Template-based artifact reduction in neuromodulation applications

Abstract

Methods and systems for recording evoked potentials evoked during electrical stimulation of nervous tissue of a patient are disclosed. These methods and systems can be used in therapeutic regimens, such as spinal cord stimulation (SCS). The disclosed methods and systems allow for reduction or removal of stimulation artifacts in recorded signals. The remainder of the stimulation artifacts may be modeled and subtracted from the recorded signal that includes the artifacts and the neural response contributions.

Description

Technical Field

[0001] The present application relates to implantable medical devices (IMDs), and more particularly to circuitry that assists in sensing neural responses to stimulation in implantable stimulator devices. Background Art

[0002] Implantable neurostimulator devices are devices that generate and deliver electrical stimulation to the body's nerves and tissues to treat various biological diseases, such as pacemakers for treating arrhythmias; defibrillators for treating cardiac fibrillation; cochlear stimulators for treating deafness; retinal stimulators for treating blindness; muscle stimulators for producing coordinated limb movements; spinal cord stimulators for treating chronic pain; cortical and deep brain stimulators for treating movement and psychological disorders; and other neurostimulators for treating urinary incontinence, sleep apnea, shoulder subluxation, etc. The following description will primarily focus on the application of the present invention in spinal cord stimulation (SCS) or deep brain stimulation (DBS) systems. However, the present invention can be applied to any stimulator device system.

[0003] The stimulator system generally includes an implantable pulse generator (IPG) 10 shown in Figure 1. The IPG 10 includes a biocompatible device housing 12 that houses circuitry and a battery 14 for powering the IPG functions. The IPG 10 is coupled to tissue stimulation electrodes 16 via one or more electrode leads forming an electrode array 17. For example, one or more percutaneous leads 15 having annular or split-ring electrodes 16 carried on a flexible body 18 can be used. In another example, a paddle-shaped lead 19 provides an electrode 16 located on one of its generally flat surfaces. Wires 20 within the lead are coupled to the electrodes 16 and to proximal contacts 21 in a lead connector 22 that can be inserted into a header 23 secured to the IPG 10, which header may include, for example, epoxy. Once inserted, the proximal contacts 21 connect to header contacts 24 within the lead connector 22, which in turn are coupled to stimulation circuitry 28 within the housing 12 by feedthrough pins 25 through a housing feedthrough 26.

[0004] In the IPG 10 shown, there are 32 electrodes (E1-E32), distributed between four percutaneous leads 15, or contained on a single paddle lead 19, so that the head 23 can include eight electrode lead connectors 22 in a 2x2 array. However, the type and number of leads and the number of electrodes in the IPG vary depending on the application and therefore may vary. The conductive housing 12 or some conductive portion of the housing may also include electrodes (Ec). In SCS applications, one or more electrode leads are typically implanted in the patient's spinal cord near the dura mater in the spine, preferably spanning the left and right sides of the patient's spine. Proximal contacts 21 are tunneled through the patient's tissue to a distal location (such as the buttocks where the IPG housing 12 is implanted) where they are coupled to the lead connectors 22. In DBS applications, the electrode leads are implanted in the brain through a hole in the skull, and lead extensions are used to connect the leads to the IPG, which is typically implanted below the clavicle (neck bone). In other IPG examples designed for direct implantation at the site requiring stimulation, the IPG may be leadless, with electrodes 16 present instead on the body of the IPG 10 for contacting the patient's tissue. In other solutions, one or more IPG leads may be integrated with the IPG 10 and permanently connected to the IPG 10. SCS therapy can relieve symptoms such as chronic back pain, while DBS therapy can relieve Parkinson's disease symptoms such as tremors and stiffness. The described IPG 10 should be understood to include an external test stimulator (ETS) that simulates the operation of the IPG 10 during a test when the leads have been implanted in the patient but the IPG 10 has not yet been implanted. See, for example, USP 9,259,574 (disclosing an ETS).

[0005] The IPG 10 may include an antenna 27a that allows it to communicate bidirectionally with multiple external devices discussed subsequently. As shown, the antenna 27a includes a conductive coil within the housing 12, although the coil antenna 27a may also be present in the head 23. When the antenna 27a is configured as a coil, near-field magnetic induction is preferably used to communicate with the external device. The IPG 10 may also include a radio frequency (RF) antenna 27b. In FIG1 , the RF antenna 27b is shown within the head 23, but it may also be within the housing 12. The RF antenna 27b may include a patch, a slot, or a wire and may operate as a monopole or a dipole. The RF antenna 27b preferably communicates using far-field electromagnetic waves and may operate according to any number of known RF communication standards (such as Bluetooth, Zigbee, WiFi, MICS, etc.).

[0006] Stimulation in the IPG 10 is typically provided by pulses, each of which may include multiple phases (30i), as shown in the example of FIG2A . Stimulation parameters typically include amplitude (current I, but voltage amplitude V may also be used); frequency (F); pulse width (PW); the electrodes 16 selected to provide stimulation; and the polarity of such selected electrodes, i.e., whether they act as anodes supplying current to tissue or as cathodes sinking current from tissue. These and possibly other stimulation parameters employed collectively comprise a stimulation program that the stimulation circuitry 28 in the IPG 10 can execute to provide therapeutic stimulation to the patient.

[0007] In the example of FIG2A , electrode E1 has been selected (during its first phase 30 a) as an anode, thereby supplying a pulse of positive current of amplitude +I to the tissue. Electrode E2 has been selected (also during the first phase 30 a) as a cathode, thereby providing a pulse of negative current of amplitude -I that is drawn from the tissue. This is an example of bipolar stimulation, in which the lead includes an anode and a cathode. Note that, as further explained in USP 10,881,859, more than one electrode on the lead can be selected as an anode electrode to form an anode at a given time, and more than one electrode can be selected as a cathode electrode to form a cathode at a given time. The stimulation provided by the IPG 10 can also be unipolar. In unipolar stimulation, the lead is programmed to be a unipolar electrode with a given polarity (e.g., the cathode pole), and the conductive housing electrode Ec acts as a loop (e.g., the anode pole). Similarly, during unipolar stimulation, more than one electrode on the lead can be active to form a pole.

[0008] The IPG 10 described above includes a stimulation circuit 28 to provide a prescribed stimulation at the patient's tissue. FIG3 shows an example of a stimulation circuit 28, which includes one or more current supply circuits and one or more current sink circuits. The supply and sink circuits may include digital-to-analog converters (DACs) and may be referred to as PDACs and NDACs based on the positive (supplied, anode) and negative (sinked, cathode) currents they emit, respectively. In the example shown, the NDACi / PDACi pairing is dedicated to (hard-wired to) a specific electrode node ei 39. For reasons explained below, each electrode node ei 39 is associated with an electrode Ei 16 via a DC blocking capacitor Ci 38. The stimulation circuit 28 in this example also supports the selection of a conductive housing 12 as an electrode (Ec12), which, as described above, is typically selected for monopolar stimulation. The PDACs and NDACs may also include a voltage source.

[0009] Appropriate control of the PDAC and NDAC allows any electrode 16 to act as an anode or cathode to generate a current through the patient's tissue R, hopefully with a good therapeutic effect. Consistent with the example provided in FIG2A , FIG3 illustrates operation during a first phase 30 a, in which electrode E1 has been selected as an anode electrode that supplies current I to the tissue R, and E2 has been selected as a cathode electrode that sinks current from the tissue. Thus, PDAC1 and NDAC2 are digitally programmed (e.g., according to a specified frequency and pulse width) to generate the desired current I with the correct timing. As described above, more than one anode electrode and more than one cathode electrode can be selected at a time, so current can flow through the tissue R between two or more electrodes 16. Other stimulation circuits 28 can also be used with the IPG 10, including stimulation circuits that include a switch matrix between the electrode node ei 39 and the N / PDAC. For example, see 6,181,969, 8,606,362, 8,620,436, 11,040,192, and 10,912,942. Most of the stimulation circuitry 28 of FIG. 3 (including the PDAC and NDAC, the switch matrix (if present), and the electrode nodes ei 39) can be integrated on one or more application-specific integrated circuits (ASICs), as described in U.S. Patent Application Publications 2012 / 0095529, 2012 / 0092031, and 2012 / 0095519. As explained in these references, the one or more ASICs can also include other circuits useful in the IPG 10, such as the IPG main control circuitry 102 (see FIG. Figure 5 ), telemetry circuit (for off-chip interface connection with telemetry antenna 27a and / or 27b), circuit for generating compliance voltage VH (as described below), various measurement circuits, etc.

[0010] The power for the stimulation circuit 28 is provided by a compliance voltage VH, as described in further detail in U.S. Patent Application Publication Nos. 2013 / 0289665 and 2018 / 0071520. The compliance voltage VH can be coupled to a supply circuit (e.g., one or more PDACs), while ground can be coupled to a sink circuit (e.g., one or more NDACs), such that the stimulation circuit 28 is powered by VH and ground. Other supply voltages can be used for the PDACs and NDACs and are explained in U.S. Patent Application Publication No. 2018 / 0071520, but for simplicity, these are not shown in FIG. 3.

[0011] Preferably, as described in USP 11,040,202, the compliance voltage VH can be generated by a VH regulator 49. VH regulator 49 receives the voltage (Vbat) of battery 14 and boosts it to the higher value required for compliance voltage VH. For example, VH regulator 49 can include an inductor-based boost converter or a capacitor-based charge pump. Regulator 49 can vary the value of VH based on measurements obtained from stimulation circuit 28. As explained in detail in the '202 patent, VH measurement circuit 51 can be used to measure the voltage drop across the active DACs in stimulation circuit 28 (e.g., PDAC1 (Vp1) and NDAC2 (Vn2) in the example shown in FIG. 3). Using such measurements allows VH to be established at an energy-efficient level: high enough to draw the specified current in the absence of a load (i.e., without drawing the specified lower current), but low enough to not unnecessarily waste power in stimulation circuit 28 when drawing the specified current. In this regard, VH can be variable, typically within a range of approximately 5 to 15 volts.

[0012] The VH measurement circuit 51 can output an enable signal VH(en1) that indicates when the VH regulator 49 should increase the level of VH, that is, when the voltage drop across the active DAC is too low. The enable signal VH(en1) can be processed at logic 53 in conjunction with other signals explained below to determine the main enable signal VH(en) for the VH regulator 49. The logic 53 can be associated with the control circuit 102 of the IPG. When the main enable signal VH(en) is asserted, the VH regulator 49 is caused to increase VH (for example, when current begins to load). De-asserting VH(en) disables the VH regulator, which allows VH to naturally decrease over time until it needs to be increased again. This feedback typically causes VH to be established at an energy-saving value suitable for the current provided by the stimulation circuit 28.

[0013] Also shown in FIG3 is a DC blocking capacitor Ci 38 placed in series in the electrode current path between electrode node ei 39 and each of electrodes ei 16 (including housing electrode Ec 12). The DC blocking capacitor 38 serves as a safety measure to prevent DC current from being injected into the patient, which could occur, for example, if there is a circuit fault in the stimulation circuit 28. The DC blocking capacitor 38 is typically provided off-chip (outside one or more ASICs), but can be provided within or on a circuit board within the IPG 10 that is used to integrate its various components, as described in U.S. Patent Application Publication 2015 / 0157861. While useful, the DC blocking capacitor 38 is not strictly required in all IPG designs and applications.

[0014] Referring again to FIG. 2A , the stimulation pulses shown are biphasic, with each pulse comprising a first phase 30a followed by a second phase 30b of opposite polarity. The biphasic pulses can be used to actively restore any charge that may be stored on capacitive elements in the electrode current path (such as the charge on the DC blocking capacitor 38). Charge restoration is illustrated in FIG. 2A and 2B . During the first pulse phase 30a, charge will (primarily) accumulate on the DC blocking capacitors C1 and C2 associated with the electrodes E1 and E2 used to generate the current, thereby generating voltages Vc1 and Vc2 (I = C*dV / dt). During the second pulse phase 30b, when the polarity of the current I reverses at the selected electrodes E1 and E2, the charge stored on capacitors C1 and C2 is restored, so that voltages Vc1 and Vc2 are expected to return to 0V by the end of the second pulse phase 30b.

[0015] Charge recovery using phases 30a and 30b is referred to as "active" because the P / NDAC in stimulation circuit 28 actively drives current, particularly during the final phase 30b, to restore charge stored after the first phase 30a. However, such active charge recovery may not be perfect, and some residual charge may remain in the capacitive structure even after phase 30b is complete. Therefore, stimulation circuit 28 can also provide passive charge recovery. As shown in FIG3 , passive charge recovery is implemented using passive charge recovery switches PRi 41. When selected via assertion of control signal <Xi>, these switches 41 couple each electrode node ei to a passive recovery voltage Vpr established on bus 43. As described in USP 10,716,937 and 10,792,491, this allows any stored charge to be restored by patient tissue R. During period 30c shown in FIG2A , control signal <Xi> is typically asserted to enable passive charge recovery after each pulse (e.g., after each final phase 30b). Because passive charge recovery involves a capacitive discharge through the resistance R of the patient's tissue, such a discharge manifests as an exponential decay of current, as shown in FIG2A . Also as discussed in the '937 patent, each of the passive charge recovery switches 41 can be associated with a variable resistor, such that each switch 41 can be controlled by a bus of signals <Xi> to control the resistance at which passive charge recovery occurs, i.e., the on-resistance when the switch 41 is closed. Passive charge recovery during period 30c can be followed by a quiet period 30d during which the DAC circuit does not drive active current and none of the passive recovery switches 41 is closed. This quiet period 30d can last until the next pulse is actively generated (e.g., phase 30a). As with the details of pulse phases 30a and 30b, the occurrence of passive charge recovery (30c) and any quiet period (30d) can be specified as part of the stimulation program.

[0016] FIG4 illustrates various external systems 60, 70, and 80 that can be in wireless data communication with the IPG 10. Such systems can be used to wirelessly transmit stimulation programs to the IPG 10, i.e., to program its stimulation circuitry 28 to produce stimulation with desired amplitude and timing as previously described. Such systems can also be used to adjust one or more stimulation parameters of a stimulation program currently being executed by the IPG 10, and / or wirelessly receive information from the IPG 10, such as various status information.

[0017] For example, the external controller 60 may be as described in U.S. Patent Application Publication 2015 / 0080982 and may comprise a portable handheld controller specifically designed for use with the IPG 10. The external controller 60 may also comprise a general-purpose mobile electronic device (such as a mobile phone) that has been programmed with a medical device application (MDA) to allow it to function as a wireless controller for the IPG 10, as described in U.S. Patent Application Publication 2015 / 0231402. The external controller 60 includes a display 61 and means for inputting commands, such as buttons 62 or selectable graphical icons provided on the display 61. The user interface of the external controller 60 enables the patient to adjust stimulation parameters, although its functionality may be limited compared to systems 70 and 80 described later. The external controller 60 may have one or more antennas capable of communicating with the IPG 10. For example, the external controller 60 may have a near-field magnetic induction coil antenna 64a capable of wirelessly communicating with the coil antenna 27a in the IPG 10. The external controller 60 may also have a far-field RF antenna 64b capable of wirelessly communicating with the RF antenna 27b in the IPG 10.

[0018] Clinician programmer 70 is further described in U.S. Patent Application Publication No. 2015 / 0360038 and can include a computing device such as a desktop, laptop or notebook computer, tablet computer, mobile smartphone, personal data assistant (PDA)-type mobile computing device, etc. In FIG4 , the computing device is shown as a laptop computer, which includes typical computer user interface devices (such as a display 71, buttons 72), as well as other user interface devices (such as a mouse, keyboard, speakers, stylus, printer, etc.), not all of which are shown for convenience. FIG4 also shows accessory devices for clinician programmer 70, which are generally specific to its operation as a stimulation controller, such as a communication "wand" 76 that can be coupled to an appropriate port on the computing device. The antenna in clinician programmer 70 used to communicate with IPG 10 can depend on the type of antenna included in IPG 10. If the patient's IPG 10 includes a coil antenna 27a, wand 76 can also include a coil antenna 74a to establish near-field magnetic induction communication over a short distance. In this case, wand 76 can be attached in close proximity to the patient, such as by placing wand 76 in a belt or holster wearable by the patient and near the patient's IPG 10. If IPG 10 includes RF antenna 27b, wand 76, computing device, or both can also include RF antenna 74b to establish communication with IPG 10 at a greater distance. Clinician programmer 70 can also communicate with other devices and networks (such as the Internet) wirelessly or via a wired link provided at an Ethernet or network port.

[0019] External system 80 comprises another device that communicates with and controls IPG 10 via network 85, which may include the Internet. Network 85 may include a server 86 programmed with communication and control functions, and may also include other communication networks or links, such as WiFi, cellular, or landline telephone links. Network 85 ultimately connects to an intermediate device 82 having antennas suitable for communicating with the antennas of the IPG, such as a near-field magnetic induction coil antenna 84a and / or a far-field RF antenna 84b. Intermediate device 82 may be located generally near IPG 10. Network 85 can be accessed by any user terminal 87, which typically includes a computer device associated with a display 88. External system 80 allows a remote user at terminal 87 to communicate with and control IPG 10 via intermediate device 82.

[0020] FIG4 also shows circuitry 90 involved in any external system 60, 70, or 80. Such circuitry may include control circuitry 92, which may include any number of devices capable of executing programs in a computing device, such as one or more microprocessors, microcomputers, FPGAs, DSPs, other digital logic structures, and the like. Such control circuitry 92 may include or be coupled to memory 94, which may store external system software 96 for controlling and communicating with IPG 10 and for presenting a graphical user interface (GUI) 99 on a display (61, 71, 88) associated with the external system. In external system 80, external system software 96 will likely reside in server 86, while control circuitry 92 may exist in either or both of server 86 and terminal 87. Summary of the Invention

[0021] Disclosed herein is a system for providing electrical stimulation to a patient's spinal cord using one or more electrode leads implantable in the patient's spine, each electrode lead including a plurality of spinal electrode contacts, the system including: a neurostimulator connectable to the one or more electrode leads, and a control circuit configured to: cause the neurostimulator to provide a first electrical stimulation to the patient's spinal cord using one or more of the spinal electrode contacts, wherein the first electrical stimulation is configured to induce a first stimulation artifact in the patient's spinal cord but not to induce a detectable neural response; cause the neurostimulator to record a first signal including a first stimulation artifact component using a second or more spinal electrode contacts, fit the first signal to a mathematical model to generate a template signal; cause the neurostimulator to provide a second electrical stimulation to the patient's spinal cord using one or more spinal electrode contacts, wherein the second electrical stimulation is configured to induce a second stimulation artifact and a neural response in the patient's spinal cord; cause the neurostimulator to record a second signal including a second stimulation artifact component and a neural response component using the one or more spinal electrode contacts, and determine a third signal using the second signal and the template signal, wherein the third signal includes a neural response component and includes a stimulation artifact component that is smaller than the second signal. According to some embodiments, the amplitude of the first electrical stimulation is less than the amplitude of the second electrical stimulation. According to some embodiments, the mathematical model includes an exponential decay. According to some embodiments, using the second signal and the template to determine the third signal includes scaling the template signal relative to the second signal. According to some embodiments, using the second signal and the template to determine the third signal includes subtracting the scaled template signal from the second signal to produce the third signal. According to some embodiments, the control circuit is a control circuit of a neurostimulator. According to some embodiments, the control circuit is a control circuit of an external computing device. According to some embodiments, the control circuit is further configured to display a representation of the third signal on a graphical display of the external computing device. According to some embodiments, the control circuit is further configured to determine one or more features in the third signal. According to some embodiments, the control circuit is configured to use the one or more features to perform closed-loop feedback regulation of the therapeutic stimulation. According to some embodiments, the therapeutic stimulation is the second electrical stimulation. According to some embodiments, the closed-loop feedback regulation is configured to keep the therapeutic stimulation within a therapeutic window. According to some embodiments, the mathematical model includes a form of where V(t) is the voltage as a function of time t, V0 is the peak voltage, and τ is the decay time constant. According to some embodiments, the decay time constant τ depends on the resistance R and capacitance C of the tissue near the electrode contact. According to some embodiments, the neural response is a compound evoked action potential (ECAP).

[0022] Also disclosed herein is a method for providing electrical stimulation to a patient's spinal cord using one or more electrode leads implantable in the patient's spine, each electrode lead comprising a plurality of spinal electrode contacts, the method comprising: providing a first electrical stimulation to the patient's spinal cord using the one or more spinal electrode contacts, wherein the first electrical stimulation is configured to induce a first stimulation artifact in the patient's spinal cord but not induce a detectable neural response; recording a first signal comprising a first stimulation artifact component using a second one or more spinal electrode contacts, fitting the first signal to a mathematical model to generate a template signal; providing a second electrical stimulation to the patient's spinal cord using the one or more spinal electrode contacts, wherein the second electrical stimulation is configured to induce a second stimulation artifact and a neural response in the patient's spinal cord; recording a second signal comprising a second stimulation artifact component and a neural response component using the one or more spinal electrode contacts, and determining a third signal using the second signal and the template signal, wherein the third signal comprises a neural response component and comprises a stimulation artifact component that is smaller than the second signal. In some embodiments, the amplitude of the first electrical stimulation is less than the amplitude of the second electrical stimulation. In some embodiments, the mathematical model comprises an exponential decay. According to some embodiments, determining the third signal using the second signal and the template comprises scaling the template signal relative to the second signal. According to some embodiments, determining the third signal using the second signal and the template comprises subtracting the scaled template signal from the second signal to produce the third signal. According to some embodiments, the control circuit is a control circuit of a neurostimulator. According to some embodiments, the control circuit is a control circuit of an external computing device. According to some embodiments, the control circuit is further configured to display a representation of the third signal on a graphical display of the external computing device. According to some embodiments, the control circuit is further configured to determine one or more characteristics of the third signal. According to some embodiments, the control circuit is configured to use the one or more characteristics to perform closed-loop feedback regulation of the therapeutic stimulation. According to some embodiments, the therapeutic stimulation is a second electrical stimulation. According to some embodiments, the closed-loop feedback regulation is configured to keep the therapeutic stimulation within a therapeutic window. According to some embodiments, the mathematical model comprises a form of where V(t) is the voltage as a function of time t, V0 is the peak voltage, and τ is the decay time constant. According to some embodiments, the decay time constant τ depends on the resistance R and capacitance C of the tissue near the electrode contact. According to some embodiments, the neural response is a compound evoked action potential (ECAP).

[0023] The present invention may also exist in the form of a programmable external device for performing the above-mentioned method (via its control circuit), a programmable IPG or ETS for performing the above-mentioned method (via its control circuit), a system including a programmable external device and an IPG or ETS for performing the above-mentioned method, or as a computer-readable medium for performing the above-mentioned method stored in an external device or an IPG or ETS. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG1 shows an implantable pulse generator (IPG) according to the prior art.

[0025] 2A and 2B illustrate examples of stimulation pulses that may be generated by an IPG according to the prior art.

[0026] FIG3 shows a stimulation circuit that may be used in an IPG according to the prior art.

[0027] FIG. 4 illustrates various external devices capable of communicating with an IPG and programming stimuli in the IPG according to the prior art.

[0028] Figure 5 An IPG with neural response sensing capabilities is shown.

[0029] Figure 6 Shown are the stimulation artifact and ECAP alone, and the two signals combined in ESG.

[0030] Figure 7 An algorithm for reducing stimulation artifacts is shown.

[0031] Figure 8 A closed-loop feedback controller is shown. DETAILED DESCRIPTION

[0032] A development of increasing interest in pulse generator systems is the addition of the ability to sense electrical potentials in patient tissue to supplement the stimulation provided by such systems. For example, as described in U.S. Patent Application Publication 2017 / 0296823, it may be beneficial to sense the neural responses produced by neural tissue that has received stimulation from an IPG. U.S. Patent Application Publication 2017 / 0296823 illustrates examples in which sensing of neural responses is useful in an SCS environment and particularly discusses sensing of evoked compound action potentials or "ECAPs." U.S. Patent No. 10,940,316 describes methods and systems that enable stimulation to be adjusted when a patient changes posture based on measurements of neural responses. U.S. Patent Application Publication 2022 / 0184399 describes methods and systems for using recorded neural responses to sense movement of an electrode array during SCS and to adjust stimulation accordingly.

[0033] Figure 5The basic circuitry used to sense neural responses in an IPG 100 is shown. The IPG 100 includes a control circuit 102, which may include a microcontroller, such as part number MSP430 manufactured by Texas Instruments, which is described in a data sheet accessible on the Internet. Other types of control circuits may be used in place of a microcontroller, such as a microprocessor, FPGA, DSP, or a combination of these. As previously mentioned, the control circuit 102 may also be formed in whole or in part in one or more application-specific integrated circuits (ASICs) in the IPG 10. The one or more ASICs may also include Figure 5 Other circuits shown in .

[0034] Figure 5 The stimulation circuit 28 ( FIG. 3 ) described above is included, including one or more DACs (PDAC and NDAC). Bus 118 provides digital control signals to the DACs to generate a current or voltage with a specified amplitude and correct timing at the electrodes selected for stimulation. The electrode current path to the electrode 16 includes the DC blocking capacitor 38 described above.

[0035] Figure 5 Also shown is a circuit for sensing neural responses. As shown, electrode nodes 39 are input to a multiplexer (MUX) 108. MUX 108 is controlled by a bus 114, which operates to select one or more electrode nodes and thus designates corresponding electrodes 16 as sensing electrodes. The one or more sensing electrodes selected via bus 114 can be automatically determined by the control circuit 102 and / or the neural response algorithm 124, as described further below. However, one or more sensing electrodes can be selected by a user (e.g., a clinician) via an external system 60, 70, or 80 ( FIG. 4 ).

[0036] The analog waveform including the sensed neural response and output by the sense amplifier circuit 110 is preferably converted to a digital signal by an analog-to-digital converter (ADC) 112 and input to the control circuit 102 of the IPG. The ADC 112 can be included in the input stage of the control circuit 102. The control circuit 102 can be programmed with a neural response algorithm 124 to evaluate the neural response and take appropriate action accordingly. For example, the neural response algorithm 124 can change the stimulation based on the sensed neural response and can send new control signals via the bus 118 to change the operation of the stimulation circuit 28 to provide better treatment for the patient. As explained in more detail below, one or more algorithms can be configured to extract (calculate) characteristic values ​​of the neural response (such as peak height, curve area, etc.) and use these calculated values ​​as indicators of therapeutic efficacy, for closed-loop adjustment of stimulation, etc.

[0037] The neural response to stimulation is typically a small amplitude AC signal on the order of microvolts or millivolts, which can make sensing difficult. The sense amplifier circuit 110 needs to be able to resolve this small signal, which is particularly difficult when one realizes that this small signal is often dependent on background voltages that are also present in the tissue. The background voltage may be caused by the stimulation itself. In particular, stimulation can produce a "stimulation artifact," which is caused by the electromagnetic field generated by the patient's tissue. The stimulation artifact waveform can be several orders of magnitude larger than the ECAP and typically decays with a time constant of hundreds of microseconds, which is long enough to overlap with the ECAP response. Both the stimulation artifact and the ECAP propagate rostrally and caudally from the location of the stimulating electrode. The propagation speeds of the two signals are typically different; typically, the stimulation artifact propagates faster. Therefore, the stimulation artifact and ECAP signals may overlap to varying degrees, depending on the electrodes used to record the signals. The overlap of the stimulation artifact with the ECAP makes it difficult to calculate the values ​​of various features of the ECAP, as described below.

[0038] Figure 6 The stimulation artifact and ECAP recorded separately are shown, along with an electrospinogram (ESG) containing each of the recorded signals. The ECAP comprises the cumulative response provided by the nerve fibers activated by the stimulus, and essentially comprises the sum of the action potentials of the activated neurons (ganglia or fibers) as they "fire". The ECAP comprises a number of peaks, which are usually labeled P for positive peaks and N for negative peaks, with P1 comprising the first positive peak, N1 comprising the first negative peak, P2 comprising the second positive peak, N2 comprising the second negative peak, and so on. Please note that not all ECAPs will have the same peaks as Figure 6 The exact shape and number of peaks shown are not accurate, as the shape of the ECAP is a function of the number and type of neurons activated and involved in its conduction. Also note that the individual ECAP and artifact signals are not drawn to scale. As mentioned above, artifact signals are typically several orders of magnitude larger than the ECAP signal. Therefore, if the ECAP signal overlaps with the artifact signal, as shown in the combined ESG, it may be difficult to discern the ECAP signal.

[0039] Various forms of artifact reduction techniques are described in the literature. Two common techniques are forward masking and alternating polarity. Both techniques are well described in the art. For example, see Akhoun et al., Electrically evoked compound action potential artifact rejection by independent component analysis: Technique validation, Hear. Res. 302:60-73, (2013).

[0040] In short, forward masking involves delivering a masking pulse that sets the neurons into a refractory state. A probe pulse is then delivered, which allows the resulting artifact (probe artifact) to be measured in the absence of any neural response. In subsequent measurements, the neural signal can be determined by subtracting the identified probe artifact from the total signal, ideally leaving only the neural response (i.e., ECAP).

[0041] Alternating polarity requires recording two buffers and summing them together: one buffer generated by the cathodal-first pulse and the other by the anodal-first pulse. It is assumed that the artifacts generated by the two pulses cancel each other and that the neural responses add together, producing an ECAP with an amplitude twice that of the summed signal.

[0042] Both the forward masking method and the alternating polarity method rely on assumptions that are known to be only approximately true. For example, the forward masking method assumes that all neurons are in a refractory state when the probe stimulus is delivered. However, when the probe stimulus is delivered, neurons that are not in a refractory state will cause a probe "artifact" signal that includes some contribution of the neural response, which is then subtracted from the resulting ECAP measurement, resulting in an inaccurate ECAP measurement. Similarly, in the alternating polarity method, the cathodal first pulse and the anodic first pulse may not produce the same neural activity; the ECAPs can have different delays and amplitudes, resulting in ECAP distortion when the two ECAPs are added. Similarly, the assumption that the stimulation artifacts of the two polarities are equal and opposite may not apply in all cases. U.S. Patent No. 11,241,580, issued on February 8, 2022, discloses a template subtraction method for artifact reduction that overcomes some of the problems associated with the above-mentioned techniques, the contents of which are incorporated herein by reference.

[0043] The inventors have identified new template-based methods for reducing or eliminating stimulation artifact interference in neural responses recorded using spinal electrodes, as described above. Specifically, the techniques and algorithms described herein are useful when neural artifacts overlap with the decaying residual charge portion 602 of the stimulation artifact, such as Figure 6 shown.

[0044] Figure 7 One embodiment of an algorithm 700 for removing stimulation artifact signals from a recorded neural response, such as an ECAP, is shown. Step 702 involves applying stimulation at a stimulation electrode configured to induce an ECAP and recording a signal containing the stimulation artifact and the ECAP signal at a recording electrode. The recorded signal x(t) can be analyzed to determine a time range (R) within which the ECAP appears within the artifact signal. For example, inserting signal 602 ( Figure 6) contains ECAP and stimulation artifact contributions. The range to which ECAP exists in such a signal can be determined. According to some embodiments, particularly when the algorithm 700 is performed in a clinical environment and the resulting recorded signal is displayed on a user interface, a person (e.g., a clinician) can simply observe the signal and try to identify peaks and / or valleys associated with features such as P1, N1, P2 of the ECAP within the signal x(t). According to some embodiments, the clinician can compare the recording channels to confirm the identification of the ECAP features within the x(t) signal, i.e., because on different channels, due to differences in propagation rates, the ECAP features will be separated differently from the artifact signal, as described above. In other embodiments, a peak detection algorithm or the like can be used to determine the range to which ECAP exists in the signal x(t). Figure 7 The range (R) of the signal x(t) containing both artifacts and ECAP is shown.

[0045] Step 704 involves determining a signal y(t) that does not contain any ECAP contribution (within a range R). For example, this may involve applying a stimulus using a stimulation electrode, where the intensity of the stimulus is insufficient to induce an ECAP. In other words, y(t) corresponds to a "subthreshold" stimulus, i.e., a stimulus below the threshold intensity required to induce a detectable ECAP. Figure 7 The artifact signal y(t) within the R range is shown.

[0046] Step 706 involves determining a function to model the subthreshold signal y(t). According to some embodiments, the residual charge decay of the stimulation artifact can be modeled as the voltage decay of an RC circuit according to equation (EQ1):

[0047]

[0048] Where V0 is the peak (or maximum) voltage, and τ is the decay time constant determined by the resistance R and capacitance C of the tissue near the electrode. As described in EQ1, exponential decay is an example of a function that models the residual artifact terms in the subthreshold signal y(t). According to some embodiments, other functions such as polynomial functions, spline functions, etc. can be used. Alternatively, a bandpass / high-pass filter can perform equivalent operation / removal of artifacts. According to some embodiments, a bandpass filter that uses optimization (e.g., least squares method) to match the RC decay can be used.

[0049] If the residual attenuation of the stimulation artifact is modeled according to equation EQ1, the function x(t) including the residual stimulation artifact and the ECAP signal can be expressed by equation EQ2:

[0050]

[0051] Where s(t) is the individual ECAP signal, V 01is the peak voltage of the signal containing ECAP, and n(t) is the noise. Similarly, the subthreshold signal function y(t) can be given by Equation 3:

[0052]

[0053] Where V 02 is the peak voltage of the subthreshold signal (usually less than V 01 ).

[0054] Step 708 involves using the y(t) function to determine a rescaled function y'(t), which can be used as a template to be subtracted from the combined function x(t) to generate the ECAP signal (s(t)). Since the R and C values ​​of the tissue can generally be assumed to be constant, any two residual artifact signals determined at different stimulation currents should differ only in their V0 values. Furthermore, the V0 value is generally linearly related to the stimulation amplitude. Therefore, y'(t) can be expressed according to Equation EQ4:

[0055]

[0056] Step 710 involves determining the signal s(t) representing only the ECAP. Once y'(t) is determined, s(t) can be determined by subtracting y'(t) from the combined function x(t), as shown in equation EQ5:

[0057] s(t)=x(t)-y'(t)EQ5

[0058] The signal s(t), which represents a neural response (e.g., an ECAP) and in which stimulation artifacts are reduced or absent, can be used in any of the applications described above. According to some embodiments, the algorithm 700 can be executed in part or in whole on an external computing device, such as a clinician programmer 70 ( FIG. 4 ). Such embodiments can be particularly suitable for use in a clinical setting. For example, a clinician may wish to record and visualize neural responses (such as ECAP). In such an embodiment, the algorithm can be configured within the external computing device and can perform manual or automatic neural response detection. The algorithm can guide the user through the steps of collecting data with and without neural responses (i.e., the y(t) and x(t) measurements described above). The algorithm can then perform the calculations described above and display the signal s(t), which represents a neural response (e.g., an ECAP) and in which stimulation artifacts are reduced or absent, on a screen of the external computing device.

[0059] According to some embodiments, the disclosed algorithms (such as algorithm 700) may be implemented in the control circuitry of an IPG, for example, as part of the neural response algorithm 124 ( Figure 5). Such embodiments are particularly suitable for allowing the IPG to use features determined from recorded neural responses to adjust the stimulation parameters of the IPG. In such embodiments, the algorithm may periodically cause the IPG to provide subthreshold and suprathreshold stimulation and collect signals y(t) and x(t) accordingly. Thus, the template signal y'(t) may be periodically calculated and stored in the IPG. In order to perform closed-loop control, the algorithm may subtract the stored template signal y'(t) from the signal recorded in response to the therapeutic stimulation provided by the IPG, thereby producing an s(t) signal that includes a neural response with reduced or absent stimulation artifacts. Alternatively, as described above, the algorithm may use an FIR filter (e.g., bandpass or highpass) that performs an approximation of template removal on the artifact remainder, thereby producing s(t). The neural response algorithm may then extract one or more features of the reduced artifact signal s(t). For example, the algorithm may determine one or more neural response features (e.g., ECAP features) that may include, but are not limited to:

[0060] The height of any peaks (e.g., N1);

[0061] The peak-to-peak height between any two peaks (e.g., from N1 to P2);

[0062] Peak height ratio (e.g., N1 / P2);

[0063] The peak width of any peak (e.g., the full width at half maximum of N1);

[0064] The area or energy under any peak;

[0065] ● Total area or energy, including the area or energy under the positive peak minus or plus the area or energy under the negative peak;

[0066] The length of any portion of the curve of the ECAP (e.g., the length of the curve from P1 to N2);

[0067] any time that defines the duration of at least part of an ECAP (e.g., the time from P1 to N2); the time delay from stimulation to the emission of the ECAP, which is indicative of the nerve conduction velocity of the ECAP, which may differ in different types of neural tissue;

[0068] The conduction velocity of the ECAP (i.e., conduction velocity), which can be determined by sensing the movement of the ECAP past different sensing electrodes;

[0069] The rate of change of any previous features, i.e. how those features change over time;

[0070] The power (or energy) determined in a specified frequency band (e.g., delta, alpha, beta, gamma, etc.) determined in a specified time window (e.g., a time window overlapping with neural responses, stimulation artifacts, etc.); Any mathematical combination or function of these variables;

[0071] Such ECAP characteristics can be approximated by a feature extraction algorithm. For example, the area under the curve can include the sum of the absolute values ​​of the digital samples sensed within a specified time interval. Similarly, the curve length can include the sum of the absolute values ​​of the differences between consecutive digital samples sensed within a specified time interval. ECAP characteristics can also be determined within specific time intervals, which can be referenced to the onset of the stimulus or referenced from the ECAP signal itself (e.g., reference peak N1).

[0072] Once the feature extraction algorithm 140 determines one or more of these features, it can be used for any useful effect in the IPG 100, and in particular can be used to adjust the stimulation provided by the IPG 100, for example by providing new data to the stimulation circuit 28 via the bus 118. This is further explained in some of the U.S. patent documents cited above. For example, if the distance between one or more stimulation electrodes and the patient's spinal cord changes (e.g., due to posture changes, coughing, movement, etc.), the stimulation can be adjusted based on the extracted features to maintain optimal therapeutic stimulation.

[0073] The IPG may include a closed-loop feedback control algorithm configured to use one or more neural response eigenvalues ​​as control variables. Closed-loop feedback control is well known in the art and will not be discussed in detail here, but the control scheme may involve a PID controller, a Kalman filter, or other controller. Figure 8 A simplified control diagram 802 is shown in which a controller (e.g., an IPG control circuit) controls stimulation based on one or more ECAP characteristics, which are determined based on recorded responses that have been processed to eliminate or reduce stimulation artifacts, as described above. For example, a feedback control algorithm can adjust stimulation parameters to seek to maintain the sensed ECAP characteristics relative to a set point, threshold, range, etc. According to some embodiments, the feedback control algorithm is configured to maintain stimulation within a therapeutic window.

[0074] It should be understood that algorithm 700 is an example of a template subtraction algorithm. Other examples of template subtraction algorithms exist in the art (some of which are described above), but they generally involve creating a template from the complete artifact, rather than just the residual as described here. Such algorithms generally work by storing the entire signal template in memory and then scaling it based on the usage, rather than storing only a few parameters to process a partial artifact as described here.

[0075] While specific embodiments of the present invention have been shown and described, the foregoing discussion is not intended to limit the present invention to these embodiments. It will be apparent to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the present invention. Therefore, the present invention is intended to encompass alternatives, modifications, and equivalents that may fall within the spirit and scope of the present invention as defined by the claims.

Claims

1. A system for providing electrical stimulation to a patient's spinal cord using one or more electrode leads implantable in the patient's spine, each electrode lead comprising a plurality of spinal electrode contacts, the system comprising: a neurostimulator connectable to the one or more electrode leads, and The control circuit is configured as follows: causing the neurostimulator to provide a first electrical stimulation to the patient's spinal cord using one or more of the spinal cord electrode contacts, wherein the first electrical stimulation is configured to induce a first stimulation artifact in the patient's spinal cord without inducing a detectable neural response, causing the neurostimulator to record a first signal including a first stimulation artifact component using a second one or more of the spinal cord electrode contacts, fitting the first signal to a mathematical model to generate a template signal, causing the neurostimulator to provide a second electrical stimulation to the patient's spinal cord using one or more of the spinal cord electrode contacts, wherein the second electrode stimulation is configured to induce a second stimulation artifact and a neural response in the patient's spinal cord, causing the neurostimulator to record a second signal comprising a second stimulation artifact component and a neural response component using one or more of the spinal cord electrode contacts, and A third signal is determined using the second signal and the template signal, wherein the third signal includes the neural response component and includes a smaller stimulation artifact component than the second signal.

2. The system of claim 1, wherein the amplitude of the first electrical stimulation is smaller than the amplitude of the second electrical stimulation.

3. The system of claim 1 or 2, wherein the mathematical model comprises an exponential decay. 4 . The system of claim 1 , wherein determining a third signal using the second signal and the template comprises scaling the template signal relative to the second signal. 5 . The system of claim 4 , wherein determining a third signal using the second signal and the template comprises subtracting a scaled template signal from the second signal to produce the third signal.

6. The system of any one of claims 1-5, wherein the control circuit is a control circuit of the neurostimulator.

7. The system of any one of claims 1-6, wherein the control circuit is a control circuit of an external computing device.

8. The system of claim 7, wherein the control circuit is further configured to display a representation of the third signal on a graphics display of the external computing device.

9. The system of any one of claims 1-8, wherein the control circuit is further configured to determine one or more characteristics in the third signal.

10. The system of claim 9, wherein the control circuit is configured to use the one or more characteristics to perform closed-loop feedback regulation of therapeutic stimulation. The system of claim 10 , wherein the therapeutic stimulus is the second electrical stimulus.

12. The system of claim 10, wherein the closed-loop feedback regulation is configured to maintain the therapeutic stimulus within a therapeutic window.

13. The system according to any one of claims 1 to 12, wherein the mathematical model comprises a where V(t) is the voltage as a function of time t, V0 is the peak voltage, and τ is the decay time constant.

14. The system of claim 13, wherein the decay time constant τ depends on the resistance R and capacitance C of tissue near the electrode contact.

15. The system of any one of claims 1-14, wherein the neural response is a compound evoked action potential (ECAP).

Images