the consecutive time points of the morphology monitoring tests.

[0169] At block 604, control circuit 80 may perform a clustering operation of the buffered data points representing the results of the consecutively performed morphology monitoring tests. Control circuit 80 may cluster the data points in at least two groups. In an example, control circuit 80 may perform an agglomerative single-link clustering method where each data point is initially a “cluster” consisting of a single point. Pairs of clusters that are closest to each other (e.g., shortest minimum Euclidean distance) are successively merged until only two clusters of data points remain. The minimum Euclidean distance between each pair of clusters may be determined as the shortest or minimum distancebetween two points of the two clusters. For instance, if cluster A and cluster B each have multiple points, a pair of points (a(x, y), b(x, y)) may be identified where point a is in cluster A and point b is in cluster B and points a and b are the two points of the two clusters that are closest to each other. The Euclidean distance between points a and b is determined as the minimum distance between the two clusters. The minimum distances determined for all pairs of clusters can be determined by control circuit and compared to each other so that the shortest minimum distance may be identified. The two clusters separated by the shortest minimum distance are merged into a single cluster and the process may be repeated until two clusters remain. In other examples, the distances between clusters may be determined by determining a centroid of each cluster and determining the distances between centroids of pairs of clusters. The shortest distance between centroids of all cluster pairs may be identified. The two clusters having centroids the shortest distance apart may be merged. A visualization of clustering methods that may be performed by control circuit 80 is shown in FIGs. 13 and 14, described below.

[0170] At block 606, control circuit 80 may determine an adaptive threshold based on the two clusters identified at block 604. Control circuit 80 may determine the centroid of each of the two clusters at block 606. Control circuit 80 may identify the cluster with the earlier centroid with respect to time, e.g., the cluster having a centroid at a time point that is earlier than the centroid of the second cluster. The centroid of the cluster that is at the earlier time point may be determined by control circuit 80 as a center metric used for determining the adaptive threshold against which the ER signal features of the second cluster can be compared to for detecting an ER morphology change.

[0171] For example, the data points of the second cluster can be compared to an adaptive threshold determined as the center metric (centroid) of the first cluster (having the earlier centroid with respect to time) plus or minus a variability metric. The variability metric may be a specified value in some examples that defines a minimum distance between the first cluster center metric and the second cluster centroid required to detect an ER morphology change. In other examples, the variability metric may be an updated variability metric determined as an average absolute distance of the data points of the first cluster to the centroid of the first cluster. An adaptive threshold may be determined as the center metric (first cluster centroid) plus and / or minus a multiple of the updated variability metric determined from the data points of the first cluster.

[0172] Control circuit 80 may detect an ER morphology change at block 610 when change criteria are met at block 608. In some examples, the change criteria are met if at least a threshold number or percentage of the second cluster data points are outside the adaptive threshold from the center metric of the first cluster and later in time than the latest data point of the first cluster. In other examples, control circuit 80 may determine that the change criteria are met at block 608 when the centroid of the second cluster is outside the adaptive threshold from than the first cluster center metric and all time points of the second cluster data points are later than the latest time point of the first cluster.

[0173] In still another example, the vector extending from the center metric of the first cluster to the centroid of the second cluster may be determined. If the length of the vector is greater than an adaptive threshold magnitude and the angle of the vector is within a threshold range (indicating at least a threshold time separation), control circuit 80 may detect the ER morphology change.

[0174] If the ER morphology change is detected at block 610, control circuit 80 may perform a response to the change detection at block 612 according to any of the examples described herein. If the ER morphology change criteria are not met at block 608, control circuit 80 may return to block 602 to wait for the next morphology monitoring test to obtain the ER signal feature(s) at the next time point and update the clusters, redetermine the centroids, and identify the earliest centroid as the updated center metric for redetermining the adaptive threshold.

[0175] FIG. 13 is a diagram 650 of the ER morphology match score (y-axis) plotted over time (x-axis). The ER morphology match scores are determined by control circuit 80 for each time point of a morphology monitoring test as described above. Control circuit 80 may buffer each of the data points as an (x,y) coordinate in memory 82, where time is the x-coordinate and the morphology match score is the y-coordinate, for example. As each new morphology monitoring test is performed, a data point is added to the set of data points stored in memory 82.

[0176] Control circuit 80 may perform a clustering method for determining an adaptive threshold against which data points can be compared to for detecting an ER morphology change. One example of a clustering method that may be performed by control circuit 80 is an agglomerative single linkage clustering method, though other methods may be used. In the agglomerative single linkage clustering method, control circuit 80 may begin bydetermining the Euclidean distance between each data point and each of the other data points in the data set (where each single data point defines a cluster at the beginning of the clustering operation). The Euclidean distance between each combination of two data points in the set of data points is buffered in memory 82. The shortest Euclidean distance between two data points is identified. Those two data points separated by the shortest distance can be merged into a cluster, e.g., cluster 1. The next shortest minimum Euclidean distance is identified between the clusters, and those two clusters can be merged into a cluster (e.g., cluster 2). As described above, the minimum Euclidean distance between a pair of clusters can be determined using one point from each cluster that together represent the shortest distance between the points of the two clusters. The shortest minimum Euclidean distance can be identified by control circuit 80 and the associated two clusters can be merged together. Cluster 3 represents the third shortest minimum Euclidean distance in the set of data points. In the example shown, the next shortest minimum Euclidean distance is between cluster 3 and the nearest data point, resulting in cluster 4. This process of identifying the shortest minimum Euclidean distance between pairs of clusters and merging those clusters continues until two clusters remain. Depending on the minimum Euclidean distances between pairs of clusters, merging of different sized clusters can occur such that the final two clusters can have differing numbers of data points.

[0177] FIG. 14 is a diagram 680 of the data set shown in FIG. 13 after control circuit 80 has completed the clustering method to obtain two clusters 682 and 686. Control circuit 80 may determine the centroids 684 and 688 of the two clusters 682 and 686, respectively. The centroids 684 and 688 may be determined as the mean x- and y-values of the data points in each respective cluster 682 and 686. Control circuit 80 may compare the time coordinate (x-coordinate in this example) of the two centroids 684 and 688 to determine which cluster 682 or 686 has an earlier centroid with respect to time. In this example, the centroid 684 of cluster 682 is earlier with respect to time. Control circuit 684 determines the centroid 684 as the as the updated center metric for determining an adaptive threshold 690 against which the data points of the second, later cluster 686 are compared to for detecting an ER morphology change.

[0178] In some examples, the adaptive threshold 690 is determined as a morphology matching score variability metric (or multiple thereof) less than the center metric 684 (inthe y-direction). The variability metric used to determine the adaptive threshold 690 as a distance from the center metric 684 may be a specified variability metric stored in memory 82, which may be programmable. In other examples the variability metric may be updated after performing morphology monitoring tests by determining the average absolute difference in the y-direction between the data points of the first cluster 682 and the center metric 685. Control circuit 80 may determine the adaptive threshold 690 as the y-value of the center metric 684 less the variability metric or multiple thereof.

[0179] Control circuit 80 may detect the ER morphology change when at least a threshold percentage of the data points in the second cluster 688 are later than the latest data point 692 of the first cluster 682 and the second cluster centroid 688 falls outside the adaptive threshold 690 relative to centroid 684 (in the y-direction in this example). Thus, the clustering operation may result in an adaptive threshold 690 determined from the earliest centroid 684, against which the y-coordinate of the later centroid 686 is compared. The clustering operation may further result in a time threshold, which may be determined as the x-coordinate of the latest data point 692 of the earliest cluster 682, against which the x- coordinates of the data points of the second, later cluster 686 may be compared.

[0180] All or at least a specified threshold percentage of the second cluster data points may be required to be later than the latest data point 692. In other examples, control circuit 80 may determine the x-distance between the time coordinates of the center metric 684 and the centroid 688 of second cluster 686 to verify that the centroids 684 and 688 are sufficiently separated in time to detect an ER morphology change. If the center metric 684 and centroid 688 are within 3 days of each other, for example, when daily morphology monitoring tests are performed, control circuit 80 may not detect an ER morphology change even if the centroid 688 falls below the adaptive threshold 690. In this way, the y- coordinate of centroid 688 of the second cluster 686 may be compared against an adaptive threshold 690 determined from the center metric (centroid) 684 of the first cluster 682, and the x-coordinate of centroid 688 of the second cluster 686 may be compared against a time threshold determined from the x-coordinate of the center metric 684 (e.g., the centroid x- coordinate plus a specified number of days) or compared against the latest time point 692.

[0181] In still other examples, control circuit 80 may define an adaptive threshold region 694 around the earliest centroid determined as the center metric 684. In this example, the adaptive threshold region 694 is shown as a rectangular region, however the region couldbe elliptical, circular or other 2D shape. Control circuit 80 may determine if at least a threshold percentage of the data points of the second, later cluster 686 are outside the adaptive threshold region 694 defined relative to the center metric 684 for detecting an ER morphology change. The adaptive threshold region 694 may or may not include all of the points of the first cluster 682 having the earliest centroid, center metric 684. The adaptive threshold region 694 may or may not be centered on the centroid 684 in the x- and / or y- directions. However, the adaptive threshold region 694 may be updated based on the earliest centroid with respect to time and / or a variability metric at each morphology monitoring time point.

[0182] Further it is noted that while time is assigned to the x-axis in FIG. 14 and morphology match score is assigned to the y-axis, the axes could be reversed such that time is the y-coordinate and the ER morphology feature is the x-axis such that the centroid having the smaller y-coordinate is the “earliest” centroid that is determined as the updated center metric used for determining the adaptive threshold against which the centroid having the higher y-coordinate and / or the datapoints of the second cluster are compared.

[0183] FIG. 15 is a diagram 700 of ER signal morphology feature data points determined during morphology monitoring tests plotted in three dimensions. FIG. 14 shows an example of an adaptive threshold determined using the centroid of a cluster of data points defined by time along one axis and one ER signal morphology feature, e.g., morphology match score, along the second axis. It is contemplated that two different ER signal morphology features may be determined for each morphology monitoring test and plotted with time in a 3D coordinate system. In other examples, an N-dimensional analysis may be performed by determining N-l ER signal features (with time being the Nth dimension) and determining the clusters in an N-dimensional space.

[0184] As shown in FIG. 15, the time of the morphology monitoring tests may be plotted along one axis, e.g., the x-axis 702, with one ER signal feature, e.g., morphology match score, plotted along a second axis, e.g., the y-axis 704, and a second ER signal feature, e.g., VAT metric, plotted along the third axis, e.g., the z-axis 706. Each of the data points 708 represents the ER morphology match score and VAT metric determined at a given time point of the respective morphology monitoring test.

[0185] Control circuit 80 may perform a clustering operation on the 3D data points 708 defined in (x, y, z) coordinates, e.g., (time, morphology match score, VAT). As describedabove, control circuit 80 may perform an agglomerative single linkage clustering method that includes starting with each data point as a single point cluster, determining the 3D minimum Euclidean distances between pairs of clusters and merging the closest pairs of clusters until two clusters remain. Control circuit 80 may determine the centroids of the two clusters, e.g., (mean time, mean morphology match score, mean VAT) to identify the 3D cluster having the earliest (lowest x-coordinate) centroid. The earliest centroid may be determined by control circuit 80 as the updated center metric used to determine an adaptive threshold against which data points in the second cluster and / or the second cluster centroid can be compared against for detecting an ER morphology change.

[0186] FIG. 16 is a diagram 750 of the projection of the data points 708 shown in FIG. 15 onto the 2D y-z plane defined by the morphology match scores along the y-axis 704 and the VATs along the z-axis 706. After the clustering operation is performed, control circuit 80 may identify a first cluster 720 having a centroid 722 that is at an earlier time (along the x-axis not shown in FIG. 16) and the second cluster 730 having centroid 732 that is at a later time along the x-axis. An adaptive threshold 740 may be determined as a threshold 2D distance from the center metric 722 defined as the centroid of the earliest cluster 720. If the second cluster centroid 732 is outside the adaptive threshold distance 740 from the center metric 722, control circuit 80 may detect the ER morphology change. Additionally, in some examples, control circuit 80 may determine if at least a threshold percentage of the data points in the second cluster 730 have an x-coordinate that is greater than the greatest x-coordinate of the data points in the first cluster 720. For instance, if at least 70%, 80%, 90% or 100%, as non-limiting examples, of the data points in the second cluster 730 have an x-coordinate that is greater than the maximum x-coordinate of the data points in the first cluster 720, and the second cluster centroid 732 is outside the adaptive threshold distance 740 from the center metric 722, control circuit 80 may detect the ER morphology change. In some examples, the Euclidean distance between the 2D centroid 732 and the 2D center metric 722 is determined and compared to the adaptive threshold distance 740.

[0187] In other examples, control circuit 80 may determine the 3D Euclidean distance between the centroids of the two clusters and compare it to a threshold distance. In still other examples, control circuit 80 may compare each of the ID distances, e.g., the x- coordinate distance, the y-coordinate distance and the z-coordinate distance between thecentroids 722 and 732 in three dimensions to respective thresholds or ranges for detecting the ER morphology change when the respective thresholds are met. In still other examples, control circuit 80 may determine an adaptive threshold region around the earlier centroid 722 and determine if at least a threshold number of data points of the second cluster 730 and / or the second centroid 732 are outside the adaptive threshold region. An adaptive threshold distance, range or region may be determined in one, two or all three dimensions, respectively, from the data points of the earliest centroid 722 against which the second cluster data points are compared for detecting the ER morphology change.

[0188] In the examples of FIGs. 12-16 involving performing a clustering operation for determining an updated center metric, control circuit 80 may perform the clustering operation after each new morphology monitoring test is performed. In some examples, when a new data point is determined, the minimum distance from the new data point to the nearest point in each of the two previously determined clusters may be determined. The new data point may be merged with the cluster for which the shortest minimum distance is found.

[0189] In still other examples, the first cluster may be established during an initialization phase then updated using data points that are within the adaptive threshold from the centroid of the first cluster with respect to the ER signal morphology feature(s). If the new data point is within the adaptive threshold distance of the first cluster centroid with respect to the ER signal morphology feature(s), the new data point may be added to the first cluster and used to update the center metric of the first cluster and update the adaptive threshold distance. If the new 2D data point is more than the adaptive threshold distance from the centroid of the first cluster with respect to the ER signal morphology feature coordinates, the new data point may be added to the second cluster and not used to update the center metric determined as the centroid of the first cluster. The adaptive threshold distance remains at the same, previous value based on the first cluster center metric. Control circuit 80 may determine a new second cluster centroid for comparison to the adaptive threshold distance.

[0190] Further disclosed herein is the subject matter of the following examples:

[0191] Example 1. A medical device system including a sensing circuit configured to sense a first cardiac electrical signal, a therapy delivery circuit configured to deliver pacing pulses and a control circuit configured to, for each of a plurality of morphologymonitoring time points, determine a first evoked response signal feature from one or more evoked response signals of the first cardiac electrical signal sensed after a respective pacing pulse delivered by the therapy delivery circuit. The control circuit may be further configured to determine a first center metric of a first portion of the first evoked response signal features determined for the plurality of morphology monitoring time points, determine an adaptive threshold based on the first center metric, compare the adaptive threshold to a second portion of the first evoked response signal features, where the second portion of the first evoked response signal features different than the first portion of the first evoked response signal features and the second portion of the first evoked response signal features include multiple morphology monitoring time points of the plurality of morphology monitoring time points. The control circuit may detect an evoked response morphology change in response to at least a threshold percentage of the second portion of the first evoked response signal features falling outside the adaptive threshold from the first center metric.

[0192] Example 2. The medical device of example 1 wherein the control circuit is further configured to determine a variability metric of the first portion of the first evoked response signal features and determine the adaptive threshold as a combination of the first center metric and the variability metric.

[0193] Example 3. The medical device of any one of examples 1 — 2 wherein the control circuit is further configured to determine a new first evoked response signal feature for a new morphology monitoring time point, determine that the new first evoked response signal feature is within the adaptive threshold from the first center metric, and update the first center metric as a weighted combination of the first center metric and the new first evoked response signal feature when the new evoked response signal feature is determined to be within the adaptive threshold from the first center metric. The control circuit may update the adaptive threshold using the updated first center metric.

[0194] Example 4. The medical device of example 3 wherein the control circuit is further configured to determine a distance from the new first evoked response signal feature and the updated first center metric, determine an updated variability metric as a weighted combination of the variability metric and the distance, and update the adaptive threshold as a combination of the updated first center metric and the updated variability metric.

[0195] Example 5. The medical device of any one of examples 1 — 4 wherein the control circuit is further configured to compare the adaptive threshold to the second portion of the first evoked response signal features by determining a second center metric from the second portion of the first evoked response signal features and comparing the second center metric to the adaptive threshold.

[0196] Example 6. The medical device of any one of examples 1 — 5 wherein the control circuit is further configured to perform a clustering operation to identify the first portion of the first evoked response signal features as a first cluster of the plurality of morphology monitoring time points and identify the second portion of the first evoked response signal features as a second cluster of the plurality of morphology monitoring time points. The control circuit may determine the first center metric as a first centroid of the first cluster.

[0197] Example 7. The medical device of example 6 wherein the control circuit is further configured to determine a second centroid of the second cluster and identify the first portion as the first cluster by determining that the first centroid of the first cluster is earlier with respect to the time than the second centroid of the second cluster.

[0198] Example 8. The medical device of any one of examples 6 — 7 wherein the control circuit is further configured to determine the adaptive threshold based on the first center metric as a distance from the first center metric, determine a second centroid of the second cluster and compare the adaptive threshold to the second portion of the evoked response signal features by comparing the second centroid to the adaptive threshold.

[0199] Example 9. The medical device of any one of examples 1 — 8 wherein the control circuit is further configured to, during an initialization phase, determine a first value of the first center metric from a nominal starting threshold and at least one of the first evoked response signal features.

[0200] Example 10. The medical device of any one of examples 1 — 9 wherein the control circuit is further configured to determine a second evoked response signal feature different than the first evoked response signal feature at each of the plurality of morphology monitoring time points and determine the first center metric in two dimensions defined by the first portion of the first evoked response signal features and a corresponding first portion of the second evoked response signal features.

[0201] Example 11. The medical device of example 10 wherein the sensing circuit is further configured to sense a second cardiac electrical signal different than the first cardiacelectrical signal, and the control circuit is further configured to determine the second evoked response signal feature from the second cardiac electrical signal.

[0202] Example 12. The medical device system of any one of examples 1 — 11 wherein the therapy delivery circuit is configured to deliver the pacing pulses as conduction system pacing pulses.

[0203] Example 13. The medical device system of any one of examples 1 — 12 wherein the control circuit is further configured to adjust a pacing control parameter in response to detecting the evoked response morphology change and control the therapy delivery circuit to deliver pacing pulses according to the adjusted pacing control parameter.

[0204] Example 14. The medical device system of example 13 wherein the control circuit is further configured to adjust the pacing control parameter by adjusting at least one of: a pacing electrode vector, a pacing pulse output, a pacing mode, or a pacing interval.

[0205] Example 15. The medical device system of any one of examples 1 — 14 wherein the therapy delivery circuit is configured to deliver the pacing pulses via a pacing electrode vector, and the control circuit is further configured to detect a change in a location of the pacing electrode vector in response to detecting the evoked response morphology change.

[0206] Example 16. The medical device of any one of examples 1 — 15 wherein the control circuit is further configured to determine the first evoked response signal features from the first cardiac electrical signal sensed during an evoked response time window following one or more respective pacing pulses for each of the plurality of morphology monitoring time points.

[0207] Example 17. The medical device of any one of examples 1 — 16 wherein the control circuit is further configured to determine the first evoked response signal feature as one of a morphology match score or a ventricular activation time.

[0208] Example 18. The medical device of any one of examples 1 — 17 further comprising a communication circuit configured to transmit a notification signal in response to the control circuit detecting the evoked response morphology change.

[0209] Example 19. The medical device system of example 18 wherein the control circuit is further configured to detect a change in pacing substrate in response to detecting the evoked response morphology change, and the communication circuit is further configured to transmit the notification as a pacing substrate change alert.

[0210] Example 20. The medical device system of example 18 wherein the control circuit is further configured to detect a change in a pacing electrode location in response to detecting the evoked response morphology change, and the communication circuit is further configured to transmit the notification as a pacing electrode location change alert.

[0211] Example 21. A method comprising sensing a first cardiac electrical signal, delivering pacing pulses, and, for each of a plurality of morphology monitoring time points, determining a first evoked response signal feature from one or more evoked response signals of the first cardiac electrical signal sensed after a respective delivered pacing pulse. The method may further include determining a first center metric of a first portion of the first evoked response signal features determined for the plurality of morphology monitoring time points, determining an adaptive threshold based on the first center metric, and comparing the adaptive threshold to a second portion of the first evoked response signal features, where the second portion of the first evoked response signal features is different than the first portion of the first evoked response signal features and the second portion of the first evoked response signal features includes multiple morphology monitoring time points of the plurality of morphology monitoring time points. The method may further include detecting an evoked response morphology change in response to at least a threshold percentage of the second portion of the first evoked response signal features falling outside the adaptive threshold from the first center metric.

[0212] Example 22. The method of example 21 further including determining a variability metric of the first portion of the first evoked response signal features and determining the adaptive threshold as a combination of the first center metric and the variability metric.

[0213] Example 23. The method of any one of examples 21 — 22 further including determining a new first evoked response signal feature for a new morphology monitoring time point, determining that the new first evoked response signal feature is within the adaptive threshold from the first center metric, updating the first center metric as a weighted combination of the first center metric and the new first evoked response signal feature when the new evoked response signal feature is determined to be within the adaptive threshold from the first center metric, and updating the adaptive threshold using the updated first center metric.

[0214] Example 24. The method of example 23 further including determining a distance from the new first evoked response signal feature and the updated first center metric, determining an updated variability metric as a weighted combination of the variability metric and the distance, and updating the adaptive threshold as a combination of the updated first center metric and the updated variability metric.

[0215] Example 25. The method of any one of examples 21 — 24 further including comparing the adaptive threshold to the second portion of the first evoked response signal features by determining a second center metric from the second portion of the first evoked response signal features and comparing the second center metric to the adaptive threshold.

[0216] Example 26. The method of any one of examples 21 — 25 further including performing a clustering operation to identify the first portion of the first evoked response signal features as a first cluster of the plurality of morphology monitoring time points and identify the second portion of the first evoked response signal features as a second cluster of the plurality of morphology monitoring time points. The method may further include determining the first center metric as a first centroid of the first cluster.

[0217] Example 27. The method of example 26 further including determining a second centroid of the second cluster and identifying the first portion as the first cluster by determining that the first centroid of the first cluster is earlier with respect to the time than the second centroid of the second cluster.

[0218] Example 28. The method of any one of examples 26 — 27 further including determining the adaptive threshold based on the first center metric as a distance from the first center metric, determining a second centroid of the second cluster, and comparing the adaptive threshold to the second portion of the evoked response signal features by comparing the second centroid to the adaptive threshold.

[0219] Example 29. The method of any one of examples 21 — 28 further including, during an initialization phase, determining a first value of the first center metric from a nominal starting threshold and at least one of the first evoked response signal features.

[0220] Example 30. The method of any one of examples 21 — 29 further including determining a second evoked response signal feature different than the first evoked response signal feature at each of the plurality of morphology monitoring time points and determining the first center metric in two dimensions defined by the first portion of thefirst evoked response signal features and a corresponding first portion of the second evoked response signal features.

[0221] Example 31. The method of example 30 further including sensing a second cardiac electrical signal different than the first cardiac electrical signal and determining the second evoked response signal feature from the second cardiac electrical signal.

[0222] Example 32. The method of any one of examples 21 — 31 further including delivering the pacing pulses as conduction system pacing pulses.

[0223] Example 33. The method of any one of examples 21 — 32 further including adjusting a pacing control parameter in response to detecting the evoked response morphology change and delivering pacing pulses according to the adjusted pacing control parameter.

[0224] Example 34. The method of example 33 wherein adjusting the pacing control parameter includes adjusting at least one of a pacing electrode vector, a pacing pulse output, a pacing mode, or a pacing interval.

[0225] Example 35. The method of any one of examples 21 — 34 further including delivering the pacing pulses via a pacing electrode vector and detecting a change in a location of the pacing electrode vector in response to detecting the evoked response morphology change.

[0226] Example 36. The method of any one of examples 21 — 35 further including determining the first evoked response signal features from the first cardiac electrical signal sensed during an evoked response time window following one or more respective pacing pulses for each of the plurality of morphology monitoring time points.

[0227] Example 37. The method of any one of examples 21 — 36 wherein determining the first evoked response signal feature includes determining one of a morphology match score or a ventricular activation time.

[0228] Example 38. The method of any one of examples 21 — 37 further including transmitting a notification signal in response to detecting the evoked response morphology change.

[0229] Example 39. The method of example 38 further including detecting a change in pacing substrate in response to detecting the evoked response morphology change and transmitting the notification as a pacing substrate change alert.

[0230] Example 40. The method of example 38 further including detecting a change in a pacing electrode location in response to detecting the evoked response morphology change and transmitting the notification as a pacing electrode location change alert.

[0231] Example 41. A non-transitory computer readable medium storing a set of instructions that, when executed by control circuitry of a medical device system, cause the medical device system to sense a cardiac electrical signal, deliver pacing pulses and, for each of a plurality of morphology monitoring time points, determine an evoked response signal feature from one or more evoked response signals of the cardiac electrical signal sensed after a respective delivered pacing pulse. The instructions may further cause the medical device system to determine a center metric of a first portion of the evoked response signal features determined for the plurality of morphology monitoring time points, determine an adaptive threshold based on the center metric, and compare the adaptive threshold to a second portion of the evoked response signal features, where the second portion of the evoked response signal features is different than the first portion of the evoked response signal features and the second portion of the evoked response signal features includes multiple morphology monitoring time points of the plurality of morphology monitoring time points. The instructions may further cause the medical device system to detect an evoked response morphology change in response to at least a threshold percentage of the second portion of the evoked response signal features falling outside the adaptive threshold from the center metric. The instructions may further cause the medical device system to transmit a notification signal in response to detecting the evoked response morphology change.

[0232] It should be understood that, depending on the example, certain acts or events of any of the methods described herein can be performed in a different sequence, in parallel, may be added, merged, or left out altogether (e.g., not all described acts or events are necessary for the practice of the method). Moreover, in certain examples, acts or events may be performed concurrently, e.g., through multi-threaded processing, interrupt processing, or multiple processors, rather than sequentially. In addition, while certain aspects of this disclosure are described as being performed by a single processor, circuit or unit for purposes of clarity, it should be understood that the techniques of this disclosure may be performed by a combination of processors, units or circuits associated with, for example, a medical device system.

[0233] In one or more examples, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored as one or more instructions or code on a computer-readable medium and executed by one or more hardware-based processing units. Computer- readable media may include computer-readable storage media, which corresponds to a tangible medium such as data storage media (e.g., RAM, ROM, EEPROM, flash memory, or any other medium that can be used to store desired program code in the form of instructions or data structures and that can be accessed by a computer).

[0234] Instructions may be executed by one or more processors, such as one or more digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), or other equivalent integrated or discrete logic circuitry. Accordingly, the term “processor,” as used herein may refer to any of the foregoing structure or any other structure suitable for implementation of the techniques described herein. Also, the techniques could be fully implemented in one or more circuits or logic elements.

[0235] Thus, a medical device system has been presented in the foregoing description with reference to specific examples. It is to be understood that various aspects disclosed herein may be combined in different combinations than the specific combinations presented in the accompanying drawings. It is appreciated that various modifications to the referenced examples may be made without departing from the scope of the disclosure and the following claims.

Claims

WHAT IS CLAIMED IS:

1. A medical device system comprising: a sensing circuit configured to sense a first cardiac electrical signal; a therapy delivery circuit configured to deliver pacing pulses; and a control circuit configured to: for each of a plurality of morphology monitoring time points, determine a first evoked response signal feature from one or more evoked response signals of the first cardiac electrical signal sensed after a respective pacing pulse delivered by the therapy delivery circuit; determine a first center metric of a first portion of the first evoked response signal features determined for the plurality of morphology monitoring time points; determine an adaptive threshold based on the first center metric; compare the adaptive threshold to a second portion of the first evoked response signal features, the second portion of the first evoked response signal features different than the first portion of the first evoked response signal features, the second portion of the first evoked response signal features comprising multiple morphology monitoring time points of the plurality of morphology monitoring time points; and detect an evoked response morphology change in response to at least a threshold percentage of the second portion of the first evoked response signal features falling outside the adaptive threshold from the first center metric.

2. The medical device of claim 1 wherein the control circuit is further configured to: determine a variability metric of the first portion of the first evoked response signal features; and determine the adaptive threshold as a combination of the first center metric and the variability metric.

3. The medical device of any one of claims 1 — 2 wherein the control circuit is further configured to: determine a new first evoked response signal feature for a new morphology monitoring time point;determine that the new first evoked response signal feature is within the adaptive threshold from the first center metric; update the first center metric as a weighted combination of the first center metric and the new first evoked response signal feature when the new evoked response signal feature is determined to be within the adaptive threshold from the first center metric; and update the adaptive threshold using the updated first center metric.

4. The medical device of claim 3 wherein the control circuit is further configured to: determine a distance from the new first evoked response signal feature and the updated first center metric; determine an updated variability metric as a weighted combination of the variability metric and the distance; and update the adaptive threshold as a combination of the updated first center metric and the updated variability metric.

5. The medical device of any one of claims 1 — 4 wherein the control circuit is further configured to compare the adaptive threshold to the second portion of the first evoked response signal features by: determining a second center metric from the second portion of the first evoked response signal features; and comparing the second center metric to the adaptive threshold.

6. The medical device of any one of claims 1 — 5 wherein the control circuit is further configured to: perform a clustering operation to: identify the first portion of the first evoked response signal features as a first cluster of the plurality of morphology monitoring time points; and identify the second portion of the first evoked response signal features as a second cluster of the plurality of morphology monitoring time points; and determine the first center metric as a first centroid of the first cluster.

7. The medical device of claim 6 wherein the control circuit is further configured to:determine a second centroid of the second cluster; and identify the first portion as the first cluster by determining that the first centroid of the first cluster is earlier with respect to the time than the second centroid of the second cluster.

8. The medical device of any one of claims 1 — 7 wherein the control circuit is further configured to: determine a second evoked response signal feature different than the first evoked response signal feature at each of the plurality of morphology monitoring time points; and determine the first center metric in two dimensions defined by the first portion of the first evoked response signal features and a corresponding first portion of the second evoked response signal features.

9. The medical device of claim 8 wherein: the sensing circuit is further configured to sense a second cardiac electrical signal different than the first cardiac electrical signal; and the control circuit is further configured to determine the second evoked response signal feature from the second cardiac electrical signal.

10. The medical device system of any one of claims 1 — 9 wherein the therapy delivery circuit is configured to deliver the pacing pulses as conduction system pacing pulses.

11. The medical device system of any one of claims 1 — 10 wherein the control circuit is further configured to: adjust a pacing control parameter in response to detecting the evoked response morphology change; and control the therapy delivery circuit to deliver pacing pulses according to the adjusted pacing control parameter.

12. The medical device of any one of claims 1 — 11 wherein the control circuit is further configured to determine the first evoked response signal feature as one of: a morphology match score; ora ventricular activation time.

13. The medical device of any one of claims 1 — 12 further comprising a communication circuit configured to transmit a notification signal in response to the control circuit detecting the evoked response morphology change.

14. The medical device system of claim 13 wherein: the control circuit is further configured to detect a change in pacing substrate in response to detecting the evoked response morphology change; and the communication circuit is further configured to transmit the notification as a pacing substrate change alert.

15. The medical device system of claim 14 wherein: the therapy delivery circuit is further configured to deliver the pacing pulses via a pacing electrode; and the control circuit is further configured to detect a change in a location of the pacing electrode in response to detecting the evoked response morphology change; and the communication circuit is further configured to transmit the notification as a pacing electrode location change alert.

Images