Controlling neural stimulation using evoked compound action potential signals

Abstract

A neurostimulation device is disclosed, comprising: at least a lead comprising a plurality of electrodes for delivering stimulation pulses to nerves of a patient; a driver circuit configured to generate stimulation pulses and apply them to the electrodes, wherein the nerves have a stimulation threshold voltage which needs to be exceeded to stimulate the nerves; a sensing circuit connected to the electrodes and configured to sense a signal that occurs responsive to a stimulation pulse, the sensed signal having an evoked compound action potential, ECAP, component, which is generated by the nerves responsive to a supra- threshold stimulation pulse, and an artifact component, which is caused by charging of the electrodes by a stimulation pulse regardless of whether the stimulation pulse is above or below threshold; a controller having a calibration mode in which the controller controls the driver circuit to apply sub-threshold stimulation pulses such that no ECAP component is generated so the sensed signal is dominated by the artifact component and a neurostimulating mode in which the controller controls the driver circuit to apply supra-threshold stimulation pulses so that the sensed signal includes an ECAP component; and a signal processing circuit operable to subtract, from the sensed signal collected in the neurostimulating mode, a scaled-up version of the sensed signal collected during the calibration mode, thereby to output a processed sensed signal with an enhanced ECAP component. The scaled-up version of the calibration mode sensed signal is at least in part obtained by amplifying the calibration mode sensed signal by a factor equal to the ratio of supra-threshold stimulation pulse current applied in the neurostimulating mode divided by the sub-threshold stimulation pulse current applied in the calibration mode.

Description

[0001] Applicant: BIOTRONIK SE & Co. KG

[0002] Our Reference: 24.020P-WO

[0003] Date: 05.08.2025

[0004] CONTROLLING NEURAL STIMULATION USING EVOKED COMPOUND ACTION POTENTIAL SIGNALS

[0005] The invention relates to a neurostimulation device that uses evoked compound action potential (ECAP) signals as a control parameter in its control of stimulation pulses applied to electrodes for neurostimulation of a patient.

[0006] Neural stimulation has various medical uses. One use is spinal cord stimulation (SCS), which is used for the treatment of chronic pain, such as back pain. SCS therapy uses an implanted pulse generator to apply therapeutic electric pulses to dorsal column fibers via electrodes implanted in the epidural space dorsal of a patient. Another use is peripheral nerve stimulation (PNS) which is another pain treatment. A still further use is in cochlear implants to stimulate the cochlear nerve with synthetic sound input.

[0007] Application of a therapeutic electric pulse above a certain threshold voltage to a nerve trunk causes or ‘evokes’ a compound action potential (CAP), resulting in these signals being referred to as evoked CAPs (ECAPs). The ECAP signal is the summation of many action potentials from the individual axons in the nerve trunk. Measuring ECAP signals is useful for quantifying the effect of the neural stimulation. For example, in SCS the ECAP signal senses neural activation of the dorsal column fibers. In the following, the term “threshold” refers to the described threshold voltage. For instance, “supra-threshold stimulation” or “subthreshold stimulation” refers to electrical stimulation above or below the threshold voltage. Analogously, a “supra-threshold current level” or a “sub-threshold current level” describe current levels which cause a voltage above or below the threshold voltage.

[0008] An ECAP signal is recorded within a short time window after the beginning of the stimulus pulse, of the order of 100 microseconds, to ensure correlation of stimulus and response. After a certain stimulation threshold is attained ECAP signal amplitude increases linearly with the stimulus current. In paresthesia-based SCS therapies, it is known to use the amplitude of ECAP signals as a control variable in a feedback loop to control pulse output delivery. Example publications disclosing this approach are:

[0009] • US 10,500,399

[0010] • Vallejo R, Chakravarthy K, Will A, Trutnau K, Dinsmoor D. A New Direction for Closed-Loop Spinal Cord Stimulation: Combining Contemporary Therapy Paradigms with Evoked Compound Action Potential Sensing. J Pain Res. 2021 Dec 29;14:3909-3918. doi: 10.2147 / JPR.S344568. PMID: 35002310; PMCID: PMC8721159.

[0011] • US 2022 / 0339442 Al

[0012] • WO 2024 / 002822 Al

[0013] Specifically, WO 2024 / 002822 Al discloses an SCS device comprising implanted electrodes that deliver a succession of therapeutic electric pulses, each of a certain amplitude. In response to each delivered pulse, the device records an ECAP signal, wherein the ECAP signal is fed back in a control loop as a control parameter to adjust the amplitude of the therapeutic electrical stimulation pulses delivered by the SCS device.

[0014] It has been reported in the two publications listed below that, as stimulation frequency increases beyond what has been traditionally used in SCS, above about 100 Hz, there is a decrease in ECAP signal amplitude. Despite the decreasing ECAP signal amplitude with increasing stimulus frequency, patients report an increase in the perceived stimulus strength, so use of higher frequencies is desired.

[0015] • Gmel GE, Santos Escapa R, Parker JL, Mugan D, Al-Kaisy A, Palmisani S. The Effect of Spinal Cord Stimulation Frequency on the Neural Response and Perceived Sensation in Patients With Chronic Pain. Front Neurosci. 2021 Jan 21 ; 15 :625835. doi: 10.3389 / fnins.2021.625835 PMID: 33551738; PMCID: PMC7859107.

[0016] • Sagalajev et al “Absence of Paresthesia During High-rate Spinal Cord Stimulation Reveals Importance of Synchrony for Sensations Evoked by Electrical Stimulation”, Neuron 112, 1-17 (15 November 2023) https: / / doi.Org / 10.1016 / i.neuron.2023.10.021 These two studies both measured stimulation frequencies up to 455 Hz.

[0017] The lower ECAP signal amplitudes at higher frequencies exacerbates the known problem that the ECAP signal is a relatively small component of the measured signal,

[0018] 24.020P-WO / 05.08.2025 since it is superposed on a much larger background signal component, referred to in the art as the ‘artifact signal’. The origin of the artifact signal is electrical charge that is present in the electrical system as a result of the stimulation pulses. In the measured signal, the artifact component is typically several orders of magnitude larger than the ECAP component. ECAP signals are typically in the microvolt range compared with artifact signals in the tens or hundreds of microvolt range.

[0019] Careful signal processing is therefore needed to ensure the ECAP signal component is detectable. In particular, owing to the small voltages involved, the sensed signal is generally amplified prior to being digitized by an analog-to-digital converter. However, transient spikes in the artifact signal can lead to gain saturation of the amplifier, so that it is difficult to operate an amplifier reliably at the desired gain levels.

[0020] According to first aspect of the disclosure there is provided:

[0021] A neurostimulation device comprising: at least a lead comprising a plurality of electrodes for delivering stimulation pulses to nerves of a patient; a driver circuit configured to generate stimulation pulses and apply them to the electrodes, wherein the nerves have a stimulation threshold voltage which needs to be exceeded to stimulate the nerves; a sensing circuit connected to the electrodes and configured to sense a signal that occurs responsive to a stimulation pulse, the sensed signal having an evoked compound action potential, ECAP, component, which is generated by the nerves responsive to a suprathreshold stimulation pulse, and an artifact component, which is caused by charging of the electrodes by a stimulation pulse regardless of whether the stimulation pulse is above or below threshold; a controller having a calibration mode in which the controller controls the driver circuit to apply sub-threshold stimulation pulses such that no ECAP component is generated so the sensed signal is dominated by the artifact component and a neurostimulating mode in which the controller controls the driver circuit to apply supra-threshold stimulation pulses so that the sensed signal includes an ECAP component; and a signal processing circuit operable to subtract, from the sensed signal collected in the neurostimulating mode, a scaled-up version of the sensed signal collected during the

[0022] 24.020P-WO / 05.08.2025 calibration mode, thereby to output a processed sensed signal with an enhanced ECAP component.

[0023] The scaled-up version of the calibration mode sensed signal is obtained by amplifying the calibration mode sensed signal by a factor equal to the ratio of supra-threshold stimulation pulse current applied in the neurostimulating mode divided by the sub-threshold stimulation pulse current applied in the calibration mode.

[0024] According to an embodiment, in the neurostimulating mode the controller is configured to control the driver circuit to vary stimulation pulse amplitude responsive to the processed sensed signal with an enhanced ECAP component.

[0025] For example, the proposed neurostimulator further comprises a memory, wherein in the calibration mode the controller is configured to record the artifact signal over a stimulation period by causing the driver circuit to apply a certain time sequence of stimulation pulses at a sub-threshold current to the electrodes over the stimulation period while storing the sensed signal as a function of time in the memory, thereby to obtain and store an artifact signal time signature for a stimulation period. Moreover, in the neurostimulating mode the controller is configured to apply multiple stimulation periods following the same time sequence of stimulation pulses as used to obtain the artifact signal time signature but at a supra-threshold current, and to control the signal processing circuit to perform said subtraction over each stimulation period synchronously with the scaled-up artifact signal time signature.

[0026] In an embodiment, the calibration mode the driver circuit is controlled to apply the time sequence of sub-threshold stimulation pulses repeatedly over multiple stimulation periods while in each stimulation period the sensed signal is sampled at offset time bases so that the scaled-up artifact signal time signature is output with a higher effective sampling rate than the sampling rate of the signal processing circuit.

[0027] According to an aspect of the present invention, the neurostimulation device further comprises an analog-to-digital converter, ADC, and a digital-to-analog converter, DAC, which are arranged in relation to the memory such that the ADC provides for digital sampling and input of the sensed signal for storage in the memory during the calibration

[0028] 24.020P-WO / 05.08.2025 mode and the DAC provides for analog output of the scaled-up artifact signal time signature to the subtractor during the neurostimulating mode.

[0029] In an embodiment, the controller stores a calibration routine for synchronizing the output from the memory of the scaled-up artifact signal time signature with the sensed signal. The calibration routine determines a delay needed to align in time the sensed signal with output from the memory, the delay being that which minimizes the sample difference deviation.

[0030] According to an embodiment of the present invention, the controller is configured to operate in the calibration mode to re-measure the scaled-up artifact signal time signature in response to an external trigger signal. In an embodiment, the external trigger signal is generated by a change of posture of the patient.

[0031] Preferably, according to an aspect, during the neurostimulating mode the signal processing circuit is operable in real-time to perform said subtraction.

[0032] Moreover, according to an embodiment, the electrodes are connected to the sensing circuit in a pseudo tripolar sensing configuration in one or more groups of three adjacent electrodes, comprising a middle electrode and two outer electrodes.

[0033] In an embodiment of the invention, the sensing circuit comprises: a differential amplifier having an inverting input connected to the middle electrode and a non-inverting input; and an impedance compensator having two inputs connected to respective ones of the two outer electrodes and an output connected to the non-inverting input of the differential amplifier.

[0034] The present invention furthermore comprises a method for determining a value for a stimulation dose threshold in neurostimulation, comprising the steps: a) deliver sub-threshold stimulation pulses (207) to nerves of a patient, sense a plurality of first sensed signals Out SubThr that occur responsive to the sub-threshold stimulation pulses, the first sensed signals possibly having an artifact component;

[0035] 24.020P-WO / 05.08.2025 al) compute an average signal from the first sensed signals, denoted as remnant stimulation artifact sub -threshold, RSA SubThr; a2) perform at least one of rectifying, bin-integrating, and squaring of the difference between the first sensed signals and the average signal (Out SubThr - RSA SubThr), wherein the result is denominated as sub-threshold dose D SubThr; b) deliver stimulation pulses (207) to nerves of a patient with a current amplitude times a scaling factor SF compared to step a), sense a plurality of second sensed signals, Out Step, that occur responsive to the stimulation pulses; bl) perform at least one of rectifying, bin-integrating, and squaring of the difference between the second sensed signals and the average signal, wherein the average signal is multiplied by scaling factor SF, and wherein the result is denominated as step dose D Step, (Out Step - RSA SubThr * SF) = D Step; c) if the difference between D Step and D SubThr multiplied by SF (D Step - D SubThr x SF) is smaller than a first threshold, either declare the averaged Out Step as new RSA SubThr and go back to step b); cl) if the difference between D Step and D SubThr multiplied by SF (D Step - D SubThr x SF) is greater than a first threshold, declare D Step as stimulation perception threshold dose DThr. or reduce the scaling factor SF and go back to step b).

[0036] According to an embodiment, the first threshold is 0.1 times D SubThr. Furthermore, according to an embodiment, the scaling factor SF is reduced by multiplication with a factor smaller than 1.

[0037] The following pseudo-code depicts an embodiment of the inventive method to calculate the stimulation perception Dose Threshold DThr:

[0038] 1. Record eight Resonance Therapy cycles of the output signal 207 (after at least 60 s to reach steady-state) for a Therapy that is known to be sub-threshold (e.g. start with the lowest Stimulation Amplitude I Stim). The signal is denominated as Out SubThr. Out SubThr is averaged over those cycles to generate RSA SubThr

[0039] 24.020P-WO / 05.08.2025 (acronym for remnant SA sub-threshold). Difference (Out SubThr - RSA SubThr) is at least one of rectified, bin-integrated, and squared over those cycles. The result signal is denominated as D SubThr (acronym for Dose Sub-threshold);

[0040] 2. Initialize the following variables: Increase lStim = TRUE; SF = 2; decrease SF = 0 / / SF = scaling factor

[0041] 3. I Stim = I Stim * SF and program such Resonance Therapy.

[0042] 4. Record eight Resonance Therapy cycles of the output signal 207 again (after at least 60 s) for the Therapy in 3. The signal is denominated as Out Step. Difference (Out Step - RSA SubThr * SF) is at least one of rectified, bin-integrated, and squared over those same cycles. The result signal is denominated as D Step (acronym for Dose Step).

[0043] 5. If (D Step - D SubThr x SF) < 0.1 x D SubThr / / This is a comparison level

[0044] IF Increase lStim = TRUE {

[0045] RSA SubThr = average (Out Step); / / Take this baseline as new remnant SA sub-threshold Go to step 3.;

[0046] {

[0047] ELSE D Thr = D Step; / / D Thr = stimulation perception Threshold Dose for Resonance Therapy DThr

[0048] ELSE

[0049] { decrease SF = decrease SF + 1 ;

[0050] Increase lStim = FALSE;

[0051] SF = SF * (1 - decrease SF * 0.1);

[0052] Repeat previous step 3) with the reduced SF.

[0053] }

[0054] Any features and or advantages of the aspects as described in here can be combined with one another as understood by the skilled person. Aspects and embodiments of the described neurostimulation device can be combined with the described method for determining a value for a stimulation dose threshold in neurostimulation and vice versa.

[0055] 24.020P-WO / 05.08.2025 This invention will now be further described, by way of example only, with reference to the accompanying drawings.

[0056] Figure 1 is a schematic diagram of an implantable medical device (IMD) with an SCS stimulator in a medical device communication system with standard architecture in which a patient remote controls therapy delivery to a patient fitted with an IMD.

[0057] Figure 2 is a schematic diagram showing an SCS stimulator attached to a lead arrangement of electrodes according to an embodiment of the invention.

[0058] Figure 3 is a schematic diagram showing more details of the stimulation and ECAP sensing arrangement used in the SCS stimulator of Figure 2.

[0059] Figure 4 is a schematic diagram showing more detail of a signal processing circuit used in the SCS stimulator of Figure 2 to extract the ECAP signal.

[0060] Figure 5 is a graph plotting artifact, emulated ECAP input signal, and extracted ECAP signals acquired during bench testing.

[0061] Figure 6 is a graph plotting remnant artifact signals after front-end cancellation as per the signal processing circuit of Figure 4 and rectified and bin-integrated versions of those artifact signals.

[0062] Figure 7 is a graph plotting artifact signals, emulated ECAP input signal, and extracted ECAP signal following removal of remnant artifact signals.

[0063] In the following detailed description, for purposes of explanation and not limitation, specific details are set forth in order to provide a better understanding of the present disclosure. It will be apparent to one skilled in the art that the present disclosure may be practiced in other embodiments that depart from these specific details.

[0064] 24.020P-WO / 05.08.2025 In particular, the embodiments described in the following are described in relation to an SCS system. However, the approach to extracting the ECAP signal component from the large background constituted by the artifact signal component is equally well applicable to other neurostimulation systems such as in PNS and cochlear implants.

[0065] An SCS device is a type of IMD that is used for chronic pain therapy of a patient. An SCS device uses electrical charge sent through electrodes on one or more leads implanted in the epidural space. Control parameters include at least: pulse width, frequency, amplitude and electrode selection. Additionally, duty cycling and burst stimulation can be introduced as further control parameters to modulate therapy and better manage battery consumption through the introduction of alternating periods of carrier and envelope stimulation.

[0066] An IMD typically forms part of a medical device communication system (MDCS) that is configured to upload clinical data from the IMD on a continual basis via a MDCS to a remotely located data repository, which may be a data center. In an SCS device, clinical data includes both patient data relating to physiological monitoring of a patient's health, for example as collected by sensors of the IMD, and device data related to status and operation of the IMD, for example data logging each time an SCS device is used and at what stimulation level in terms of electrical pulse width and pulse frequency as well as control signals issued to the SCS device to optimize delivery of pain relief. These clinical data can then be accessed by health care staff accessing the data center using suitably designed software applications. Treatment plans and, if necessary, interventions can then be decided upon based on analysis of these clinical data. The analysis may be expert analysis by a health care professional or computer-automated analysis with a software program, e.g., running on computing resource at a neuro-service data center, or a combination of both.

[0067] Figure 1 shows a standard architecture of a MDCS 1 for communication between an IMD 10 implanted in a patient and a remotely located data center 60, which is accessible to health care staff using suitable application software. The IMD 10 includes therapeutic components for patient treatment, such as an SCS device 11 (or another IMD stimulator) incorporating an implantable pulse generator (IPG), and therapeutic components for patient monitoring, such as an IMD sensor 13. The IMD 10 further comprises chipsets providing computing and

[0068] 24.020P-WO / 05.08.2025 telecommunications resource in the form of memory 18, a processor 19 and a wireless personal area network (WPAN) transceiver 12. The IMD 10 is powered by an electrical energy source 14, in this example by a rechargeable battery 14. The IMD's rechargeable battery 14 of an implanted IMD can be charged in a contactless manner by a charger 23, for example a resonant inductive charger, which is placed on the patient's skin adjacent the implanted IMD 10 to charge its rechargeable battery 14. The charger 23 is itself provided with an energy source 24, here a rechargeable battery 24, as well as an external mains power connection 29, for example an external power jack, to power the charger 23 so that its rechargeable battery 24 can be recharged. Alternatively, the charger's battery 24 can be recharged wirelessly by placing the charger 23 on a mains-powered charging pad. The IMD WPAN transceiver 12 uses a suitable WPAN protocol such as Bluetooth Low Energy (BLE), Medical Implant Communication System (MICS) or Medical Device Radiocommunications Service (MedRadio), the latter two being almost identical protocols.

[0069] The patient is provided with a patient remote controller 30 (hereinafter also called “patient remote 30”), for example a smartphone with wireless transceivers 32, 35, 36 respectively for WPAN (Wireless Personal Area Network, e.g. BLE), LPWAN (Low Power Wide Area Networks), cellular (e.g., 4G / LTE / 5G) and WLAN (Wireless Local Area Network) communication. If a smartphone is used, this is a smartphone that is possessed, e.g., owned, by the patient in which the IMD 10 is implanted and on which a software application ('app') is installed. The app is then a so-called Software as a Medical Device (SaMD) which is defined by the United States Food and Drug Administration (FDA) as software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device.

[0070] The cellular transceiver 35 and WLAN transceiver 36 provide two different data communication paths for the patient remote 30 to upload clinical data from the IMD 10 via an internet connection 50 to a remotely located data center 60 (hereinafter also called “backend 60”) acting as repository for storage of clinical data and / or as a host for the services, for example a neuro-service data center. The remote's WLAN transceiver 36 can upload data to the backend 60 via a router 38 and a telephone line 40 (or telephone network 40) using a wired internet connection 50. The internet connection 50 may provide access to

[0071] 24.020P-WO / 05.08.2025 one or more distributed networks and / or cloud services. The remote's cellular transceiver 35 can upload data to the backend 60 via one or more cellular network base stations 42 (cellular towers), for example LPWAN-capable cellular network base station. In some cases, instead of a public communication network, a dedicated point-to-point transmission, e.g., via a dedicated telephone line, may be provided for uploading clinical data. In all these scenarios, uploading of clinical data from the IMD 10 to the backend 60 takes place via the intermediary of the patient remote 30, the latter thereby acting as a relay device.

[0072] Health care staff, such as health care professionals (HCPs), clinical specialists and representatives and remote care team members have access to the backend 60 via suitable portals 70 with the aid of a software application running on the backend 60 and / or the portal 70 to provide the necessary user interfacing, diagnostics and so forth. At least one portal 70 is provided by at least one workstation for health care staff to access a data center, e.g., the backend 60. Analysis and diagnostic software may also be run at the backend 60 to analyze clinical data from individual patients or groups of patients.

[0073] Figure 2 is a schematic diagram showing the SCS device 11 of the IMD 10 attached to a lead arrangement 80 comprising multiple leads 1001 to 100N with each lead incorporating multiple electrodes 101. a to lOl.h. In the illustrated example, there are multiple leads but in other examples there may be only one lead. The SCS device 11 comprises an electrode driver circuit 16 for providing a suitable drive current to each of the electrodes 101. a to lOl.h according to a set of control parameters that follow a treatment plan as controlled by a controller 15 which may be implanted wholly in hardware as a control circuit but may include a combination of hardware, firmware (e.g. programmable logic array(s)) and / or software elements. The controller 15 delivers control signals to the electrode driver circuit 16 according to a treatment plan devised by a computer program running on the processor 19 (e.g., a microprocessor) of the IMD 10. The computer program is stored in the IMD memory 18. These hardware and software components operate collectively to provide an intelligent pulse generator to deliver electrical pulses to the electrodes 101. a to lOl.h conforming to a particular set of parameters, including amplitude, pulse width, frequency and / or duty cycle to provide stimulation therapy. For SCS, the leads 1001 to 100N are implanted at or near a patient’s spinal cord to direct electrical charge into the patient’s tissue

[0074] 24.020P-WO / 05.08.2025 for spinal cord stimulation. The SCS device is implanted subcutaneously. To incorporate ECAP signal feedback to the controller 15, an IMD electrode sensing circuit 17 is provided to collect electrical signals generated responsive to application of stimulation pulses from the electrodes 101. a to lOl.h.

[0075] Figure 3 is a schematic diagram of a part of an example neurostimulation device. A percutaneous or paddle lead 100nis shown with its electrodes 101. a to 101. h. The electrodes 101. f, 101. g and 101. h are connected by a pseudo tripolar sensing configuration 103 to a differential amplifier 105. The electrodes 101. a, lOl.b and lOl.c are used to deliver the drive pulses from the driver circuit 16 for multiphase rotating electrode therapy, in which the three electrodes 101. a, 101. b and 101. c in this example are actuated with successive cathodic pulses (“rotating cathodes”) as described in W02024002822A1. That is, the controller 15 is configured to control the driver circuit 16 to deliver multiphase rotating electrode therapy current to the patient via electrodes 101. a, 101. c and 101.e. The “resonance” therapy operates in an antidromic 102 or local field potential fashion, so that signal measurement of the ECAP signal is performed using electrodes lOl.f, 10 Eg and lOl.h. WO 2024 / 002822 Al discloses for further details of the functionality of resonance stimulation therapy and is hereby referenced. The outer electrodes of the group of three, i.e., electrodes 101. f and lOl.h, are connected to an impedance compensator 104 (e.g. a potentiometer) whose output (e.g. potentiometer tap) is connected to the non-inverting input of the differential amplifier 105. The middle electrode of the group of three, i.e., electrode 101.g, is connected to the inverting input of the differential amplifier 105. Impedance compensator 104 corrects for impedance imbalance in that its output resembles the voltage swing of electrode 10 Eg. The compensation can be relatively coarse, since all that is required is for the output signal from the impedance compensator 104 to remain within the specified range for linear behavior of the differential amplifier 105. Block 108 further filters and converts the output of the differential amplifier 105 to a single-ended output, referred to in the following as the pre- processed sensed signal 106. The pre-processed sensed signal 106 has the relatively small ECAP component superposed on the relatively large artifact component.

[0076] Figure 4 is a schematic diagram of a signal processing circuit 20 used to process the sensed signal to enhance its ECAP signal component. The signal processing circuit 20 has the

[0077] 24.020P-WO / 05.08.2025 function of filtering the sensed signal as received from the sensing circuit 17 to remove as much as possible of the artificial signal component, and to subtract a scaled sampled artifact, ideally only to leave the ECAP signal component.

[0078] The pre-processed sensed signal 106 is initially supplied to one input of a subtractor 209. The output of the subtractor 209 is amplified by a back-end amplifier 205 and then output from the signal processing circuit 20 as a processed output signal 207 after passing through a bandpass filter 206 (e.g. for a chain total passband of 300 Hz - 5000 Hz).

[0079] The signal processing circuit 20 is used in two different modes of operation, referred to as the calibration mode and the neurostimulating mode.

[0080] In the calibration mode, the controller 15 controls the driver circuit 16 to apply a certain time sequence of stimulation pulses corresponding to the intended treatment plan but at subthreshold current levels so that nerves are not stimulated and, consequently, the sensed signal 106 picked up by the sensing circuit 17 has no ECAP component and hence consists essentially only of the artifact component. The artifact component is caused by charging of the electrodes by a stimulation pulse, regardless of whether the stimulation pulse is above or below threshold. The stimulation pulse time sequence has a time span of a time period, referred to as a stimulation period. The signal processing circuit 20 records the sensed signal collected over the stimulation period, which is taken as a proxy for the sub-threshold artifact signal. To do this, a portion of the pre-processed sensed signal 106 is tapped off and recorded in a local memory 201 after digitization by an ADC 200. The local memory 201 thus has stored in it an artifact signal time signature for a stimulation period. It will be understood that the artifact signal time signature will typically be built up over many sub-threshold stimulation periods with the average being taken to ensure the artifact signal time signature most faithfully represents the artifact signal as a function of time over a stimulation period. This can be done by iteratively updating the stored artifact signal time signature by subtracting the sub-threshold sensed signal 106 from the corresponding samples of the presently stored artifact signal time signature until the digitized samples reach a specified deviation and maximum difference. The maximum difference is specified as the largest

[0081] 24.020P-WO / 05.08.2025 difference from the center voltage measured by the reconstruction filter 204. The deviation is specified as mean squared value of the difference of each sample from the center voltage.

[0082] It will also be understood that owing to the digital nature of the memory the artifact signal time signature will consist of a finite number of samples over the stimulation period separated by a certain sampling interval as defined by the specification and operation of the ADC 200. To reduce the effective sampling interval, the stimulation pulse time sequence can be applied multiple times over multiple stimulation periods during the calibration mode, while in each stimulation period the sensed signal 106 is sampled at offset time bases. In this way, the artifact signal time signature can be built up by combining the samples collected at different offset times to arrive at an artificial signal time signature having a higher effective sampling rate than the sampling rate of the ADC 200. For example, if data is collected at four stimulation periods each offset by a quarter of the sampling interval, then the effective sampling rate is multiplied by four. This improves signal subtraction performance during the neurostimulating mode by providing an artifact signal time signature that more closely follows rapid transients in the artificial signal. For example, if the sampling rate is 100 ksps (kilo samples per second) and four stimulation periods are run through with a 2.5 ps delay between acquisitions cycles (e.g. at 0 ps, 2.5 ps, 5.0 ps, and 7.5 ps), then the effective sampling rate of the stored artificial signal time signature is 400 ksps.

[0083] In the neurostimulating mode, the same stimulation pulse time sequence is applied, in practice repeatedly over multiple stimulation periods, but this time at supra-threshold current levels to cause neurostimulation. The sensed signal 106 thus includes both an ECAP component and an artifact component. The signal processing circuit 20 feeds the recorded time sequence of the artifact signal collected during the calibration mode and stored in the local memory 201 via a DAC 203 and reconstruction filter 204 to the other input of the subtractor 209, so that artifact component recorded at the same time in the stimulation period during the calibration mode is subtracted from the sensed signal obtained during the neurostimulating mode. Since the artifact signal time signature was obtained sub-threshold and the sensed signal obtained during the neurostimulating mode is supra-threshold, the artifact signal time signature needs to be scaled up by the ratio of the respective drive currents. For example, if the sub-threshold current was 0.7 times the threshold current for

[0084] 24.020P-WO / 05.08.2025 neurostimulation and the supra-threshold current is 1.4 times the threshold current then a multiplier of two needs to be applied. The scaling-up function can be applied by supplying a scale factor, SF, to the DAC 203 as schematically illustrated, so that individual samples are multiplied up in real time as they are supplied to the DAC 203 during the neurostimulating mode. The SF can be written to the local memory 201 by the controller 15 once the current level of the treatment plan is known as a set-up step when entering the neurostimulating mode. Another possibility for performing the scaling-up function is for a processor to access the memory 201 and store a scaled-up version of the whole artifact signal time signature back into the local memory 201 before entering the neurostimulating mode. A further possibility is for the scaling-up function may be subsumed in the reconstruction filter 204 and carried out in real time during the neurostimulating mode on a sample-by- sample basis as samples are fed through the reconstruction filter 204 to the subtractor 209.

[0085] In the neurostimulating mode, in order to achieve proper cancellation of the artifact signal, there is a need to align the phases in the stimulation period between the two signals arriving at the subtractor 209, so that the artifact signal time signature being applied to the subtractor 209 is at the same phase of the stimulation period as the sensed signal 106. This phase alignment can be achieved by applying an appropriate amount of delay to the samples being provided by the DAC 203 to correctly align in time the single-ended output of the differential amplifier 105 provided by the converter 108 with the output of the reconstruction filter 204. In a preferred embodiment, this is automatically achieved by running a phase alignment routine that tests the maximum sample difference and deviation, and determines the delay that results in the smallest difference deviation. Since this delay is hardware dependent, firmware only needs to run the phase alignment routine once, and then save the values in the local memory 201 for use during the neurostimulating mode.

[0086] Importantly rapid transients (e.g. spikes) in the artifact component are removed, or at least heavily suppressed to leave only a remnant of the artifact component, making the signal after subtraction amenable to subsequent amplification without saturation. The signal processing circuit 20 includes a back-end amplifier 205 for this purpose and also a back-end band-pass filter 206 (e.g. for a chain total passband of 300 Hz - 5000 Hz). The signal processing circuit output 207 is then applied to back-end processing (as described further below).

[0087] 24.020P-WO / 05.08.2025 The embodiment described above does not require blanking of the front-end differential amplifier 105 at any time. Moreover, removing the artifact component does not employ any pre-defined artifact signal template for cancellation but rather drives the neurostimulation device with a particular time sequence of stimulation pulses at sub-threshold current to generate a bespoke artifact signal time signature covering a stimulation period, this bespoke artifact signal time signature then being used subsequently to process the sensed signal obtained by applying the same stimulation pulse time sequence with supra-threshold current.

[0088] If the supra-threshold therapy is changed, or a body posture change of the patient is sensed, or acquisition of a new artifact signal time signature is required for any other reason, then the recalibration can be carried out using the same procedure of applying a sub-threshold version of a new stimulation pulse time sequence intended to be applied supra-threshold according to the new treatment plan. For example, recalibration can be carried out based on any of the following events: after a certain time period has elapsed since the last calibration (e.g. on a timer basis); whenever an accelerometer in the IPG detects a body posture change of the patient; and / or whenever the treatment plan is to be changed to follow a different stimulation pulse time sequence.

[0089] Figure 5 is a graph showing examples of signals acquired during bench testing utilizing an RC network load representative of the tissue-electrode interface in SCS and a supra- threshold 3-electrode resonance therapy at a stimulation current of 1 mA, pulse width of of 300 ps, and a stimulation frequency of 300 Hz).

[0090] The upper two traces show the two inputs to the subtractor 209. Signal 300 is the recorded subtractive input of the subtractor 209, i.e., the samples of the sub-threshold artifact signal time signature multiplied by SF. Signal 301 is the recorded summative input of the subtractor 209, i.e., the pre-processed sensed signal 106 obtained during the neurostimulating mode during supra-threshold therapy.

[0091] In the lower two traces, signal 302 shows the emulated signal processing circuit output 207 (continuous curve) and signal 303 is an emulated ECAP signal component that is injected

[0092] 24.020P-WO / 05.08.2025 via audio transformers to the RC network load (discontinuous lines, DAC-generated output). As can be seen the extracted signal 207 (shown by signal 302) closely conforms to the ECAP signal component demonstrating the effectiveness of the artificial signal component removal.

[0093] BACK-END RESIDUAL SA REMOVAL

[0094] Figure 6 shows examples of sub-threshold averaged artifact time signatures 400, 401 (acquired at pulse currents of 1 mA and 1.3 mA respectively) as measured at the signal processing circuit output 207. Signal 402 is the scaled-up version of signal 400 (1 mA pulse current) multiplied by an SF of 1.3 (corresponding to the amplitude ratio in this example). As can be seen, signal 401 and signal 207 / 402 follow each other in the areas of largest deviation from the baseline which are highlighted by the dashed boxes. Further, if the signal 401 and the signal 207 / 402 are rectified and integrated in the sensing window 403, the delta (signal 404 minus signal 405) between the final and initial values for both signals coincide. WO 2024 / 002822 Al discloses rectifying, bin-integrating and squaring the sensed signal output 207 over several consecutive cycles (e.g. eight) to compute the therapy total dose Di-otai in SCS. According to an embodiment, the following method / pseudo code forback-end processing of the processed sensed signal 207 / 302 to calculate the stimulation perception Dose Threshold DTIH- can be used:

[0095] 1. Record eight Resonance Therapy cycles of the output signal 207 (after at least 60 s to reach steady-state) for a Therapy that is known to be sub-threshold (e.g. start with the lowest Stimulation Amplitude I Stim). The signal is denominated as Out SubThr. Out SubThr is averaged over those cycles to generate RSA SubThr (acronym for remnant SA sub-threshold). Difference (Out SubThr - RSA SubThr) is at least one of rectified, bin-integrated, and squared over those cycles. The result signal is denominated as D SubThr (acronym for Dose Sub-threshold);

[0096] 2. Initialize the following variables: Increase lStim = TRUE; SF = 2; decrease SF = 0 / / SF = scaling factor

[0097] 3. I Stim = I Stim * SF and program such Resonance Therapy.

[0098] 4. Record eight Resonance Therapy cycles of the output signal 207 again (after at least 60 s) for the Therapy in 3. The signal is denominated as Out Step. Difference

[0099] 24.020P-WO / 05.08.2025 (Out Step - RSA SubThr * SF) is at least one of rectified, bin-integrated, and squared over those same cycles. The result signal is denominated as D Step (acronym for Dose Step).

[0100] 5. If (D Step - D SubThr x SF) < 0.1 x D SubThr / / This is a comparison level IF Increase lStim = TRUE

[0101] {

[0102] RSA SubThr = average (Out Step); / / Take this baseline as new remnant SA sub-threshold Go to step 3.;

[0103] {

[0104] ELSE D Thr = D Step; / / D Thr = stimulation perception Threshold Dose for Resonance Therapy DThr

[0105] ELSE

[0106] { decrease SF = decrease SF + 1 ;

[0107] Increase lStim = FALSE;

[0108] SF = SF * (1 - decrease SF * 0.1);

[0109] Repeat previous step 3) with the reduced SF. }

[0110] ECAP MORPHOLOGY EXTRACTION

[0111] Figure 7 shows a graph plotting artifact signals 300, 301, emulated ECAP input signal 303, and extracted ECAP signal 702 following removal of remnant artifact signals.

[0112] Referring to the method / pseudo code above, to observe the ECAP morphology (for suprathreshold therapy) eight (8) cycles of (Output 207 - RSA SubThr * SF SuptoThr), where RSA SubThr is the last signal saved in the back-end residual SA removal algorithm described above and SF SuptoThr is the ratio of the desired supra-threshold Stimulation Amplitude Istim to that Istim found in the algorithm that defined perception Dose Threshold DThr, are averaged. An eight-point median filtering may then be applied to such signal and the result fit using a morphological filter to extract the ECAP 702 morphology. The morphological filter is defined by a first slope, a first peak, an opposite slope to the first

[0113] 24.020P-WO / 05.08.2025 slope, an opposite peak to the first peak, a second peak in the same direction as the first peak, and an opposite slope to the first slope. Figure 7 shows an example of such processing (no median filtering applied) and the expected morphology (two peaks in one direction, another peak in the opposite direction, rising and falling times in the 100s of ps). As it can be seen, the large spike at the beginning of the recording frame that was present in Figure 6 was removed by subtracting RSA SubThr * SF SuptoThr.

[0114] TRADITIONAL THERAPY CASE The same signal processing approach to enhance the ECAP signal component can be applied for feedback control with traditional therapy (i.e. classical biphasic stimulation pulse with active balance delivered at tens of Hz). In this case, since active biphasic stimulation is utilized and ECAP recording happens after charge balancing, the scaling factor can remain 1. There is no need for scaling with different stimulation amplitudes Istim.

[0115] 24.020P-WO / 05.08.2025 REFERENCE NUMERALS

[0116] I Medical Device Communication System (MDCS)

[0117] 10 Implantable Medical Device (IMD)

[0118] I I IMD stimulator, e.g., SCS device with attached electrodes

[0119] 12 IMD wireless personal area network (WPAN) transceiver

[0120] 13 IMD sensor

[0121] 14 IMD electrical energy source

[0122] 15 IMD controller

[0123] 16 IMD electrode driver circuit

[0124] 17 IMD electrode sensing circuit

[0125] 18 IMD memory

[0126] 19 IMD processor

[0127] 20 signal processing circuit

[0128] 23 charger for IMD

[0129] 24 charger energy source

[0130] 29 charger external mains power connection

[0131] 30 patient remote controller (patient remote)

[0132] 32 smartphone wireless personal area network (WPAN) transceiver

[0133] 35 smartphone cellular transceiver

[0134] 36 smartphone wireless local area network (WLAN) transceiver

[0135] 38 router

[0136] 40 telephone line / telephone network

[0137] 42 cellular network base station (cellular tower)

[0138] 50 internet connection

[0139] 60 remotely located data center (backend)

[0140] 70 portal

[0141] 80 lead arrangement

[0142] 1001 to 100N leads lOl.a to lOl.h electrodes

[0143] 102 antidromic potential

[0144] 103 pseudo tripolar sensing configuration

[0145] 24.020P-WO / 05.08.2025 impedance compensator differential amplifier pre-processed sensed signal single-ended output converter analog-to-digital converter, ADC memory of signal processing circuit 20 single-ended output converter for differential amplifier 105 analog-to-digital converter, ADC analog signal reconstruction filter amplifier (gain stage) bandpass filter processed sensed signal subtractor , 301 artifact signals emulated signal processing circuit output 207 emulated ECAP signal component , 401 sub-threshold averaged artifact time signatures scaled-up version of signal 400 sensing window , 405 rectified and integrated signals 400, 401 extracted ECAP signal

[0146] 24.020P-WO / 05.08.2025

Claims

Claims1. A neurostimulation device comprising: at least a lead (100N) comprising a plurality of electrodes (101) for delivering stimulation pulses to nerves of a patient; a driver circuit (16) configured to generate stimulation pulses and apply them to the electrodes, wherein the nerves have a stimulation threshold voltage which needs to be exceeded to stimulate the nerves; a sensing circuit (17) connected to the electrodes and configured to sense a signal that occurs responsive to a stimulation pulse, the sensed signal (106) having an evoked compound action potential, ECAP, component, which is generated by the nerves responsive to a supra-threshold stimulation pulse, and an artifact component, which is caused by charging of the electrodes by a stimulation pulse regardless of whether the stimulation pulse is above or below threshold; a controller (15) having a calibration mode in which the controller controls the driver circuit to apply sub-threshold stimulation pulses such that no ECAP component is generated so the sensed signal is dominated by the artifact component (300) and a neurostimulating mode in which the controller controls the driver circuit to apply supra-threshold stimulation pulses so that the sensed signal includes an ECAP component (301); and a signal processing circuit (20) operable to subtract, from the sensed signal collected in the neurostimulating mode, a scaled-up version of the sensed signal collected during the calibration mode, thereby to output a processed sensed signal (207) with an enhanced ECAP component (302), wherein the scaled-up version of the calibration mode sensed signal is at least in part obtained by amplifying the calibration mode sensed signal by a factor equal to the ratio of supra-threshold stimulation pulse current applied in the neurostimulating mode divided by the sub-threshold stimulation pulse current applied in the calibration mode.

2. The neurostimulation device of claim 1, wherein in the neurostimulating mode the controller is configured to control the driver circuit to vary stimulation pulse amplitude responsive to the processed sensed signal with an enhanced ECAP component.24.020P-WO / 05.08.20253. The neurostimulation device of claim 1 or 2, further comprising a memory (201); wherein in the calibration mode the controller is configured to record the artifact signal over a stimulation period by causing the driver circuit to apply a certain time sequence of stimulation pulses at a sub-threshold current to the electrodes over the stimulation period while storing the sensed signal as a function of time in the memory, thereby to obtain and store an artifact signal time signature for a stimulation period, and wherein in the neurostimulating mode the controller is configured to apply multiple stimulation periods following the same time sequence of stimulation pulses as used to obtain the artifact signal time signature but at a supra-threshold current, and to control the signal processing circuit to perform said subtraction over each stimulation period synchronously with the scaled-up artifact signal time signature.

4. The neurostimulation device of claim 3, wherein in the calibration mode the driver circuit is controlled to apply the time sequence of sub-threshold stimulation pulses repeatedly over multiple stimulation periods while in each stimulation period the sensed signal is sampled at offset timebases so that the scaled-up artifact signal time signature is output with a higher effective sampling rate than the sampling rate of the signal processing circuit.

5. The neurostimulation device of claim 3 or 4, further comprising an analog-to-digital converter, ADC, (200) and a digital-to-analog converter, DAC, (203) which are arranged in relation to the memory (201) such that the ADC provides for digital sampling and input of the sensed signal for storage in the memory during the calibration mode and the DAC provides for analog output of the scaled-up artifact signal time signature to the subtractor during the neurostimulating mode.

6. The neurostimulation device of claim 5, wherein the controller stores a calibration routine for synchronizing the output from the memory of the scaled-up artifact signal time signature with the sensed signal, wherein the calibration routine determines a delay needed24.020P-WO / 05.08.2025to align in time the sensed signal with output from the memory, the delay being that which minimizes the sample difference deviation.

7. The neurostimulation device of any one of the preceding claims, wherein the controller is configured to operate in the calibration mode to re-measure the scaled-up artifact signal time signature in response to an external trigger signal.

8. The neurostimulation device of claim 7, wherein the external trigger signal is generated by a change of posture of the patient.

9. The neurostimulation device of any one of the preceding claims, wherein during the neurostimulating mode the signal processing circuit is operable in real-time to perform said subtraction.

10. The neurostimulation device of any one of the preceding claims, wherein the electrodes are connected to the sensing circuit in a pseudo tripolar sensing configuration (103) in one or more groups of three adjacent electrodes, comprising a middle electrode and two outer electrodes.

11. The neurostimulation device of claim 10, wherein the sensing circuit comprises: a differential amplifier (105) having an inverting input connected to the middle electrode and a non-inverting input; and an impedance compensator (104) having two inputs connected to respective ones of the two outer electrodes and an output connected to the non-inverting input of the differential amplifier.

12. Method for determining a value for a stimulation dose threshold in neurostimulation, comprising the steps: a) deliver sub-threshold stimulation pulses (207) to nerves of a patient, sense a plurality of first sensed signals Out SubThr that occur responsive to the sub-threshold stimulation pulses, the first sensed signals possibly having an artifact component;24.020P-WO / 05.08.2025al) compute an average signal from the first sensed signals, denoted as remnant stimulation artifact sub -threshold, RSA SubThr; a2) perform at least one of rectifying, bin-integrating, and squaring of the difference between the first sensed signals and the average signal (Out SubThr - RSA SubThr), wherein the result is denominated as sub-threshold dose D SubThr; b) deliver stimulation pulses (207) to nerves of a patient with a current amplitude times a scaling factor SF compared to step a), sense a plurality of second sensed signals, Out Step, that occur responsive to the stimulation pulses; bl) perform at least one of rectifying, bin-integrating, and squaring of the difference between the second sensed signals and the average signal, wherein the average signal is multiplied by scaling factor SF, and wherein the result is denominated as step dose D Step, (Out Step - RSA SubThr * SF) = D Step; c) if the difference between D Step and D SubThr multiplied by SF (D Step - D SubThr x SF) is smaller than a first threshold, either declare the averaged Out Step as new RSA SubThr and go back to step b); cl) if the difference between D Step and D SubThr multiplied by SF (D Step - D SubThr x SF) is greater than a first threshold, declare D Step as stimulation perception threshold dose DThr. or reduce the scaling factor SF and go back to step b).

13. The method according to claim 12, wherein the first threshold is 0.1 times D SubThr.

14. The method according to claim 12 or 13, wherein the scaling factor SF is reduced by multiplication with a factor smaller than 1.24.020P-WO / 05.08.2025

Images