Minimize neural stimulation artifacts

Abstract

A neural stimulation device has stimulation and measurement electrodes positioned relative to the stimulation, the stimulation and relative positions being configured such that in an artifact generated relative to the distance from the stimulation electrode, the minimum area of ​​the artifact is substantially co-located with the measurement electrode. Alternatively, the ratio of the inter-electrode spacing to the electrode length is between 2 and 3.66. Alternatively, an impedance is connected to a passive electrode and is configured to reduce artifacts generated on the measurement electrode.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of Australian Provisional Patent Application No. 2018904012 filed on 23 October 2018, which is incorporated herein by reference. Technical Field

[0003] The present invention relates to the measurement of compound action potentials evoked by a neural stimulator and, in particular, to minimizing artifacts caused by the application of electrical stimulation. Background of the Invention

[0005] There are many situations where it is desirable to apply neurostimulation in order to generate a compound action potential (CAP). For example, neuromodulation is used to treat a variety of conditions, including chronic pain, Parkinson's disease, and migraine. Neuromodulation systems apply electrical pulses to tissues in order to produce a therapeutic effect. When used to relieve chronic pain, electrical pulses are applied to the dorsal column of the spinal cord (DC), known as spinal cord stimulation (SCS). Neuromodulation systems typically include an implantable electrical pulse generator and a power source, such as a battery that can be recharged by transcutaneous inductive transmission. An electrode array is connected to the pulse generator and positioned in the dorsal epidural space above the dorsal column. The electrical pulses applied to the dorsal column by the electrode cause neuronal depolarization and generate a propagating action potential. The fibers stimulated in this way inhibit the transmission of pain from that section of the spinal cord to the brain. In order to maintain the pain relief effect, stimulation is applied continuously, for example, at a frequency in the range of 50 Hz to 100 Hz.

[0006] Neuromodulation can also be used to stimulate efferent nerve fibers, for example, to induce motor function. Generally speaking, the electrical stimulation generated in a neuromodulation system triggers a nerve action potential, which later has either an inhibitory effect or a stimulatory effect. The inhibitory effect can be used to modulate an undesirable process such as pain transmission, or to induce a desired effect such as muscle contraction.

[0007] There are a range of situations where it is desirable to obtain electrical measurements of compound action potentials (CAPs) induced on a neural pathway by an electrical stimulus applied to the neural pathway. However, this can be a difficult task because the observed CAP signal will typically have a maximum amplitude of tens of microvolts or less, while the stimulus applied to induce the CAP is typically several volts. Electrode artifacts are typically generated by the stimulus and appear as a decaying output of a few millivolts or hundreds of microvolts over the time the entire CAP occurs, which poses a significant obstacle to isolating the much smaller CAP of interest. Because neural responses can occur simultaneously with stimulation and / or stimulation artifacts, CAP measurements present difficult challenges to implant design. In practice, many non-ideal aspects of the circuit lead to artifacts, and because these primarily have decaying exponential characteristics that can have either positive or negative polarity, identification and elimination of the source of the artifacts can be laborious. Various methods have been proposed to record CAP, including those of King (US Pat. No. 5,913,882), Nygard (US Pat. No. 5,785,651), Daly (US Pat. App. No. 2007 / 0225767), and the present applicant (US Pat. No. 9,386,934).

[0008] When evoked responses occur later than the time at which the artifacts occur, or when the signal-to-noise ratio is high enough, they are less difficult to detect. Artifacts are typically confined to a time of 1-2 ms after stimulation and therefore data can be obtained if neural responses are detected after this time window. This is the case in surgical monitoring, where the distance between the stimulating and recording electrodes is large, so that the neural response propagation time from the stimulation site to the recording electrode exceeds 2 ms. However, in order to characterize responses evoked by a single implant (such as responses from the dorsal columns to the SCS, for example), high stimulation currents and close proximity between electrodes are required, and therefore the measurement process must directly overcome simultaneous artifacts, which greatly exacerbates the difficulty of neural measurements.

[0009] Similar considerations may arise in deep brain stimulation, where it may be desirable to stimulate a neural structure and measure the evoked compound action potential generated in that structure immediately before the neural response propagates elsewhere in the brain. Artifacts remain a significant obstacle to measuring neural responses near the stimulation site, with the result that most, if not all, conventional neural stimulation implants (which must be compact devices) are unable to make any measurements of neural responses evoked by the implant's stimulation anyway.

[0010] Any discussion of documents, acts, materials, devices, articles and the like that has been included in this specification is solely for the purpose of providing a context for the present invention. It is not to be taken as an admission that any or all of these matters form part of the prior art base or were common general knowledge in the field relevant to the present invention simply because they existed before the priority date of each claim of this application.

[0011] Throughout this specification, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or any other group of elements, integers or steps.

[0012] In the present specification, a statement that an element may be “at least one of” a list of options will be understood that the element may be any one of the listed options, or may be any combination of two or more of the listed options. Summary of the invention

[0013] According to a first aspect, the present invention provides a nerve stimulation device, the nerve stimulation device comprising:

[0014] at least one stimulation electrode configured to deliver electrical stimulation to neural tissue; and

[0015] at least one measuring electrode configured to record a response of the neural tissue to the stimulus,

[0016] Therein, the positions of the stimulation and measurement electrodes relative to the stimulation electrode are configured such that, in an artifact generated relative to the distance from the stimulation electrode, a minimum area of ​​the artifact is located substantially at the same position as the measurement electrode.

[0017] According to a second aspect, the present invention provides a method for neural stimulation, the method comprising:

[0018] delivering electrical stimulation to neural tissue using at least one stimulation electrode; and

[0019] recording the response of the neural tissue to the stimulus using at least one measuring electrode,

[0020] Therein, the positions of the stimulation and measurement electrodes relative to the stimulation electrode are configured such that, in an artifact generated relative to the distance from the stimulation electrode, a minimum area of ​​the artifact is located substantially at the same position as the measurement electrode.

[0021] The minimum area of ​​the artifact is defined as a spatial area in which the amplitude of the artifact is reduced relative to the peak artifact in the spatial area farther from the stimulation electrode. In some embodiments, the minimum area of ​​the artifact may include a zero-crossing area of ​​the artifact, which is a spatial area in which the amplitude of the artifact is reduced relative to the peak artifact in the spatial area farther from the stimulation electrode, and wherein the zero-crossing area of ​​the artifact contains the zero-crossing of the artifact. For example, in some embodiments, the minimum area of ​​the artifact may include a spatial area in which the amplitude of the artifact is less than 75%, more preferably less than 50%, and more preferably less than 25% of the peak artifact generated in the spatial area farther from the stimulation electrode. It is worth noting that by locating the minimum area of ​​the artifact in the same position as the measurement electrode, it allows the measurement electrode to be positioned closer to the stimulation site and thus allows the capture of stronger and less dispersed evoked compound action potentials (ECAPs) while being subject to fewer artifacts than would occur at some locations farther from the stimulation site. In a preferred embodiment, the stimulation and measurement electrode positions relative to the stimulation electrode are configured so that an artifact minimum (such as an artifact zero crossing) is located substantially or preferably exactly at the same position as the measurement electrode, so that the artifact experienced by the measurement electrode is negligible relative to peak artifacts produced at other distances from the stimulation electrode.

[0022] To this end, in at least some embodiments, the present invention provides techniques relating to the design of stimulation patterns for spinal cord stimulation and other neuromodulation methods, and provides a set of methods that can be used to reduce or eliminate artifacts and thereby improve the measurement of tissue responses to stimulation, such as neural compound action potentials evoked by the applied stimulation.

[0023] In some embodiments of the first aspect of the invention, the stimulation is configured so that the minimum area of ​​the artifact is substantially co-located with the measuring electrodes by delivering the stimulation in a multipolar manner using more than two stimulation electrodes and applying a mismatch on the corresponding stimulation electrode pairs so that the minimum area of ​​the artifact is co-located with the measuring electrodes. For example, the stimulation can be delivered in a tripolar manner by three stimulation electrodes, the three stimulation electrodes comprising a central electrode carrying the entire stimulation current and two peripheral electrodes carrying two corresponding portions of the stimulation current to maintain charge balanced stimulation, whereby the corresponding portions of the stimulation current carried by the peripheral stimulation electrodes are mismatched in such a way that the minimum area of ​​the artifact is substantially co-located with the measuring electrodes.

[0024] The desired ratio between the respective portions of the stimulation current carried by the peripheral electrodes (referred to herein as the stimulation ratio) differs depending on the measurement electrode(s) used. Thus, a preferred embodiment provides an artifact minimization algorithm that delivers a series of stimulations of different stimulation ratios at subthreshold levels that do not recruit neural responses, and observes the artifacts induced by each such stimulation at the measurement electrodes, in order to seek a stimulation ratio that minimizes the artifacts observed at the measurement electrodes used.

[0025] Other embodiments of the first aspect of the present invention may employ other stimulation configurations so that the minimum area of ​​the artifact is substantially co-located with the measuring electrode. For example, a quadrupolar stimulation configuration employing four stimulation electrodes, or a multipolar stimulation configuration employing more than four electrodes may be employed, and may be pre-configured or adaptively configured to deliver (multiple) mismatched stimulation ratios in order to manipulate the position of the minimum area of ​​the artifact so that the minimum area of ​​the artifact is substantially co-located with the measuring electrode.

[0026] According to a third aspect of the present invention, there is provided an implantable lead for neural stimulation, the lead comprising a plurality of electrodes, each electrode having an electrode length and an electrode width, and the electrodes being longitudinally spaced apart by an inter-electrode spacing,

[0027] Wherein, the electrode width is greater than the electrode length;

[0028] Wherein, the length of the electrode is less than 3 mm, and

[0029] Wherein, the ratio of the inter-electrode spacing to the electrode length is between 2 and 3.66.

[0030] In an embodiment of the third aspect, wherein the lead is a percutaneous lead comprising a skin-engraving electrode that passes substantially or completely around the circumference of the lead, or in the case of a non-cylindrical lead around the cross-sectional periphery of the lead, the electrode width is defined herein as being equal to the width or diameter or maximum cross-sectional dimension of the lead.

[0031] In an embodiment of the third aspect, the electrode length is preferably greater than 1.5 mm, such as greater than 1.6 mm, more preferably greater than 1.8 mm. The electrode length is preferably less than 2.9 mm, more preferably less than 2.55 mm, more preferably less than 2.2 mm. The electrode length is preferably 2.0 mm.

[0032] In an embodiment of the third aspect, the ratio of the inter-electrode spacing to the electrode length is preferably greater than 2.1, such as greater than 2.25, more preferably greater than 2.4. The ratio of the inter-electrode spacing to the electrode length is preferably less than 3.5, such as less than 3.1, more preferably less than 2.7. The ratio of the inter-electrode spacing to the electrode length is preferably 2.5.

[0033] According to a fourth aspect, the present invention provides a nerve stimulation device, comprising:

[0034] at least one stimulation electrode configured to deliver electrical stimulation to neural tissue;

[0035] at least one measuring electrode configured to record a response of the neural tissue to the stimulation;

[0036] at least one passive electrode near the measuring electrode; and

[0037] An impedance coupled to the passive electrode is configured to reduce artifacts produced at the measuring electrode.

[0038] According to a fifth aspect, the present invention provides a method for neural stimulation, the method comprising:

[0039] delivering electrical stimulation to neural tissue using at least one stimulation electrode;

[0040] using measurement electrodes to record the response of the neural tissue to the stimulation;

[0041] An impedance connected to a passive electrode located proximate to the measuring electrode is configured to reduce artifacts generated at the measuring electrode.

[0042] In the embodiments of the fourth and fifth aspects of the present invention, the impedance can be preconfigured and can be fixed. In alternative embodiments, the impedance can be a variable impedance and can be adaptively configured by any suitable means in order to reduce the artifacts observed on the (multiple) measuring electrodes or preferably seek the minimum value of the artifacts. For example, the variable impedance can be adaptively configured by using an artifact minimization algorithm, the algorithm delivers a series of stimuli while adjusting the variable impedance to take a series of different values, the stimulation is at a subthreshold level that does not recruit neural responses, and observes the artifacts caused by each such stimulus at the (multiple) measuring electrodes, so as to seek the impedance value that minimizes the artifacts observed on the (multiple) measuring electrodes used at the time. Alternatively, this technique can use suprathreshold stimulation that does recruit neural responses, and can measure the total energy of neural responses + artifacts, and can adjust the variable impedance in a way that seeks to reduce or minimize such energy. The artifact minimization algorithm of such an embodiment can operate simultaneously with the above-mentioned artifact minimization algorithm of some embodiments of the first and second aspects of the present invention, for example by using two feedback loops that run simultaneously or concurrently.

[0043] The impedance may comprise a resistance or a reactance.The impedance may be variable by being implemented in the form of a switched capacitor resistor, configured to present a controllable resistance as electronically defined by a switching rate or pulse width modulation of the switched capacitor.

[0044] In some embodiments, the impedance is connected between a pair of passive electrodes located on either side of the measuring electrode. In such embodiments where more than one measuring electrode is used, one of the passive electrodes is preferably located between the measuring electrodes, and another of the passive electrodes is preferably located between the stimulation electrode(s) and the measuring electrode. Thus, a pair of electrodes not involved in the stimulation or measurement can be used to direct an artifact minimum or zero to the measuring electrode.

[0045] According to a sixth aspect, the present invention provides a nerve stimulation device, the nerve stimulation device comprising:

[0046] at least one stimulation electrode configured to deliver electrical stimulation to neural tissue; and

[0047] at least one measuring electrode configured to record a response of the neural tissue to the stimulation;

[0048] The electrodes are arranged in a longitudinal array, and at least one of the electrodes is configured to exhibit a resistance in the longitudinal direction that is greater than a resistance of the electrode in the transverse direction.

[0049] In some embodiments of the sixth aspect, the electrode can be configured to exhibit greater resistance in the longitudinal direction by providing the electrode with transverse grooves or ribs, wherein the transverse grooves or ribs are configured to increase the longitudinal current path length.

[0050] According to a seventh aspect, the present invention provides a nerve stimulation device, comprising:

[0051] at least one stimulation electrode configured to deliver electrical stimulation to neural tissue;

[0052] a current source configured to generate the electrical stimulation; and

[0053] at least one measuring electrode configured to record a response of the neural tissue to the stimulation;

[0054] wherein the stimulation electrode is segmented into at least two electrode portions, the at least two electrode portions being configured to create a discontinuity in a tissue-electrode interface, and wherein the current source is connected to the electrode portions via respective resistors.

[0055] In some embodiments of the seventh aspect, a first resistor connecting the current source to a first portion of the segmented electrode is different from a second resistor connecting the current source to a second portion of the segmented electrode. The first resistor and the second resistor are preferably selected or adaptively controlled in a manner to offset asymmetric voltages at the electrode-tissue interface. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Examples of the present invention will now be described with reference to the accompanying drawings, in which:

[0057] Figure 1 An implantable spinal cord stimulator is schematically shown;

[0058] Figure 2 is a block diagram of an implantable neurostimulator;

[0059] Figure 3 is a schematic diagram showing the interaction of an implanted stimulator with a nerve;

[0060] Figure 4 An implantable lead electrode array was demonstrated;

[0061] Figure 5 It is the equivalent circuit of a single card skin electrode;

[0062] Figure 6 is a graph of predicted artifact amplitudes in simulations as a function of the number of segments into which the card skin electrode is nominally segmented;

[0063] Figure 7 The stimulation and measurement configurations of the electrode leads are graphically depicted;

[0064] Figure 8 is a plot of stimulation voltage versus the resulting saline artifact;

[0065] Fig. 9 is a simplified circuit diagram showing the simulation of cutting each of the four electrodes into two slices, and the interaction of the slices with the grid resistance representation of the saline;

[0066] Fig.10 is a plot of the simulated potential generated on the interface layer of electrode 4 as a function of the nominal slice position in response to a stimulus applied between electrode 2 and electrode 1 at the end of the stimulus and after 1 ms of relaxation in a 21-slice simulation;

[0067] Fig.11 is a plot of the simulated potentials generated on the stimulating electrodes at the end of biphasic stimulation in a 21-slice simulation.

[0068] Fig.12 are plots of the (measured and simulated) artifacts as a function of stimulation current;

[0069] Fig.13 is a plot of the measured (measured and simulated) artifacts between adjacent electrode pairs as a function of the displacement of the stimulating electrodes;

[0070] Fig.14a and Fig.14b Illustration of the current flow paths in two stimulating electrodes and one passive electrode during and after a stimulation pulse, respectively;

[0071] Fig.15 is a simulation of the electric field in saline surrounding two stimulating electrodes and two passive electrodes shortly after stimulation is completed;

[0072] Fig.16 is a simulation of the electric field in saline surrounding two stimulating electrodes and one passive electrode shortly after stimulation is completed;

[0073] Fig.17 A tripolar stimulation method for artifact minimization is shown;

[0074] Fig.18 demonstrated the decomposition of tripolar stimulation into bipolar stimulation;

[0075] Fig.19 Shows from Fig.18 The imaginary sum of the components;

[0076] Fig. 20 shows the simulation artifact of tripolar stimulation when term A varies;

[0077] Fig.21 A variable tripolar stimulation waveform according to an embodiment of the present invention is shown;

[0078] Fig. 22 shows the artifact amplitudes at electrodes E4 to E8 when the tripolar stimulation ratio is varied;

[0079] Figures 23a to 23d A feedback control loop for automatically adjusting the multipolar stimulation ratio to seek a minimum of artifacts according to an embodiment of the present invention is presented;

[0080] Fig.24 Another embodiment is presented wherein a bridge resistor minimizes artifacts;

[0081] Fig.25 A 4-contact epidural electrode array with conventionally sized electrodes and an epidural electrode array including shortened electrodes according to an embodiment of the present invention are shown;

[0082] Fig.26 A circuit model of a single electrode in a saline bath is shown;

[0083] Fig. 27Shown along Fig.26 The voltage on the electrode surface;

[0084] Fig.28 A circuit model of a pair of stimulating electrodes in a saline bath is shown;

[0085] Fig.29 Shown along Fig.28 The voltage on the electrode surface;

[0086] Fig.30 A circuit model of a pair of recording electrodes in a saline bath is shown;

[0087] Fig.31 Shown along Fig.30 The voltage on the electrode surface;

[0088] Fig.32 Shown for highlighting Fig.34 Compression algorithm for artifacts in ;

[0089] Fig.33 Stimulation waveforms are shown;

[0090] Fig.34 Compressed fields in response to biphasic stimulation in a saline bath are shown;

[0091] Fig.35 The uncompressed voltage along the electrode surface is shown;

[0092] Fig.36 A paddle electrode array with shortened electrodes according to another embodiment of the present invention is shown;

[0093] Fig.37 A neural stimulation lead with segmented electrodes according to an embodiment of the present invention is shown;

[0094] Fig.38 Another neural stimulation lead with split electrodes according to another embodiment of the present invention is shown; and

[0095] Fig.39 A neural stimulation electrode with increased longitudinal resistance according to a further embodiment of the invention is presented. DETAILED DESCRIPTION

[0096] Figure 1An implantable spinal cord stimulator 100 is schematically illustrated. The stimulator 100 includes an electronic device module 110 implanted at a suitable location in the patient's lower abdominal region or upper posterior buttock region, and an electrode assembly 150 implanted in the epidural space and connected to the module 110 by suitable leads. Many aspects of the operation of the implantable neural device 100 can be reconfigured by an external control device 192. In addition, the implantable neural device 100 functions as a data acquisition device, wherein the acquired data is transmitted to an external device 192 via any suitable percutaneous communication channel 190.

[0097] Figure 2 1 is a block diagram of an implantable neurostimulator 100. The module 110 includes a battery 112 and a telemetry module 114. In an embodiment of the present invention, the telemetry module 114 can use any suitable type of transcutaneous communication 190 such as infrared (IR) transmission, electromagnetic transmission, capacitive transmission, and inductive transmission to transmit power and / or data between an external device 192 and the electronic device module 110. The module controller 116 has an associated memory 118 that stores patient settings 120, a control program 122, etc. The controller 116 controls the pulse generator 124 to generate stimulation in the form of current pulses according to the patient settings 120 and the control program 122. The electrode selection module 126 switches the generated pulses to the appropriate electrode(s) of the electrode array 150 for delivering the current pulses to the tissue surrounding the selected electrode(s). The measurement circuit system 128 is configured to capture measurements of the neural responses sensed at the sensing electrode(s) of the electrode array as selected by the electrode selection module 126.

[0098] Figure 3 1 is a schematic diagram illustrating the interaction of an implantable stimulator 100 with a nerve 180, however, in this case, the spinal cord alternative embodiment may be positioned adjacent to any desired neural tissue, including peripheral nerves, visceral nerves, parasympathetic nerves, or brain structures. The electrode selection module 126 selects the stimulation electrode 2 of the electrode array 150 to deliver current pulses to the surrounding tissue including the nerve 180, and also selects the return electrode 4 of the array 150 for stimulation current recovery to maintain zero net charge transfer.

[0099] Appropriate stimulation is delivered to the nerve 180 to induce a neural response, which includes a compound action potential that will propagate along the nerve 180 as shown for therapeutic purposes, which in the case of a spinal cord stimulator for chronic pain can be to produce paresthesia at a desired location. To this end, the stimulating electrodes are used to deliver stimulation at any therapeutically suitable frequency (e.g., 30 Hz, but other frequencies including up to the kHz range can also be used), and / or the stimulation can be delivered in a non-periodic manner (depending on the patient), such as in bursts or sporadic. In order to adapt the device, clinicians apply various configurations of stimulation that attempt to produce the sensation that the user experiences as paresthesia. When a stimulation configuration that induces paresthesia is found, the location and size of the paresthesia is consistent with the area of ​​the user's body affected by pain, and the clinician recommends continued use of the configuration.

[0100] The device 100 is further configured to sense the presence and strength of a compound action potential (CAP) propagating along the nerve 180, whether such a CAP is induced by stimulation from electrodes 2 and 4, or induced in other ways. To this end, the electrode selection module 126 can select any electrode in the array 150 to act as a measurement electrode 6 and a measurement reference electrode 8. The signals sensed by the measurement electrodes 6 and 8 are transmitted to the measurement circuit system 128, which can, for example, operate according to the teachings of the applicant's international patent application publication number WO 2012155183, the contents of which are incorporated herein by reference. However, artifacts remain a significant obstacle to measuring neural responses near the stimulation location. The present disclosure first studies the artifact phenomenon in more detail, and then provides a number of novel solutions based on these findings.

[0101] Current pulses delivered by medical implants through platinum electrodes to recruit neurons produce slowly decaying voltage tails, called "artifacts". These tails make it very difficult to measure the evoked potential following the pulse. We provide evidence that, in typical clinical scenarios, these tails are mainly caused by concentration gradients of species induced in the double layer adsorbed onto the platinum surfaces of the stimulating and passive electrodes. A compact model is proposed that allows simulation of these artifacts. It is expected that the model may prove useful for predicting the effectiveness of techniques to reduce the amplitude of artifacts.

[0102] Recently, there has been a great deal of interest in rapidly sensing neural activity in the presence and after neural stimulation pulses in order to apply feedback in neuromodulation systems. The evoked compound action potential (ECAP), measured synchronously with the stimulation and near the stimulation site, exposes the extent of neural recruitment caused by the pulse. Therefore, the measured ECAP can be of great help during implantation surgery and for automatically adjusting the stimulation pulse amplitude and position in daily use.

[0103] Typical neuromodulation pulses have amplitudes of several volts, while the amplitude of the signal visible at the electrode due to neural firing may be only a few microvolts. To be useful, recording of the evoked response must begin within about a hundred microseconds of the end of the stimulation pulse. Making the job of the recording electronics more difficult is that the amplitude of the artifact tail ranges from millivolts to tens of microvolts, depending on the spacing between the stimulation and recording electrodes. The sensor electrodes closest to the stimulation site are the best locations to assess recruitment but have larger artifacts because the ECAPs disperse as they propagate from the recruitment site. Amplifiers designed to process these signals for analog-to-digital conversion, particularly CMOS amplifiers, are required.

[0104] In view of this, a compact model of electrode-electrolyte interface impedance is proposed. The model uses constant phase elements (CPE) and other more common elements in the equivalent circuit representation of the interface between the electrode and the electrolyte. The model is able to accurately represent the impedance of the interface. In the circuit simulator, the model allows the prediction of residual voltage and artifact tailing in some cases (but not all). As seen below, the model predicts that the artifacts in certain circuit configurations are zero, while the measurement results show that significant artifacts actually appear in such configurations. This clearly shows that there are mechanisms at work that are not captured by the single CPE compact model.

[0105] The inventors have therefore introduced a split electrode model. It has been hypothesized that the artifact can be created not only by residual charge left in the double layer around the electrode after the electrode conducts current, but also by an uneven distribution of this charge along the surface of the electrode. Such an uneven charge distribution can occur even if there is no net current flowing through the electrode into the electronic device. One might see that countercharges in the diffuse region of the double layer accumulate at one end of the electrode as a result of surface conduction caused by the field generated by the stimulating electrode. Although for many purposes a more sophisticated model is needed, this idea of ​​charge accumulation at one end of the electrode suggests a straightforward way in which a compact model can be constructed to simulate this phenomenon.

[0106] In the proposed segmented electrode model, the electrode is segmented into multiple separately modeled segments or "slices". The single-branch model of the electrode is replaced by a model with n branches, each contributing 1 / n times the total admittance (area) presented by the original electrode. The branches are connected together at a single junction on the metal side of the electrode, but are connected only through a resistor mesh representing the bulk fluid on the fluid side of the interface (e.g. Fig. 9 and discussed further below).

[0107] Considering the rotationally symmetric case, the simulation is therefore straightforward using a two-dimensional representation, and comparative measurements can be performed with cylindrical implantable platinum electrode arrays, such as an octrode (eight-electrode array) or a dodecatrode (twelve-electrode array). Figure 4 A typical implantable lead electrode array is depicted in FIG. 1 , which has a set of 8 electrodes (cylindrical platinum card skin) and is suitable for use as Figures 1 to 3 The array 150 in the embodiment of the present invention is shown in Figure 1. Each card skin is about 1.3mm in diameter, 3mm in length, and is separated by 4mm insulating segments.

[0108] exist Figure 5 The equivalent circuit of a single card skin is shown in , omitting the diode and memristor for clarity. Figure 5 As can be seen in the figure, the equivalent circuit of a single 3mm card of the electrode array becomes a sequence of n sections, each with its own CPE, and each connected to a different geometric point in the resistor grid representing the electrolyte. Although diodes must be included in such a model except for small-signal simulations, such diodes are omitted in this figure for clarity. For small-signal simulations of scenarios that are safe for use in humans, the diode-memristor branch elements can be ignored, but at least the diodes must be included to detect and model the nonlinear Faraday effect. The basic model parameters of the CPE used in the circuit are obtained, adjusted to show consistency with the observed impedance of electrodes 1 and 2, and scaled according to the division factor used. The base values ​​used for our simulations are m=1.5, |Z|=6500Ω (at 1Hz), and Rs=12Ω. The SPICE equivalent network is generated from 10mHz to 500kHz using a density of k=1.3.

[0109] The resistor grid values ​​are calculated using the measured values ​​of the brine conductivity at 6400 mm and on a grid chosen to be appropriate for the split factor. Figure 7 We will mainly consider one particular circuit arrangement, as graphically depicted in FIG. 1 . The stimulus current is applied to the second cylindrical plug of the lead, Figure 4 and Figure 7 The stimulation current return will be via the first card (electrode 1). The voltage of interest is the voltage that appears between the fourth card and the seventh card (electrode 4 and electrode 7). This represents a typical, clinically desirable arrangement for a feedback implant. The stimulation pulse will be a so-called biphasic pulse, consisting of delivering +5mA for 240μs, zero current for 200μs, -5mA for 240μs, and finally disconnecting from the drive electronics. The form of the stimulation pulse is not particularly closely related to the problem of how the artifact is generated, so the solution proposed below can also be applied to other stimulation schemes.

[0110] This configuration uses a single CPE model for the electrodes in 0.1x phosphate buffered saline (PBS). In the case of the single branch electrode model, assuming no load on the electronics, an absolute zero artifact is predicted between electrodes 4 and 7. However, as the electrode model is split into more and more slices, the prediction changes. Figure 6 This influence is shown significantly. Figure 6 The predicted artifact amplitudes as a function of the number of segments into which the electrode skin is segmented in a SPICE simulation are shown. Figure 6 It is shown that it is necessary to split the electrode into 7 or more slices in the simulation to provide reasonable artifact amplitude accuracy. Therefore, unless otherwise stated, a split factor of 11 will be used in the following. Fig. 9 ) are also divided into slices of equal geometric size, such as 3 / 11 mm or finer, so the number of mesh nodes involved in simulating an electrode extending over 60 mm of electrolyte can become quite large. SPICE requires several seconds to several minutes to run each such simulation on a typical personal computer.

[0111] Comparison with the measured results is performed. A Saluda Medical Pty Ltd "inducing" implant is used to generate the pulses and amplify the measured signal. Wiring arrangement as Figure 7 . The stimulus is delivered to electrode 2 as shown and returned via electrode 1 to ground. The response is measured at electrode 4 relative to electrode 7. 0.1x PBS is prepared using Medicago 09-2051-100 PBS tablets and deionized water at a ratio of 1 tablet per liter. This is a common phantom of tissue in the cerebrospinal fluid (CSF). The received signal is digitized by the implant, but also from the test port provided by the implant IC using a Tektronix TPS2014 isolated input oscilloscope. The same data can be obtained from the implant, but only at a lower sampling rate, approximately 16k samples per second. Due to the limited common mode rejection ratio (CMRR) of the implanted amplifier (no less than 75dB), there is an undefined additive contribution from the common mode signal.

[0112] Figure 8 A 5 mA pulse stimulation voltage 810 (left axis) and a typical measured response signal 820 (right axis) are shown. Figure 7As depicted in , the pulse is applied from the second electrode to the first electrode and the voltage present between the fourth and seventh electrodes is measured. The response 820 (in μV, right axis) measured between the fourth and seventh electrodes is blanked until approximately 800 μs after the stimulation 810 (in volts, left axis) begins. Response 820 is a manifestation of an artifact that obscures the in vivo neural measurements, and it will be noted that the dynamic range of artifact 820 is approximately 150 uV, which is much larger than the voltage observed by the evoked neural response, illustrating the severity of the problem such an artifact presents. The series of small pulses on the received pulse tail 820 are caused by the implanted ADC operation and are not part of the signal. From Figure 8 It can be observed that the front end amplifier (FEA) used to record signal 820 is blanked during pulse 810 and the blanking is released approximately 100μs after the stimulation current is turned off. Since the implantable pulse generator (IPG) FEA is AC coupled, any DC offset is subtracted from the test port voltage. The array of small spikes spaced approximately 31μs apart are clock feedthrough associated with the analog-to-digital conversion inside the implantable chip and should be ignored. Figure 8 Switching spikes can be seen on the stimulation voltage trace 810 in . These cause a contribution to the artifact that does not vary with the stimulation amplitude. As will be seen below, this contributes a constant offset to the amplitude of the measured artifact.

[0113] It turns out that it is difficult to Figure 8 820 in . The inventors now attribute this difficulty to the fact that the artifact is the sum of several components that largely cancel each other out. Therefore, small deviations in any one component that may be associated with a particular contact (contact point, also referred to herein as an electrode) can result in relatively large changes in the final artifact value. We believe that this explains the anecdotal observation that the observed artifact may be better or worse at different contact points for no apparent reason, even in saline where tissue inhomogeneities are not an issue. Our model allows these components to be identified individually, as follows.

[0114] The first such artifact component is that from passive electrodes. During the pulse, all electrodes on the lead are exposed to a voltage gradient along their length. Charge accumulates at one end of each conductive lead compared to the other end. A voltage gradient is created in the electrolyte in which the tissue or lead is immersed, through the dipoles of the stimulating electrode and the return electrode (electrode 2 and electrode 1 in this example). Charge moves more easily along the surface of the lead than the surrounding medium. Once the pulse ends, even an electrode that is not electrically connected and conducts zero net current will acquire a charge imbalance along its surface, which manifests itself as a transient net potential difference between the metal of the lead and the bulk medium.

[0115] Fig. 9 A simplified circuit with 4 electrodes is shown, each electrode being split into only two slices. Although the simulations herein used 11 slices, this two-slice simplified circuit is provided to help visualize the generation of artifacts in the passive electrodes. Fig. 9 The arrows in show that the current induced by the stimulation flows from one end of the passive electrode to the other end of the passive electrode. The inventors have realized that the CPE of each such single electrode is charged by a "circulating current".

[0116] Fig.10 The simulated potential due to the accumulated charge in the interface layer is plotted as a function of position along electrode 4, purely as a result of the 5 mA stimulation current flowing between electrode 2 and electrode 1. For this simulation, the electrode was segmented into 21 slices. Rectangular symbol 1010 indicates the position of the electrode 4. Figure 8 Simulated voltages recorded at the end of the biphasic stimulation pulse depicted in . Point 1020 represents the simulated data at a time of 1 ms, showing relaxation. The electrodes are not connected, so the net electrode current remains zero at all times. Note that the distribution along the length of the electrode is not symmetrical, for example, the slice potential 1010 at a displacement of -1.4 mm is about 2.3 mV, while the slice potential 1010 at a displacement of +1.4 mm is about -1.9 mV. The simulation shows that the average or net potential across electrode 4 is about 100 μV from the time of contact with body saline. The residual potential on a single electrode closer to the stimulation pair is larger, while the residual potential on an electrode farther from the stimulation pair is smaller. The distribution relaxes at the end of the pulse as the charge is redistributed.

[0117] The same mechanism operates at electrode 7, but the result is less than 2 μV compared to the net 100 μV at electrode 4. Therefore, a total contribution of ≈100 μV is made to V(4,7), which is the sensed voltage due primarily to the net voltage between the tissue and metal sides of electrode 4. The first artifact component (i.e., the artifact from the passive electrode) is therefore a significant contributor to the artifact problem.

[0118] The second artifact component is that from the stimulation electrode. The stimulation electrode (referenced 2 in our example) produces a charge gradient about 25 times greater along the length of its surface than electrode 4 because one end of it is closer to the ground return path provided by the adjacent electrode 1.

[0119] Fig.11 The simulated potential caused by the accumulated charge in the interface layer is shown as a function of position along electrode 2, as a result of a 5 mA stimulation pulse. Fig.11 The simulation data shown in Figure 8 The end of the biphasic stimulation pulse 810 depicted in FIG. Fig.10The units of the y-axis in are the same. At the edge of this electrode, even a 5 mA stimulation pulse induces nonlinear effects, but if the current is uniform across the stuck area, the nonlinear effects are expected to be much smaller. This large charge gradient relaxes at the end of the pulse, resulting in a potential difference between all electrodes along the lead. The contribution to the measured voltage V(4,7) is ≈100 μV. Further note that the average deviation from zero represents the charge accumulated after stimulation. The variation in potential along the electrode indicates that the charge accumulation is not uniform. The charge or species concentration in the double layer is not the same at every point along the electrode.

[0120] Ground return electrode 1 responds in the same way as stimulation electrode 2, but its contribution has the opposite polarity and is at a different distance from receive pair 4, 7. Therefore, ground return electrode 1 also makes an artifact contribution to the measured voltage V(4,7). For a 5 mA stimulus, the difference between the contribution from the stimulation electrode and the ground electrode is typically ≈50 μV. The second artifact component (i.e., the artifact from the stimulation electrode) is therefore also an important contributor to the artifact problem.

[0121] The third artifact component is the common mode artifact. Fig.11 As seen in , the total charge accumulated on each stimulation electrode and return electrode between the metal and the dielectric (as opposed to the difference along the skin only on the electrolyte side) is hundreds of mV. The receive amplifier in the prior art Saluda Medical Evoke implant typically has a common mode rejection ratio (CMRR) in the 80 dB range. The circuit ground remains connected to the return current electrode. The average voltage across the ground electrode appears as a common mode signal. This produces an additional artifact of ≈50 μV. The third artifact component (i.e., the common mode artifact) is therefore also an important contributor to the artifact problem.

[0122] By noting these artifact components, it is possible to address the problem of the overall artifact produced when these mechanisms operate together. Fig.12 Artifacts (measured and simulated) are shown as a function of peak stimulation current. The three measured traces 1210, 1212, 1214 represent normal variations between leads that are assumed to be identical. Note the zero offset on the measured traces caused by the common mode signal and the switching signal. The amplitude difference between the simulated trace 1220 and the measured trace is attributed to the ideal properties of a compact model in which all electrodes are identical and the common mode signal does not contribute. In this case, agreement is considered excellent.

[0123] Fig.13 The measured and predicted results of the artifacts measured between adjacent electrode pairs v(3,4), v(4,5), v(5,6), v(6,7) and v(7,8) are shown, the electrode pairs being indicated on the x-axis by their respective displacements from the stimulation electrode (in units of electrode numbers). Fig.13 The measured data shown in Fig.12 Observed shift at zero stimulation current of about 20 μV. As previously described, a 5 mA stimulus was applied to electrode 2 with a ground return through electrode 1. The simulation predicts a contradictory change in the sign of the artifact between the first electrode pair (v(3,4), with a displacement of 1) and the second electrode pair (v(4,5), with a displacement of 2). This change in the sign of the artifact has been observed previously, but appeared contradictory and added considerable difficulty to the problem of resolving the artifact. The present model predicts this sign change, which provides considerable confidence in the authenticity of the model. Moreover, this indicates that there is an artifact zero crossing, or more generally, there is a zero crossing region indicated by 1310, which provides the opportunity for low artifact neural measurements, an observation exploited by the embodiments presented in the present invention.

[0124] The fourth artifact component is the contribution from a single passive skin-attached electrode. The effect of the charge "passively displaced" by the flowing stimulation current is key to understanding and cannot be underestimated. To emphasize this point and aid understanding, consider Figure 14, which is a simplified case of two stimulation electrodes and one passive electrode during a stimulation pulse ( Fig.14a ) and after ( Fig.14b ) is a diagram of the current flow path of a current in a Gouy-Stern-Chapman layer. The current labeled "5" may be 1 / 100 of the currents "1", "2" and "3", but it causes e3 to have a significant displacement charge. The diagram shows how a completely passive metallic structure can become "charged", and that is how it can accommodate the displacement of populations of mobile matter in the Gouy-Stern-Chapman layer.

[0125] then, Fig.15 and Fig.16 The effect of including or omitting a single passive skin on the residual field is shown. Fig.15 is a plot of the electric field in saline surrounding one set of electrodes shortly after stimulation is removed (completed) for two stimulating electrodes and two passive electrodes. The electrodes are placed along the y-axis. The electrodes are completely disconnected so that no current flows into or out of the electrolyte. Fig.16 is a plot of the electric field in saline surrounding a set of electrodes with two stimulating electrodes and one passive electrode shortly after stimulation is removed; Fig.15 compared to, Fig.16 The only change in the figure is the removal of one of the passive electrodes (e3). Comparing the two figures, it is quickly apparent that the addition of a passive disconnected piece of metal ( Fig.15 A new region (around slice 100 near electrode e3) is introduced in which the field changes polarity, passing through a zero point, whereas previously (in Fig.16 Of course, this phenomenon occurs with any metal structure near the stimulation electrodes, not just the components of the implanted leads themselves.

[0126] The observation that the electrode model predicts the artifact only when the electrode is modeled as a series of parallel "slices" that are free to accumulate unequal charges confirms the conjecture that uneven charge distribution on the surface of individual electrodes contributes to the long pulse tailing known as the "artifact" observed on implanted measurement electrodes. Further note that the model suggests that surface charge imbalance is the main or possibly the only contributor to the artifact, because Fig.12 We show that this model can explain the artifact amplitudes observed in well-designed implants with good common-mode rejection ratios in the front-end amplifiers used for recording.

[0127] Thus, the preceding analysis confirms an electrode-intrinsic mechanism that causes electrode artifacts. "Electrode-intrinsic" refers to a phenomenon that is an unavoidable physical consequence of electrode design, is inherent to the electrode's geometry and materials, and is independent of the electrical action, connection, or load of the associated electronic device. While electrical efforts to minimize artifact tailing produced by the electrical action, connection, or load of the associated electronic device in neuromodulation systems are important and ongoing, it is clear that the artifacts measured in this specification inevitably appear in the electrode-electrolyte system itself. Therefore, even a perfect front-end amplifier will encounter this signal.

[0128] With the new understanding that surface charge balance will place a lower (best case) limit on the artifacts that must be added to (i.e., occur simultaneously with and obscure) the evoked compound action potential, it becomes possible and important to consider how electrodes can be designed to intrinsically minimize this phenomenon, and what electrical measures can be taken to accommodate or even beneficially exploit this electrode-intrinsic mechanism and the previous analytical results. The following solution is proposed.

[0129] Specifically, we describe several techniques conceived based on the results of the previous analysis. These techniques include tripolar stimulation, bridging electrodes, and changes in electrode size and shape, any or all of which can be implemented according to embodiments of the present invention.

[0130] Reference again Fig.13 , it should be noted that the sign of the artifact changes with position along the lead as the distance from the stimulation electrode increases. In particular, the artifact is negative when V(3,4) is measured at 1 electrode displacement, but positive when V(4,5) is measured at 2 electrode displacements from the stimulation electrode, and at larger displacements. Fig.13 This is confirmed by both simulation and measurement. Fig.13, but other measurement electrode configurations (such as measuring V(3,5) or V(3,6)) may also experience such artifacts as changes in sign with changes in displacement from the stimulation site.

[0131] The first aspect of the invention recognizes that the change in sign of this artifact with increasing displacement from the stimulation electrode indicates that there is at least one location along the lead where the artifact components sum to zero. We therefore proceed to consider ways in which the artifact can be minimized or substantially eliminated at the location of the measuring electrode by making the location where the artifact crosses zero substantially co-located with the measuring electrode (even if the artifact is non-zero at locations remote from the measuring electrode). We believe this can be accomplished in one or more of a variety of ways, including by changing the layout of the electrodes, changing the interconnections of the electrodes, and / or changing the spatiotemporal distribution of the delivered stimulation current.

[0132] The present embodiment recognizes that an artifact null can be induced by careful tripolar stimulation. In particular, the inventors recognize that the artifact null can be shifted to the 4th electrode by adding a certain contribution current to the 3rd electrode, so that the 2nd and 3rd electrodes in combination provide a stimulation current that returns via the 1st electrode. In general, the principle is to stimulate through three or more electrodes in a way that minimizes artifacts on another electrode or electrodes, usually with asymmetric current distribution.

[0133] In saline, simulations predict that the artifact on electrode 4 will be zero when the stimulation current X into electrode 1 is fed by yX from electrode 2, and (1-y)X from electrode 3. Measurements place these currents close to 0.5X and 0.5X, respectively, with y=0.5.

[0134] Fig.17 A tripolar stimulation method is shown where the current to the external contact point can be varied using a parameter α. Concise terminology is needed to describe the general method of distributing the current between multiple electrodes: the term [E,e,A] is used to describe the proportion "A" of the total current C injected between electrode E and electrode e. In this notation, the neutral electrode (the "earth" point of the system) is defined as electrode 0. This can also be written as [1,2,1]. A list of such definitions describes the complete distribution pattern, for example [1,0,1], [2,0,-1] describes a bipolar stimulation between electrode 1 and electrode 2. Since scaled versions of the same stimulation pattern are applied to different electrodes, the definition is independent of the amplitude or waveform (which can be biphasic, triphasic, or any other waveform). For bipolar and tripolar, the convention is that the second phase is the cathode of the stimulating electrode. This is done to activate the nerve as late as possible.

[0135] Tripolar stimulation is described herein by the general definitions [X,x,A], [Y,y,-1], [Z,z,1-A], where X, Y, and Z are contact point numbers, and x, y, and z are the junctions to which each of the other ends of the current sources are connected. In this document, in most cases, x = y = z = 0, as it allows for a simpler explanation of the current sources between the electrodes. Typically, X < Y < Z indicates that the electrode numbers are sequential, and A < 1. If A = 1 or A = 0, this becomes bipolar stimulation.

[0136] Fig.18 It is shown that tripolar stimulation can be decomposed into two simultaneous bipolar stimulations. Fig.19 It is shown how artifacts from these two components will appear at the amplifier output. They will have opposite phases, and the artifacts from the more distant electrode pair will have a lower amplitude. However, by adjusting the relative amplitudes of these two bipolar stimulations, it is expected that these components will cancel. Note that in Fig.18 the central current source e2 can be omitted and replaced with a ground connection or a switched connection to a voltage supply rail suitable for cathode or anode operation.

[0137] Fig. 20 It is shown the simulated artifacts of tripolar stimulation as the term A varies - showing that it has a zero value as expected. The zero value occurs when A is approximately equal to 0.7. This shows that Fig.19 the method described in Fig. 20 is valid in principle. Fig. 20 It is also shown that the ideal current ratio (ratio approximately 0.4) that produces an artifact zero on E4 is different from E5 to E8, which all enjoy an artifact zero at the same current ratio (approximately 0.7). Therefore, a preferred embodiment will utilize the differential neural measurement results between two of E5 to E8 to avoid E4 and allow for optimal minimization of artifacts simultaneously on two measurement electrodes. This method of varying the current ratio in tripolar stimulation to achieve an artifact zero represents one aspect of the present invention. Other embodiments of the present invention may employ quadripolar stimulation in order to direct the artifact minimum to be in the same location as the measurement electrodes. Other embodiments of the present invention may seek the observed minimum in the artifacts that occur when obtaining the differential measurement results between two measurement electrodes, not when the artifact is zero or at a minimum on each electrode measurement electrode, but when the amplitude of the artifact on each measurement electrode is the same, for example, if using

[0138] The quadrupole can be described by the notations [1,0,A.B], [2,0,A.(1-B)], [3,0,-1], [4,0,1-A], where 0 < A < 1 and 0 < B < 1. "A" describes the charge distribution between electrodes E1 and E2 (which are the proximal and distal ends of E3) and E3, while B defines the charge distribution between E1 and E2. Fig.21 A typical stimulation waveform is shown for the case where A = 0.75 and B = 0.5. From this waveform, it can be observed that the cathodic stimulation on E3 is more than twice the amplitude of the cathodic stimulation on the other electrodes. Thus, E3 will be the only electrode to generate an ECAP. It will also be observed that since there are three electrodes in addition to E3 and there are different combinations of charge distributions on these three electrodes, and the current on none of these electrodes exceeds 50% of E3, there is an opportunity to adjust these ratios to produce a spurious zero while generating only a single ECAP. Fig. 22 The spurious amplitudes on electrodes E4 to E8 are shown when the ratio A is varied. This shows that there is a spurious zero when A is approximately equal to 0.7.

[0139] When A is equal to 0.7 and assuming similar resistances of the electrodes, the current ratios in E1, E2, and E4 will be 35%, 35%, and 30%. When electrodes E1, E2, and E4 are simply connected together, a similar effect is expected, which will result in a distribution ratio of 33%, 33%, and 33% under the constant impedance approximation. This is supported by the clinical observations made by the applicant that generally "adding an anode will reduce the spurious". Since the individual electrode impedances may be different, using separate current sources on each electrode may provide more robust results. Simulations show that when the current distribution is set to 37.5%, 37.5%, and 25%, the spurious reaches zero, and zeros on E5 and E6 occur at similar ratios (E5 is the first available recording electrode in quadrupole stimulation; E4 is in tripolar stimulation). Thus, this quadrupole method represents a further aspect of the present invention. It will be appreciated that the multipole method can be extended as long as the number of return electrodes is greater than 2. For example, it is expected that the simulation curves of [1,0,A / 2], [2,0,A / 2], [3,0,-1], [4,0,(1-A) / 2], [4,0,(1-A) / 2] will produce spurious zeros while maintaining the condition of a single cathode on E3.

[0140] Another embodiment of the present invention is described as a modification of the tripolar stimulation [X,0,A], [Y,0,-1], [Z,0,1-A], where X, Y, and Z are any three electrodes, with Z used as the recording electrode and 0 < A < 1. Thus, in this embodiment, the spurious is made zero by injecting a small amount of current into the recording electrode, and thus the spurious can be titrated.

[0141] Another variant embodiment of tripolar stimulation can utilize automatic adjustment. In this embodiment, parameter A is automatically adjusted to minimize artifacts, and this is embedded in the feedback loop to maintain the induced response at a preset level. This is shown in Figure 23. This embodiment has an ECAP feedback loop composed of an ECAP detector and an SCS loop controller, and the feedback loop produces a current value I. The present embodiment further provides a second feedback loop, which is configured to minimize artifacts, and the second loop includes an AC energy detector, an artifact loop controller, and a current source controller. The second loop controls parameter A, and the parameter A controls the current distribution between electrode 1 and electrode 3.

[0142] The detector of the illusion loop measures the total energy on the recording electrode. This energy will include ECAP and the illusion. However, since the current remains constant, the current distribution between the cathodes will not affect the ECAP. However, this change in distribution will affect the illusion. At the point where the illusion is zero, only the ECAP will remain. The way the illusion zeroing feedback loop works is to first measure the total energy of the signal. This is best done by measuring the standard deviation of the signal samples over the measurement time. Based on the definition of standard deviation, such a measurement is not affected by DC offsets in the signal. Next, the illusion loop controller adjusts the control parameter A. Initially, this change can be in either direction. Next, the illusion loop controller measures the total energy a second time to determine whether it is increasing or decreasing. Next, the illusion loop controller changes the value A in the direction of reducing the detected energy.

[0143] In another embodiment, the architecture of Figure 23 can be adapted for subthreshold determination of a preferred value for parameter A. This embodiment provides an artifact minimization algorithm that delivers a series of stimulations of different stimulation ratios at a subthreshold level, where the subthreshold level is a stimulation level that is non-zero but small enough that it does not recruit any neural response. Artifacts caused by each such subthreshold stimulation at the measuring electrode are then observed to seek a stimulation ratio A that minimizes the artifacts observed at the measuring electrode. This A value is then used to configure the stimulation ratio for suprathreshold stimulation in ongoing therapeutic stimulation. The inventors have discovered that the A value optimized to minimize artifacts at subthreshold stimulation levels is also unexpectedly well suited to minimizing artifacts at suprathreshold stimulation currents.

[0144] exist Fig.23c The current distribution options are shown in . Fig.23c use Figure 23b The boxes are symbols for the electrodes and their connections. The gray boxes indicate the current return (ground) connection for that electrode. The white boxes indicate the current source connection for that electrode and show the current percentage for that electrode. The recording electrode on the right is connected to the amplifier input. For simplicity, Fig.23c The recording electrodes are not shown.

[0145] like Fig.23c As shown in , bipolar stimulation uses a single current source and ground return. Tripolar stimulation uses two ground returns, usually on either side of the stimulation electrode. This is referred to as a "passive triode" in this document to avoid ambiguity. In contrast, an active triode grounds the central contact point and connects the current source to the outer electrodes. If the electrode impedances are equal, a 50 / 50 active triode behaves similarly to a passive triode, but the current distribution ratio will vary with tissue impedance and therefore over time as tissue grows around the stimulation electrode. Therefore, an active triode is a method superior to a passive triode. Both passive triodes and 50 / 50 active triodes reduce artifacts compared to bipolar stimulation, particularly for adjacent electrodes (as already discussed herein). The inventors have recognized that a particularly preferred method is to use an active triode with a current distribution ratio of 75 / 25, because this produces zero values ​​at several recording contact points simultaneously. In some embodiments, the active triode or multipole can even provide current sources for all stimulation electrodes, where no stimulation electrode is grounded, as long as the net cathode current in each stimulation phase is balanced with the net anode current in that phase. In such embodiments, any mismatch in current source matching can be resolved according to the teachings of International Patent Publication No. WO 2014071446, the contents of which are incorporated herein by reference.

[0146] A further approach is to use a passive quadrupole. Assuming equal tissue impedance on all return electrodes, this achieves a current sharing ratio of 33 / 33 / 33, but it should be noted that this simple assumption ignores the unequal distances of the corresponding ground return electrodes, which will tend to cause unequal return current sharing ratios. In any case, the passive quadrupole is preferred over the passive tripolar because it produces lower artifacts due to this difference in current sharing. However, the passive quadrupole does require space on the lead to place additional stimulation electrodes. The active quadrupole sharing ratio of 37.5 / 37.5 / 25 has similar artifact characteristics to the tripolar sharing ratio of 75 / 25. There are many variations of these methods, depending on the exact patient situation. However, the principles have been described, and it will be appreciated that such other variations are within the scope of the present invention.

[0147] Fig.23d Further shown are multipolar stimulation configurations according to some embodiments of the present invention as well as some measurement electrode configurations that may be used. Fig.23c A stimulation configuration for use with an implantable lead comprising 12 electrodes is shown. Fig.23c Electrodes 1 to 3 are shown for tripolar stimulation, and electrodes 1 to 4 are shown for quadripolar stimulation, but as previously described with respect to Figure 1 and Figure 2As described, other embodiments may deliver stimulation from any suitable subset of the twelve electrodes. Fig.23d , each row shows a different electrode stimulation configuration. For example, stimulation configuration B shows a tripolar stimulation configuration in which one central stimulation electrode (2) and two lateral return electrodes (1 and 3) are connected together. If the tissue is homogeneous, 50% of the current will be returned via each return, but, as this varies with tissue impedance and possibly also with posture, the individual return currents cannot be ensured. Therefore, each return current value is shown in brackets, and the exact value of the corresponding return current through return electrodes 1 and 3 (or for example electrodes 1, 2 and 4 in configuration E) will depend on the patient anatomy. Again, as shown in row D, a particularly preferred approach is to use an active triode with a current sharing ratio of 75 / 25, in which the larger return current is forced away from the measuring electrode. As Fig.23d As shown in the stimulation configuration of , any other electrode of the lead can be selected as a measurement electrode, and any one of the three or four electrodes closest to the stimulation can be selected as a sensing electrode, with the farthest electrode (12) serving as a measurement reference electrode.

[0148] It will be noted that Fig.23c Certain stimulation configurations presented in, or more generally throughout, other described embodiments may not produce artifactual zero crossings. Additionally or alternatively, depending on the type of measurement circuitry used, only the power output of the sensed signal may be retrieved. In this case, zero crossing artifacts may not exist or may not be detected. Regardless, however, current embodiments of the invention recognize that artifactual spatial minima may be produced and that measures may be taken to "guide" or intentionally position the minima on or near the measurement electrodes.

[0149] A further method of reducing artifacts can be to use a pair of electrodes not involved in stimulation or measurement to direct the zero point to the measuring electrode, e.g. Fig.24 as shown in .

[0150] exist Fig.24 In the example of , simply adding a resistor between electrode 3 and electrode 5 will allow the artifact on electrode 4 to be reduced to near zero by adjusting the junction resistance. For the above case, the simulation predicted a value of 400 ohms. The measurement result is close to 200 ohms, which is a good match. This resistance can be provided by switching the impedance achieved by the pulse width modulation connection. There may be multiple such "bridge" connections that allow the artifact amplitude to be controlled, and all such bridge resistor arrangements are within the scope of the present invention.

[0151] A third approach to minimizing artifacts is to vary the number, size, and deployment of electrodes in a multi-electrode lead to minimize artifacts on electrodes adjacent to the stimulating electrodes.

[0152] Reference again Fig.13 . The graph is measured on a standard lead with a period of 7mm with 3mm electrodes and 4mm of insulating space. At its simplest, the lead can be constructed with smaller sensing electrodes at the zero crossings where minimal artifacts will be measured. Simulations show that the third and fourth electrodes help produce the zero value, so they are left as is and a smaller electrode is added in the correct position to locate the zero point between the third and fourth electrodes. There will be more than one pattern of electrodes and insulators that will zero the artifacts of the electrodes, and such variations are within the scope of the invention.

[0153] A further solution provided by the present invention involves an improved electrode design which, in a preferred embodiment, reduces artefacts by a factor of 1 / 3 compared to prior art electrode designs. Fig.25 The upper portion of the diagram shows a 4-contact version of a prior art epidural electrode. Most such electrodes have 8 contacts, each 3 mm long and spaced 7 mm apart as shown. In contrast, Fig.25 The preferred embodiment shown in the lower portion of provides 2 mm long contacts with a pitch of 7 mm.

[0154] exist Fig.25 In the present invention, some electrodes are marked as 'S' to indicate stimulation electrodes, and some are marked as 'R' to indicate recording electrodes. It will be understood that the electrodes are substantially the same and any electrode can be used for stimulation and recording. However, this nomenclature is introduced at this time to support the subsequent description. Fig.25 A scenario is shown where a 4-contact point lead can be used to feedback stimulation in an SCS system, with two stimulation electrodes and two recording electrodes. As will be appreciated, in a system with 8 or 12 electrodes, more electrode combinations can be used.

[0155] Fig.26 shows a circuit model of a single electrode in a saline bath that is subject to a current flowing along the electrode. In the SCS case, this would be head-to-tail. In this figure, the three-dimensional structure of the saline bath is modeled as a two-dimensional array of resistors, and the electrode is modeled as a series of constant phase elements (CPEs), as described previously. However, in Fig.26 In the diagram, the resistor grid and CPEs have been expanded to show higher resolution, and the electrode becomes a series of CPEs directly connected on the metal side but distributed on the saline side. Experiments have found that 7 or more slices are needed to adequately describe a 3mm electrode, but 4 slices are shown in this figure for clarity, and this is a discrete model of the continuous case, so not relevant to the results. In this figure, the resistors at the perimeter are not connected. In practice, this grid would end at the limit of the physical volume it models.

[0156] It will be appreciated from this figure that if a voltage is applied between the right and left sides of the grid, current will flow, and some of the current will flow into the leftmost CPE, then into the metal electrode, and then out of the rightmost CPE. Current will also flow symmetrically (but to a lesser extent) through the center two CPE elements. Thus, at the end of a certain period, the voltage along the surface of the array (relative to the metal) will be as follows: Fig. 27 as shown in .

[0157] Now consider Fig.28 A pair of adjacent stimulation electrodes is shown in Figure 1. The current will flow preferentially to the area that is closest to the metal. After a period of stimulation, the voltage on the electrode surface relative to the metal will be as follows: Fig.29 as shown in .

[0158] Now consider Fig.30 A pair of adjacent recording electrodes is shown in Figure 1. The stimulation pair on the left but not shown will produce a vertical current gradient in the grid. This will spread throughout the resistor grid, affecting the closer recording electrodes more than the more distant ones. This current will be a small fraction of the total stimulation current and is therefore indicated by a lowercase i. After a period of stimulation, the voltage at the surface of the recording electrode relative to the metal will be as follows Fig.31 as shown in .

[0159] This again illustrates the main mechanism for the artifact generation: the voltage along the recording electrode is larger for electrodes closer to the stimulation source, and the gradient on the stimulation electrode is largest at the edge where the stimulation contacts are close to each other. The artifact occurs when this charge is redistributed during recording, resulting in a varying differential voltage between the recording electrode metal contacts. The simulations above show that these effects are comparable in magnitude.

[0160] These phenomena are caused by dipoles generated on the stimulating and recording electrodes. The electric moment generated by the dipoles is proportional to the size and distance between the opposite charges. Therefore, reducing the length of the dipole will reduce the field and therefore reduce the artifact. The dipoles become unbalanced (e.g., due to the close proximity of the stimulating electrodes) Fig.29 ), so increasing the gap between the electrodes is expected to result in a more symmetric charge distribution and therefore a smaller dipole. Simulations show that reducing the length of the electrodes from 3 mm to 2 mm reduces the artifacts by a factor of three. This suggests that both mechanisms are present.

[0161] These phenomena can be seen in simulations of this shortened electrode configuration. The simulations were performed using the methods described previously. Fig.32 A compression algorithm for highlighting phenomena in a two-dimensional field diagram is shown. Fig.33 Stimulus waveforms are shown to provide a reference for subsequent figures. Fig.34Shown are the compressed fields in a saline bath at different times in response to biphasic stimulation. Fig.35 The uncompressed voltage along the electrode surface is shown.

[0162] The effects of reducing the electrode size on the system include the following. The impedance of the electrode varies with its area and therefore its length. Reducing the electrode length from 3mm to 2mm will increase its impedance (as simulated in a saline bath of 1 / 10 saline in tripolar mode) from 750 ohms to 950 ohms. This will increase the power required to drive the electrode and therefore shorten battery life. However, this tradeoff is acceptable and can be recovered by other system design changes, such as the improved current delivery mechanism described in Australian Provisional Patent Application No. 2018900480, the contents of which are incorporated by reference.

[0163] Another system impact involves safe charging. The maximum charge that can be delivered by the electrode before unsafe free radicals (such as Cl and H) are generated depends on the electrode area. A 3mm×1.3mm electrode can be used to deliver 14.5uC. Since most patients need a charge below 7uC for comfort, and a 2mm long electrode can deliver 9.7uC before reaching its unsafe limit, an electrode length of 2mm is more than enough.

[0164] Although Figure 25 to Figure 35 Related to the lead electrodes, the paddle electrode contacts can also be made shorter in both the craniocaudal dimension to ameliorate artifacts to which such arrays are subject. Fig.36 The right side of FIG. 1 shows a contact point shortened for this purpose according to one embodiment of the present invention.

[0165] Fig.37 Another embodiment of the present invention is shown which exploits the previous observations. In this embodiment, each stimulation electrode is split into two loops which are driven independently. Splitting each electrode into two parts ensures that each part of the electrode carries the same current, i.e., I / 2. Thus, this arrangement will cancel out the asymmetric voltages generated on a single electrode, such as Fig.35 Similarly, splitting the recording electrode during stimulation will prevent current from passing through the ring, thereby counteracting the development of voltage asymmetry on the recording electrode. If desired, during the recording phase, the two parts of the recording electrode can be electrically connected together by switch 3702. The second recording electrode (not shown) will have a similar arrangement. Although Fig.37 The embodiment of the present invention utilizes two current sources for each stimulation electrode, one of which is associated with a respective portion of the electrode, but in another embodiment, the current flowing along the ring of electrodes can be attenuated by using split electrodes and increasing the resistance of the wire, such as Fig.38This resistor can be added as a separate component or by using hookup wire with insulated strands where each strand is connected to a single ring.

[0166] Fig.39 A further variation is shown in . This embodiment involves increasing the resistance in a direction-specific manner, whereby the contact material itself is used to preferentially increase the resistance in a direction along the length of the contact. As previously described, this resistance inhibits the flow of current and the development of asymmetric voltage distributions associated with the artifact. Fig.39 In an embodiment of the invention, additional longitudinal resistance is added by cutting slots in the ring. To achieve a resistance comparable to tissue impedance, the slots increase the current path length and therefore the resistance. The ring is fed by a contact point in the middle. Such slots can be made using laser etching.

[0167] Certain embodiments of the present invention may utilize 3D printing to construct a device. Thus, in some embodiments, the present invention may reside in a digital blueprint comprising a digital file in a format configured for use with rapid prototyping and computer-aided design (CAD) and / or manufacturing, such as an STL (stereolithography) file format. Such digital blueprint files (whether generated by performing three-dimensional scanning of embodiments of the present invention, or generated by CAD development software tools, etc.) are within the scope of the present invention.

[0168] Some embodiments of the present invention may be implemented in conjunction with other techniques for artifact minimization or repair, including, for example, using a triphasic stimulation technique according to the teachings of the applicant's International Patent Publication No. WO 2017219096, the contents of which are incorporated herein by reference.

[0169] Those skilled in the art will appreciate that, without departing from the spirit or scope of the invention as broadly described, many changes and / or modifications may be made to the invention as shown in the specific embodiments. Therefore, the present embodiments should be considered in all aspects as illustrative rather than restrictive or limiting.

Claims

1. A neural stimulation device comprising: at least one stimulation electrode configured to deliver tripolar electrical stimulation to neural tissue via three stimulation electrodes including a central electrode carrying a stimulation current and two peripheral electrodes carrying two respective portions of the stimulation current to maintain charge balanced stimulation; as well as at least one measuring electrode configured to record the response of the neural tissue to the tripolar electrical stimulation, wherein the artifact appears on the measuring electrode as a function of the distance between the stimulation electrode and the measuring electrode; wherein the corresponding parts of the stimulation current carried by the peripheral electrodes are mismatched in such a way that a minimum area of ​​the artifact function occurs on the measuring electrode; and Therein, the device is configured to adaptively change a stimulation ratio, which is a ratio between respective portions of the stimulation current carried by the peripheral electrodes, depending on which measurement electrode or electrodes are used.

2. The nerve stimulation device according to claim 1, wherein The artifacts produced at the measuring electrode are less than 75% of the peak artifacts produced in a spatial region far from the stimulation electrode.

3. The nerve stimulation device according to claim 2, wherein: The artifacts produced at the measuring electrode are less than 50% of the peak artifacts produced in a spatial region far from the stimulation electrode.

4. The nerve stimulation device according to claim 3, wherein: The artifacts produced at the measuring electrodes are less than 25% of the peak artifacts produced in a spatial region far from the stimulation electrodes.

5. The nerve stimulation device according to any one of claims 1 to 4, wherein: The minimum region of the artifact includes a zero-crossing region of the artifact.

6. A neural stimulation device as described in claim 1, configured to perform artifact minimization by delivering a series of stimulations with different stimulation ratios at subthreshold levels that do not recruit neural responses, observing the artifacts caused by each such stimulation at the at least one measuring electrode, and seeking a stimulation ratio that minimizes the artifacts observed at the at least one measuring electrode used.

7. A method for nerve stimulation, the method comprising: delivering tripolar electrical stimulation to the neural tissue via three stimulation electrodes including a central electrode carrying a stimulation current and two peripheral electrodes carrying two respective portions of the stimulation current to maintain charge balanced stimulation; as well as recording the response of the neural tissue to the tripolar electrical stimulation using at least one measuring electrode, wherein the artifact appears on the measuring electrode as a function of the distance between the stimulation electrode and the measuring electrode; wherein the corresponding parts of the stimulation current carried by the peripheral electrodes are mismatched in such a way that a minimum area of ​​the artifact function occurs on the measuring electrode; and Depending on which measuring electrode or electrodes are used, the stimulation ratio is adaptively varied, the stimulation ratio being the ratio between the respective portions of the stimulation current carried by the peripheral electrodes.

Images