Automatic Calibration in an Implantable Stimulator Device Having Neural Sensing Capability

Abstract

System and methods are disclosed to automatically set or update physiological thresholds such as perception threshold (pth) and discomfort thresholds (dth) in an implantable stimulator system. The system monitors neural responses such as ECAPs resulting from stimulation provided to the patient. Extracted neural thresholds (ENTs) are determined, which can comprise a smallest stimulation amplitude at which a neural response can be reliably detected. A correlation between ENTs and physiological thresholds such as pth and dth is used to allow the physiological thresholds to be estimated and updated using the measured ENT values.

Claims

1. A method for determining one or more first thresholds for therapeutic stimulation provided to a patient by an implantable neurostimulator, comprising: (a) providing the therapeutic stimulation to the patient via the implantable neurostimulator, wherein the therapeutic stimulation comprises a plurality of stimulation parameters; (b) determining a value for a neural response, wherein the neural response is formed in response to the therapeutic stimulation; and (c) determining one or more first thresholds for the therapeutic stimulation using the determined value for the neural response, wherein each first threshold is determined using a first mathematical relationship that models each first threshold as a function of values of the neural response.

2. The method of claim 1, wherein the one or more first thresholds comprise thresholds for one of the stimulation parameters that causes a physiological response in the patient, wherein the physiological response comprises one or more of paresthesia and discomfort.

3. The method of claim 1, wherein the one or more first thresholds comprise physiological thresholds, wherein the one or more physiological thresholds comprise one or more of a perception threshold and a discomfort threshold.

4. The method of claim 3, wherein the one or more first thresholds comprise thresholds for an amplitude of the therapeutic stimulation.

5. The method of claim 1, wherein the therapeutic stimulation is provided to the spinal column of the patient, and wherein the neural response comprises an Evoked Compound Action Potential.

6. The method of claim 1, wherein the value for the neural response comprises a value of one of the stimulation parameters, wherein the value for the neural response comprises a minimum value of the one of the stimulation parameters at which the neural response is detectable.

7. The method of claim 6, wherein the one of the stimulation parameters comprises an amplitude of the therapeutic stimulation.

8. The method of claim 1, wherein the first mathematical relationship that models each first threshold is a linear function of the values of the neural response.

9. The method of claim 1, wherein the value for the neural response comprises an extracted neural threshold.

10. The method of claim 1, wherein an external device communicates with the implantable neurostimulator.

11. The method of claim 10, wherein the value for the neural response is determined in the external device.

12. The method of claim 10, wherein the method is initiated at a user interface of the external device.

13. The method of claim 10, wherein the first mathematical relationship for each of the first thresholds is stored in the external device.

14. The method of claim 13, wherein the one or more first thresholds is determined in the external device.

15. The method of claim 1, further comprising prior to step (a), providing test stimulation to the patient via the implantable neurostimulator, wherein the test stimulation is provided at a plurality of different test pulse widths.

16. The method of claim 15, further comprising determining values for a neural response at each of the test pulse widths, wherein the neural response is formed in response to the test stimulation.

17. The method of claim 16, further comprising determining a second mathematical relationship that models values for the neural response as a function of pulse width using the values for the neural response as determined at each of the test pulse widths.

18. The method of claim 17, wherein in step (a) the therapeutic stimulation is provided to the patient at a therapeutic pulse width, wherein in step (b) the value for the neural response is determined using the second mathematical relationship determined at the therapeutic pulse width.

19. A system, comprising: an external device configured to communicate with an implantable neurostimulator, wherein the external device is configured to: (a) program the implantable neurostimulator to provide therapeutic stimulation to a patient, wherein the therapeutic stimulation comprises a plurality of stimulation parameters; (b) determine a value for a neural response, wherein the neural response is formed in response to the therapeutic stimulation; and (c) determine one or more first thresholds for the therapeutic stimulation using the determined value for the neural response, wherein each first threshold is determined using a first mathematical relationship that models each first threshold as a function of values of the neural response.

20. A non-transitory computer readable medium comprising instructions executable on an external device configured to communicate with an implantable neurostimulator, wherein the instructions are configured to: (a) render a user interface on the external device to allow a user program the implantable neurostimulator to provide therapeutic stimulation to a patient, wherein the therapeutic stimulation comprises a plurality of stimulation parameters; (b) determine a value for a neural response, wherein the neural response is formed in response to the therapeutic stimulation; and (c) determine one or more first thresholds for the therapeutic stimulation using the determined value for the neural response, wherein each first threshold is determined using a first mathematical relationship that models each first threshold as a function of values of the neural response.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG. 1 shows an Implantable Pulse Generator (IPG), in accordance with the prior art.

[0024] FIGS. 2A and 2B show an example of stimulation pulses producible by the IPG, in accordance with the prior art.

[0025] FIG. 3 shows stimulation circuitry useable in the IPG, in accordance with the prior art.

[0026] FIG. 4 shows various external devices capable of communicating with and programming stimulation in an IPG, in accordance with the prior art.

[0027] FIG. 5 shows an improved IPG having neural response sensing, and the ability to adjust stimulation dependent on such sensing.

[0028] FIG. 6 shows stimulation producing a neural response, and the sensing of that neural response at at least one electrode of the IPG.

[0029] FIG. 7 shows different physiological thresholds such as pth and dth, and how they relate to amplitude I of stimulation.

[0030] FIG. 8 shows a Graphical User Interface (GUI) of an external device such as a clinician programmer, and use of the interface to set stimulation for a patient and to determine physiological thresholds for that stimulation.

[0031] FIG. 9 shows graphs relating physiological thresholds such as pth and dth to extracted neural thresholds (ENTs).

[0032] FIG. 10 shows an algorithm using measured ENTs to determine physiological thresholds using the relationships of FIG. 9.

[0033] FIG. 11 shows strength-duration curves, and modeling ENTs in accordance with such curves to establish ENTs as a function of duration (pulse width).

[0034] FIG. 12 shows an algorithm using ENTs measured at different pulse widths to determine physiological thresholds using the relationships of FIG. 9 and the ENT modelling of FIG. 11.

DETAILED DESCRIPTION

[0035] An increasingly interesting development in pulse generator systems, and in Spinal Cord Stimulator (SCS) pulse generator systems specifically, is the addition of sensing capability to complement the stimulation that such systems provide. For example, and as explained in U.S. Patent Application Publication 2017/0296823, it can be beneficial to sense a neural response in neural tissue that has received stimulation from an SCS pulse generator.

[0036] FIG. 5 shows circuitry for an SCS IPG 100 having neural response sensing capability. The IPG 100 includes control circuitry 102, which may comprise a microcontroller for example, such as Part Number MSP430, manufactured by Texas Instruments, which is described in data sheets at http://www.ti.com. Other types of control circuitry may be used in lieu of a microcontroller as well, such as microprocessors, FPGAs, DSPs, or combinations of these, etc. Control circuitry 102 may also be formed in whole or in part in one or more Application Specific Integrated Circuits (ASICs) in the IPG 100 as described earlier, which ASIC(s) may additionally include the other circuitry shown in FIG. 5.

[0037] FIG. 5 includes the stimulation circuitry 28 described earlier (FIG. 3), including one or more DACs (PDACs and NDACs). A bus 118 provides digital control signals to the DACs to produce currents or voltages of prescribed amplitudes (I) and with the correct timing (PW, F) at the electrodes selected for stimulation. The electrode current paths to the electrodes 16 include the DC-blocking capacitors 38 described earlier.

[0038] The control circuitry 102 is programmed with a neural response algorithm 124 to evaluate a neural response of neurons that fire (are recruited) by the stimulation that the IPG 100 provides. One such neural response depicted in FIGS. 5 and 6 is an Evoked Compound Action Potential, or “ECAP,” although other types of neural responses also exist and can be sensed by the IPG 100. As its name implies, an ECAP comprises a compound (summation) of various action potentials issued from a plurality of recruited neurons, and its amplitude and shape varies depending on the number and type of neural fibers that are firing. Generally speaking, an ECAP can vary between tens of microVolts to tens of milliVolts. The neural response algorithm 124 assesses the ECAP and can, for example, adjust the stimulation program in a closed loop fashion. In this regard, the neural response algorithm 124 can attempt to remove other signals that may be present at the sensing electrode (stimulation artifacts, background signals and noise) and determine one of more features indicative of the size and shape of the ECAP (e.g., peak-to-peak heights, peak areas, line lengths, etc.). Such ECAP features are described in further details in U.S. Patent Application Publication 2020/0305744, which is incorporated herein by reference in its entirety.

[0039] The control circuitry 102 and/or the neural response algorithm 124 can also enable one or more sense electrodes (S) to sense the ECAP, either automatically or based on a user selection of the sense electrode(s) as entered into an external device (see FIG. 4). As shown in FIG. 6, the ECAP will be initiated upon stimulation of neural fibers in a recruited neural population 95 proximate to the electrodes chosen for stimulation (e.g., E1 and E2), and will move through the patient's tissue via neural conduction. In the simple example of FIG. 6, electrode E6 is chosen as a sense electrode S, and thus this electrode will detect the ECAP as it moves past. The speed at which the ECAP moves depends on the several factors, and is variable.

[0040] To assist with selection of the sensing electrode(s), and referring again to FIG. 5, each electrode node ei 39 is made coupleable to at least one sense amp 110. In this example, for simplicity, all of the electrode nodes are shown as sharing a single sense amp 110. Thus, any one sensing electrode (e.g., electrode node e6) can be coupled to the sense amp 110 (e.g., Ve6) at a given time per multiplexer 108, as controlled by bus 114. However, although not shown, each electrode node can also be coupleable to its own dedicated sense amp 110. ECAP sensing can also involve differential sensing of the ECAP at more than one electrode (e.g., at electrodes E5 and E6), and thus two electrode nodes (e.g., Ve5 and Ve6) can be input to a differential sense amp 110, as explained for example in U.S. Patent Application Publication 2020/0305744. After the ECAP is sensed, the analog waveform comprising the ECAP is preferably converted to digital signals by an Analog-to-Digital converter 112, which may also reside within the control circuitry 102. The neural response algorithm 124 can then assess the size and shape of the ECAP as already described, and if necessary, make adjustments to stimulation via bus 118. In an alternative, the neural response algorithm 124 may also transmit data to an external device such as the clinician programmer 70 or the patient external controller 60 (FIG. 4), which permits those external systems to perform some of the processing. For example, the algorithm 124 can cause the IPG 100 to transmit digital or analog representations of the waveforms, leaving the external system to determine ECAP features from the waveforms. The algorithm 124 may also cause the IPG 100 to transmit determined ECAP features to the external systems for analysis. In this regard, notice that all or a portion of the neural sensing algorithm 124 can operate in an external system, such as in conjunction with the clinician programmer's software 84 (FIG. 4).

[0041] Stimulation in IPG 100 can be provided with reference to a number of different physiological thresholds, which will different from patient to patient. Generally speaking, a physiological threshold comprises a threshold that causes a physiological response in the patient. Reaching a physiological threshold may be perceptible to the patient, such as paresthesia or discomfort. FIG. 7 shows an example of two particular physiological thresholds, which are called the perception threshold (pth) and the discomfort threshold (dth). Other physiological thresholds also exist, but are not shown for simplicity. In this example, the physiological thresholds are expressed in terms of current amplitude I of the stimulation therapy that is provided to the patient, although other parameters of the stimulation therapy (e.g., pulse width, frequency) could be used to define these thresholds as well. The perception threshold pth comprises a lowest amplitude at which the patient can still feel the stimulation (as paresthesia), or as a highest amplitude at which the patient cannot feel the stimulation. As such, the perception threshold pth demarks a boundary between sub-perception stimulation therapy (I<pth) and supra-perception stimulation therapy (I>pth). The discomfort threshold dth comprises a highest amplitude at which stimulation is still comfortable for the patient. In other words, stimulation amplitudes above dth (I>dth) are uncomfortable for the patient, and thus these higher amplitudes are generally avoided.

[0042] It is normally useful for the clinician to determine at least one of these physiological thresholds for a given patient, because this can be useful to programming or controlling the patient's stimulation therapy. For example, pth can be important to determine if it is desired that a patient receive sub-perception stimulation therapy or supra-perception therapy. dth can be important to determine to ensure that stimulation therapy does not cause patient discomfort. Physiological thresholds can be determined by the clinician using the GUI 82 of the clinician programmer 70 as shown in FIG. 8, which can also be used to generally determine optimal stimulation parameters for the patient. The GUI 82 as shown in FIG. 8 is fairly basic, and an actual implementation can be more complicated and can provide more options to tailor stimulation and/or to provide active or passive charge recovery schemes, although these more advanced details aren't shown. In this example, the GUI 82 includes a lead interface 130 which shows the electrode array 17 as implanted in the patient. The lead interface 130 can show the relative positions of the leads in the electrode array to each other, and can also show the position of the leads relative to the patient's tissue. For example, although not shown, the locations of the patient's vertebrae can also be displayed in the leads interface 130.

[0043] The GUI 82 can also include a stimulation parameters interface 132 which is used to set the stimulation parameters of the stimulation that the patient will receives. This can include means to adjust the amplitude (I), pulse width (PW) and frequency (F) of the stimulation pulses. The GUI can also include means to set the location of stimulation in the electrode array 17. This can involve selecting active electrodes (E), the polarity of those active electrodes (P; anode or cathode), and the percentage of amplitude I (X %) that each active electrode should receive. In this example, electrode E1 has been selected as an anode and E2 as a cathode, with each receiving 100% of current amplitude I as an anodic current (+I) and as a cathodic current (−I). However, and as mentioned earlier, more than one electrode can be selected as an anode and more than one electrode can be selected as a cathode at a given time by sharing the anodic or cathodic current between those anode/cathode electrodes, as dictated by percentage X %. Sharing the anodic and cathodic currents between different numbers of electrodes can set the position of anode and cathode poles (+ and −) as virtual poles between the physical location of the electrodes, as is known. Typically, the location of the stimulation in the electrode array 17 can be manipulated by the clinician, such as by using a computer mouse to move the location of the stimulation within the leads interface 130, and this is done with the goal of locating a position that treats the patient's symptoms (e.g., pain). Note that an electrode configuration algorithm can be used to automatically determine active electrodes (E), polarities (P), and percentages (X %) as the clinician positions the stimulation in the electrode array, as explained further in U.S. Pat. No. 10,881,859, which is incorporated herein by reference in its entirety. GUI 82 can also include a program interface 134 to allow a clinician to store and load stimulation programs for the patient.

[0044] Once generally optimal stimulation parameters (I, PW, F, E, P, X) have been determined for the patient, the clinician can determine one or more physiological thresholds discussed earlier. This can generally involve an amplitude “sweep” where the amplitude is set to 0 and is gradually increased until the patient first starts to perceive the stimulation (which establishes pth). Further increasing the amplitude until the patient experiences discomfort similarly establishes dth. Once these thresholds have been determined in this manner, they can be stored by the clinician in a threshold interface 136 in the GUI 82 along with the other stimulation parameters, which then allows these physiological thresholds to be used in setting or controlling the patient's stimulation. For example, once dth is set, the GUI 82 may set this as a maximum value for amplitude I, and may limit amplitude I adjustments to values lower than this maximum for patient safety. In another example, if the clinician decides that the patient should receive paresthesia-based therapy (i.e., supra-perception therapy where the patient perceives a sensation produced by the stimulation), the GUI 82 may set pth as an amplitude minimum, while also setting dth as an amplitude maximum for safety. Percentages of these values can be used as well. For example, the maximum amplitude may be set to 90% of dth to ensure some guardband against patient discomfort. Likewise, if the clinician decides that the patient should receive paresthesia-free (sub-perception) therapy, the GUI 82 may set pth as an amplitude maximum. Again, the maximum amplitude may be limited to a percentage of pth, such as 90% of pth to guardband against the possibility that the patient may feel the stimulation. The optimal stimulation parameters and any relevant physiological thresholds such as pth and/or dth can also be transmitted to and stored in the patient's external controller 60 and/or the patient's IPG 100, as shown in FIG. 8. Having the physiological thresholds transmitted to such patient devices is useful to control the extent to which a patient can adjust his situation therapy, such as by limiting amplitude adjustments to different treatment regimes (supra-perception, sub-perception, sub-discomfort, etc.) as just described.

[0045] While establishing physiological thresholds such as pth and dth can be useful, it does take some time for the clinician to perform. This can create problems for the clinician when trying to determine optimal stimulation parameters for the patient. As noted above, the clinician can attempt to move the location of the stimulation in the electrode array to try and find a location that best treats the patient's symptoms. Typically, the values of physiological thresholds such as pth and dth will change as the location of the stimulation changes. This can mean that the clinician may need to determine these thresholds at each new stimulation location, which as noted takes some time to manually establish.

[0046] Another shortcoming to determining pth and dth as described is that these thresholds are typically set once at the beginning of stimulation therapy, and may thereafter only be altered by the clinician from time to time. This is unfortunate, because it may be useful to adjust such thresholds in between clinician visits. In this regard, it is known that it can be necessary to adjust a patient's stimulation, because the stimulation environment has changed. If a patient changes position, such as going from sitting to standing, this can bring the electrodes closer to or farther from the spinal neural tissue. This would suggest that the intensity of stimulation (e.g., amplitude) may need to be decreased or increased to bring about the same therapeutic effect when treating a patient's symptoms. Scar tissue or changes to the electrode/tissue interface may also naturally change over time, which would also suggest that it may be beneficial to adjust a patient's stimulation. It would be expected that such changes to the stimulation environment would suggest the need to adjust the physiological thresholds. For example, if it is necessary to generally increase the amplitude of simulation given such environmental changes, it would be expected that pth and dth should also increase. However, pth and dth as just discussed are typically set or adjusted by the clinician only infrequently, as described above.

[0047] It would be beneficial to automatically change or update physiological thresholds such as pth and dth using measurements taken from the patient. This would allow clinician to more quickly establish values for such thresholds, and would allow such thresholds to be adjusted even after leaving a clinician's office. In this disclosure, neural response measurements are used to estimate, adjust, and set therapeutic thresholds. More specifically, extracted neural thresholds (ENTs) are determined. An ENT may be expressed in terms of current amplitude I of the stimulation therapy that is provided to the patient, and comprises the minimum amplitude at which a neural response can be reliably detected, as described further below. The inventors have noticed a correlation between ENTs and physiological thresholds such as pth and dth, which allows such thresholds to be estimated and updated using measured ENT values. In particular, the inventors have noticed a parallel between the strength-duration curves for ENTs and physiological thresholds such as pth and dth, which again allows physiological thresholds to be estimated and updated using measured ENT values. This is beneficial, because ENTs can be objectively measured, which allows physiological thresholds to be automatically and quickly adjusted on the fly. This both assists the clinician in determining physiological thresholds, and also allows for updating of these thresholds without a clinician's assistance.

[0048] An extracted neural threshold (ENT) as just noted may be expressed in terms of current amplitude I of the stimulation therapy that is provided to the patient, as shown in FIG. 7, and comprises the smallest amplitude at which an neural response to the stimulation can be reliably detected. One such neural response comprises an ECAP, which as mentioned above is a small signal which may be difficult to detect. The ability to detect an ECAP—and thus determine an ENT value (in mA)—depends on many factors, such as the design of the sense amp(s) 110 (FIG. 5), the resolution with which the sensed ECAP is sampled (via ADC 112), and the particulars of the neural response algorithm 124 (which as noted above can operate at least in part in external systems as well). Furthermore, sensing an ECAP is made difficult because the voltage in a patient's tissue can vary, both as a result of the stimulation and background noise in the tissue. See, e.g., U.S. Patent Application Publications 2020/0305744 and PCT (Int'l) Publication WO 2020/251899. Still further, the sensing electrode(s) used to sense ECAPs may not be constant with respect to stimulation provided in different patients; the distance between the stimulating and sensing electrodes may vary for example. Adding further variability is the fact that the ECAPs can be detected in different ways. For example, an ENT can be determined by the neural response algorithm 124 using signal averaging. See, e.g., U.S. Pat. No. 10,926,092. An ENT can also be determined “visually”—i.e., as a lowest amplitude at which an ECAP (e.g., its shape) can be visually noticed by a clinician and input into the GUI. (Normally it would be expected that ENT extracted mathematically would be lower than an ENT extracted visually).

[0049] For these reasons, an ENT, although measured objectively, does not comprise an absolute value, but instead has a value that may be system and/or patient dependent. In this regard, and referring again to FIG. 7, an ENT may have an uncertain relationship to other therapeutic thresholds such as the perception threshold pth. For example, ECAPs are present in sub-perception stimulation therapy, and so one could expect ENTs to be of lower values than pth. However, such ENTs could be higher than pth depending on how ENTs are measured by the system, and depending on the specific patient lead implantation location and its distance to the spinal cord. Note that ECAPs comprise only one type neural response to stimulation, and that other types of neural responses exist for which extracted thresholds can be determined. For example, in Deep Brain Stimulation, stimulation of brain tissue can give rise to Evoked Resonant Neural Activity (ERNA) responses, as explained in U.S. patent application Ser. No. 17/388,818, filed Jul. 29, 2021. Neural thresholds of still evoked potentials are possible. Note also that ENTs as described here also comprise a type of physiological threshold in that it comprises a threshold (e.g., of current amplitude) that causes a physiological response (a detectable ECAP response) in the patient.

[0050] Even though ENT values may have some variability, the inventors have noticed based on empirical measurements that ENTs as measured in a given system vary predictably with physiological thresholds such as pth and dth otherwise determined by the system. This is shown in FIG. 9, which shows data taken from different patients. For each patient, an extracted neural threshold (ENT) was measured. Each patient's perception threshold pth (top graph) and discomfort threshold dth (bottom graph) were also determined as described above. As the figures shown, both pth and dth show a significant correlation to ENT. Specifically, using curve fitting (e.g., least squared) techniques, a mathematical relationship 140a between pth and ENT has been determined, and a mathematical relationship 140b between dth and ENT has been determined. Notice that these relationships 140a and 140b do not comprise a mere scalar between ENT as measured and the therapeutic thresholds pth and dth (i.e., pth and dth are not just equal to c*ENT, where c is a constant). In this example, the relationships 140a and 140b are linear, but could be modelled as different functions (e.g., exponential s, logarithms, polynomials, etc.).

[0051] FIG. 10 shows a first algorithm 150 which can be used to measure ENTs, and to automatically estimate or calibrate one or more physiological thresholds such as pth or dth using the relationships 140a or 140b. In the example of FIG. 10, certain steps in the algorithm 150 are performed by an external device, such as a clinician programmer 70 or patient external controller 60. However, this is not strictly necessary, and instead the entirety of algorithm 150 can be performed within the IPG 100 itself (as programmed in its control circuitry 102), particularly if the relationships 140a or 140b are stored in the IPG 100. It is assumed here that algorithm 150 is initiated by the GUI at option 148 (FIG. 148). Note that aspects of algorithm 150 may be embodied in instructions stored in non-transitory computer readable media such as solid state, optical or magnetic memories, which may be included whole or in part in the external device or the IPG 100. Algorithm 150 may comprise part of the external device's software (e.g., 84, FIG. 4), and may be integrated with the software used to render the GUI 82.

[0052] In step 152, an ENT value is measured using stimulation parameters determined earlier for the patient. Specifically, and assuming option 148 is used to start the algorithm 150, the external device causes the IPG 100 to increase the amplitude I starting from zero, until a neural response such as an ECAP is detectable (either using extraction or visualization). The ENT could also be determined by decreasing I until ECAPs are no longer detectable. Because step 152 implicates use of the neural response algorithm 124, and because this algorithm 124 can also operate in part in the external device, step 152 may be performed at least in part in the external device or wholly within the IPG 100. Again, the neural response algorithm 124 can determine the ENT value using extraction or visualization techniques as described earlier.

[0053] In step 154, the determined ENT value is used to determine at least one neural threshold, such as pth (using relationship 140a) or dth (using relationship 140b). These relationships may be stored in the external device in conjunction with other aspects of algorithm 150. One skilled will understand that the physiological thresholds can be determined by entering the ENT value into the relationships 140a and 140b and solving for pth and/or dth. Again, if the relationships 140a and 140b are stored in the IPG 100, this step 154 can also be performed entirely within the IPG 100.

[0054] In step 156, the physiological thresholds pth and/or dth determined in step 156 are stored in the external device. In particular, the algorithm 150 may automatically populated these determined physiological thresholds into the threshold interface 136 of the GUI (FIG. 8).

[0055] At this point, an optional step 158 may be performed to confirm that the determined physiological thresholds are at proper values by testing them on the patient. This is simpler and faster than determining these thresholds as described above using a full amplitude sweep. For example, in a manual mode, the amplitude value is set to the determined pth value (say pth=I=4.8 mA) and tested on the patient, perhaps by manually moving the amplitude up and down a slight amount from this value. From this, a slightly different value for pth may be determined (e.g., pth=4.9 mA or 4.7 mA), and this would occur more quickly because a full range of amplitude values is not tested. This example in effect provides an estimated pth value, which can then be quickly updated based on testing. dth may be similarly tested and confirmed. In a more automated approach, a small range of amplitude values is swept around the determined thresholds (e.g., from 4.5 mA to 5.1 mA), with the patient pressing a button (e.g., at I=4.9 mA) when (in this example) he can first start to feel paresthesia. pth in this example would be calibrated from 4.8 to 4.9 mA.

[0056] At step 160, the physiological thresholds as so determined (and perhaps confirmed at step 158) are transmitted to the patient's external controller 60 or to the patient's IPG 10 directly. As noted above, these thresholds can be put to useful ends in controlling patient stimulation therapy. If algorithm 150 runs exclusively in the IPG 100, such transmission of the determined physiological thresholds would not be necessary.

[0057] FIG. 11 illustrates a manner in which the disclosed technique for determining physiological thresholds using ENT can be extended to include consideration of other stimulation parameters. The left graph in FIG. 11 shows a well-known strength-duration curve, which generally shows a relation between strength (e.g., amplitude I) and the duration (e.g., pulse width PW) necessary to recruit neural tissue and generate an action potential. This curve is typically defined by the rheobase (Irb) which generally comprises the minimal current amplitude that results in depolarization, and a chronaxie time (c) which essentially sets the time constant of the displayed curve. This strength-duration curve has been modeled in the literature using different mathematical equations, and two of them (Weiss-Lapicque, and Lapicque-Blair) are shown. ECAPs are the result of multiple action potentials triggered almost at the same time and added together to form the evoked compound action potential. It stands to reason that recruited neural tissue would cause ECAPs to issue, and therefore that ENTs (also measured in intensity or amplitude) would follow these same curves, as shown in the right graph of FIG. 11. As such, ENTs can be modelled as a function of duration (pulse width), using the same constants (Irh, c) present in the strength-duration equations.

[0058] This suggests to the inventor that it may be beneficial to measure ENTs at more than one pulse width. Doing so allows a mathematical relationship 170 to be determined relating ENT values and pulse widths (i.e., ENT=f(PW)). This is shown in further detail in FIG. 12, where ENT values have been measured for different pulse widths: (PW, ENT)=(50 μs, 5 mA), (100 μs, 3 mA), and (300 μs, 1 mA). Using the Weiss-Lapicque equation and curve fitting techniques, values for the rheobase (Irh=0.2 mA) and the chronaxie time (c=1200 μs) can be determined. In other words, a relationship 170 is determined for the patient which relates ENTs to various pulse widths for their stimulation—i.e., ENT=0.2 (1+1200/PW).

[0059] Establishing relationship 170 is useful in the context of the disclosed technique, because it allows physiological thresholds like pth and dth to be estimated for different pulse widths, and algorithm 180 in FIG. 12 shows an example of the use of such information. In the example of FIG. 12, certain steps in the algorithm 180 are performed by an external device, such as a clinician programmer 70 or patient external controller 60. However, this is not strictly necessary, and instead the entirety of algorithm 180 can be performed within the IPG 100 itself (as programmed in its control circuitry 102), particularly if the relationships 140a, 140b, and 170 are stored in the IPG 100. It is again assumed here that algorithm 180 is initiated by the GUI at option 148 (FIG. 148). Like algorithm 150, algorithm 180 may be embodied in instructions stored in non-transitory computer readable media.

[0060] In step 182, an ENT value is measured using stimulation parameters determined earlier for the patient, but with different pulse widths. Specifically, and assuming option 148 is used to start the algorithm 150, the external device may provide an instruction 200 on the GUI 82 instructing the clinician to provide stimulation at a next (first) pulse width value, as shown in FIG. 8. This first pule width value (e.g., 50 μs) can be entered using the stimulation parameter interface 132. The clinician can then select instruction 200 to measure the ENT for this pulse width. As before, the external device can cause the IPG 100 to increase the amplitude I until an ECAP is detectable (either using extraction or visualization), or can decreasing I until ECAPs are no longer detectable. The user can then enter a next pulse width to test (e.g., 100 μs), and measure the ENT at this pulse width by again selecting instruction 200, etc. This results in building a data table 200 of (PW, ENT) values, as shown in FIG. 8, which data table can be stored in the external device. It should be understood that step 182 essentially involves test stimulation in the sense that such stimulation may involve testing at pulse widths that aren't necessarily therapeutically optimal for the patient.

[0061] Step 184 determines the ENT=f(PW) relationship 170 using the data in data table 200. This step was described earlier with respect to FIG. 11. Once relationship 170 is determined, it can be stored in the external deice in step 186.

[0062] Next is step 188, a therapeutic pulse width to be used for the patient is entered into the GUI (again using stimulation parameters interface 132 for example). In the depicted example, this pulse width value is 200 μs, which is assumed here to be the pulse width that has otherwise been deemed optimal to provide therapeutic stimulation for the patient. While an ENT value could be measured at this optimal pulse width, this is not necessary, because the ENT at PW=200 μs can be estimated using relationship 170 as just determined, which occurs at step 190. In this example, it is assumed using relationship 170 that ENT=1.6 mA at PW=200 μs.

[0063] From this estimated ENT value, and at step 192, one or more physiological thresholds like pth or dth can be estimated using the relationship 140a and 140b described earlier. For example purposes, it is only assumed that a single physiological threshold (pth) is determined at step 192. Plugging ENT=1.6 mA into relationship 140a yields pth=1.50, which as before can be auto-populated in GUI 82 at threshold interface 136 (FIG. 8). This determined value for pth can be optionally confirmed at step 194, similar to what was described earlier at step 158 in FIG. 10, and as before can be transmitted to the patient external controller 60 or 10 to control stimulation to useful effect at step 196.

[0064] If later the pulse width of the patient's stimulation is changed (step 196), the physiological threshold(s) can be automatically adjusted without need to take further ENT measurements, because relationships 170, 140a, and/or 140b can be used to adjust the threshold(s). In this regard, and as shown in FIG. 12, algorithm 180 can revert back to step 190, where a new ENT value is predicted at the new pulse width using relationship 170, and at step 192 a new physiological threshold(s) (e.g., pth) can be determined using the new ENT value using relationship 140a. In short, algorithm 180 is able to estimate physiological thresholds such as pth and dth using both measured ENT values and pulse width. As was the case earlier, aspects of algorithm 180 can be executed both at the external device and the IPG 100.

[0065] To summarize, algorithm 180 (FIG. 12) is more flexible than algorithm 150 (FIG. 10) in that physiological thresholds can be calibrated for a wider range of potential variations in a patient's stimulation program, and in particular when the pule width of the pulses is varied. However, in either case, physiological thresholds are determined automatically, and based on objective ENT measurements, which makes determining these physiological thresholds easier for both the clinician and the patient. It also allows physiological thresholds to be updated without the need for intervention by the clinician. For example, physiological thresholds such as pth and dth can be adjusted on the fly by the patient, or even automatically by the system. In this regard, note that the algorithms 150 can also operate at least in part on a patient's external controller 60. The controller 60 can automatically periodically run either of algorithms 150 or 180 to measure ENTs (possibly at different pulse widths), and to adjust physiological thresholds accordingly, which in turn affects the patient's stimulation. Adjusting dth at the external controller 60 can limit the amplitude I the patient can prescribe, which keeps the patient safe and comfortable. Adjusting pth at the external controller 60 helps to ensure for example that the patient's stimulation will reliably remain sub-perception or supra-perception by constrain the amplitude that the patient can prescribe, etc.

[0066] While described in the context of determining physiological thresholds such as pth and dth, it should be understood that the disclosed techniques may also be used to determine target values for stimulation. In this regard, an optimal stimulation amplitude I for a patient can relate to physiological thresholds such as pth and dth. For example, an optimal stimulation amplitude I may comprise pth, a percentage of pth (e.g., I=70% pth), or a particular value between pth and dth (e.g., a midpoint value such as pth+[dth−pth/2]). Because physiological thresholds pth and dth can be determined using ENTs as described above, and because a desired amplitude threshold I can be based on or predicted using pth and/or dth, ENTs can be used to predict and/or adjust amplitude I.

[0067] Although particular embodiments of the present invention have been shown and described, the above discussion is not intended to limit the present invention to these embodiments. It will be obvious to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the present invention. Thus, the present invention is intended to cover alternatives, modifications, and equivalents that may fall within the spirit and scope of the present invention as defined by the claims.