WO2022081470A1 - Systems and methods for spatial optimization of spinal cord stimulation - Google Patents

Abstract

This present disclosure provides systems and methods relating to neuromodulation. In particular, the present disclosure provides systems and methods for identifying spatially optimized electrical stimulation patterns for reducing pain in a subject. The systems and methods of neuromodulation disclosed herein facilitate the treatment of neuropathic pain associated with various disease states and clinical indications.

Description

SYSTEMS AND METHODS FOR SPATIAL OPTIMIZATION OF SPINAL

CORD STIMULATION

RELATED APPLICATIONS

[0001] This application claims priority to and the benefit of U.S. Provisional Patent Application No. 63/090,798 filed October 13, 2020, which is incorporated herein by reference in its entirety for all purposes.

FIELD

|0002] This present disclosure provides systems and methods relating to neuromodulation. In particular, the present disclosure provides systems and methods for identifying spatially optimized electrical stimulation patterns for reducing pain in a subject. The systems and methods of neuromodulation disclosed herein facilitate the treatment of neuropathic pain associated with various disease states and clinical indications.

BACKGROUND

[0003] Spinal cord stimulation (SCS) is a surgically implanted device therapy to treat chronic pain that applies electrical pulses to a targeted area of the spinal cord. SCS efficacy is dependent on the amplitude, pulse duration, and pulse repetition rate or frequency of the applied pulses, and these parameters are typically set by a device programmer. Furthermore, new electrode lead designs can have as many as 32 individual electrode contacts, and each contact can be programmed to deliver a different frequency, amplitude, or pulse duration. Technological innovation has made possible significant improvements in SCS programming, but the very large parameter space makes it challenging to arrive at optimal or even appropriate parameters to provide maximal pain relief. There is little guidance and few criteria that can be used for programming, apart from feedback from the patient about the effects of each parameter set on their perceived pain at that moment. Further, these is a practical limitation on the number of stimulation parameter combinations that can feasibly be tested during a session. In the temporal domain, programmers typically use a single pulse repetition frequency within a narrow range (40-100 Hz). In the spatial domain, programmers select active electrode contacts to maximize the overlap of the region where pain is felt with the area where stimulation creates a paresthesia, the referred sensation of tingling or buzzing that occurs corresponding to location in the spinal cord where stimulation occurs. However, stimulation in regions beyond the pain area can also activate inhibitory mechanisms in the spinal cord and may be beneficial in improving pain relief.

SUMMARY

[0004] Embodiments of the present disclosure include a method of identifying an optimized electrical stimulation pattern for delivering electrical stimulation to a subject for pain reduction. In accordance with these embodiments, the method includes selecting an electrode geometry and/or electrode position, selecting at least one stimulation pulse parameter, evaluating the electrode geometry and/or the electrode position and the at least one stimulation pulse parameter based on neural activity in at least one target zone using a computational model of a neuronal network, and identifying at least one spatially optimized candidate stimulation pattern capable of reducing pain.

[0005] In some embodiments, the neural activity comprises a reduction in the activity in pain-transmitting neurons.

|0006] In some embodiments, the at least one target zone comprises a receptive field zone. [0007] In some embodiments, the spatially optimized candidate stimulation pattern activates a target zone comprising a center of a receptive field and/or a target zone comprising an area of surround inhibition.

[0008] In some embodiments, the at least one target zone comprises three or more target zones. In some embodiments, the at least one target zone is located in the spinal cord. In some embodiments, the at least one target zone is located in the peripheral nervous system.

|0009] In some embodiments, selecting the electrode geometry comprises selecting or altering which electrode contacts are active. In some embodiments, selecting the electrode position comprises selecting or altering a physical position of the electrode lead in the subject. In some embodiments, selecting at least one stimulation pulse parameter comprises selecting or altering pulse repetition frequency, pulse amplitude, pulse duration, pulse shape, temporal pattern, and any combinations thereof.

[0010] In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 400 Hz. In some embodiments, the pulse amplitude ranges from about 1 mA to about 20 mA. In some embodiments, the pulse width ranges from about 30 ps to about 600 ps. In some embodiments, the temporal pattern comprises a random, stochastic, periodic, and/or bursting pattern of stimulation pulses. [0011] In some embodiments, the computational model of the neuronal network simulates activity of a wide dynamic range (WDR) neuron. In some embodiments, the activity of the WDR neuron in the computational model is a proxy for pain.

[0012] In some embodiments, the computational model of the neural network comprises input responses from computational models of dorsal column axons to the spatially optimized candidate stimulation patterns. In some embodiments, the dorsal column axons input responses are calculated based on diameter and position of the dorsal column axons, wherein the position corresponds to different receptive field zones. In some embodiments, the spatially optimized candidate stimulation patterns activate specific populations of dorsal column axons.

[0013] In some embodiments, the computational model of the neural network comprises three network zones comprising heterogeneous inhibitory and excitatory neural connections. |0014] In some embodiments, the method further comprises programming a neuromodulation device to deliver the at least one spatially optimized candidate stimulation pattern to a subject.

[0015] In some embodiments, the method further comprises delivering the at least one spatially optimized candidate stimulation pattern to the subject to reduce pain. In some embodiments, the at least one spatially optimized candidate stimulation pattern is delivered to an area of the spinal cord. In some embodiments, the at least one spatially optimized candidate stimulation pattern is delivered to an area of the peripheral nervous system.

[0016] Embodiments of the present disclosure also include a method using electrical stimulation to reduce pain in a subject. In accordance with these embodiments, the method includes programming a neuromodulation device to deliver at least one spatially optimized candidate stimulation pattern to a subject, wherein the at least one spatially optimized candidate stimulation pattern is optimized with respect to electrode geometry and/or electrode position and at least one stimulation pulse parameter, and delivering the at least one spatially optimized candidate stimulation pattern to the subject to reduce pain.

[0017] Embodiments of the present disclosure also include a system for using electrical stimulation to reduce pain in a subject. In accordance with these embodiments, the system includes an electrode sized and configured for implantation in proximity to neural tissue, and a pulse generator coupled to the electrode, the pulse generator including a power source comprising a battery and a microprocessor coupled to the battery, wherein the pulse generator is configured to generate electrical signals for delivering a spatially optimized candidate stimulation pattern to the subject to reduce pain. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] FIG. 1: Representative patient pain rating scores (mean+s.e.m., n=5) during different frequencies of SCS.

[0019] FIG. 2: Representative diagrams demonstrating center- surround architecture in peripheral receptive fields, which is preserved in the organization of dorsal column axons and networks of neurons in the dorsal horn. Zone 1 represents the pain area in the peripheral receptive fields. Zones 2 and 3 send inhibitory and excitatory connections to zone 1.

[0020] FIGS. 3A-3B: Representative diagrams of the architecture of distributed biophysically-based network model of afferent signal processing in the dorsal horn and responses to SCS. FIG. 3A (i-iii) includes synaptic connections for each neuron in the model (i), network architecture within the model for a single node (ii), and the distributed multimodal model architecture. FIG. 3B represents default model inputs and outputs in neuropathic pain condition for zone 1 model neurons.

[0021] FIGS. 4A-4C: Design of a computational model for simulating dorsal column axon activity. FIG. 4A is a model based on a mammalian myelinated axon fiber model. FIG. 4B is a model with a double cable structure nodal dynamics. FIG. 4C is a dorsal column fiber model.

[0022] FIGS. 5A-5D: Representative data modeling the response of dorsal column axons to different parameter settings. FIG. 5A includes a representative range of firing rates for axons activated by 50 Hz/300 ps stimulation waveform. FIG. 5B includes representative data of the organization of axons into different receptive field zones (Zl, Z2, and Z3) based on position. FIG. 5C includes a representative frequency-pulse width curve used to sample parameters. FIG. 5D includes data from an example fiber population.

[0023] FIGS. 6A-6I: Representative data of dorsal horn responses to stimulation. FIGS. 6A-6D represent model responses, whereas FIGS. 6E-6I include experimental responses.

[0024] FIGS. 7A-7D: Representative data using the distributed model for optimization with different amplitude/frequency/targeting combinations. FIG. 7A includes representative response of 25 sample populations to changing frequency and amplitude. FIG. 7B includes the same data as FIG. 7A, but response of all neurons were sorted by neuron response amplitude. FIG. 7C includes the percent of the target population activated by amplitude and frequency. FIG. 7D includes the percent of model nerve fibers firing faithfully with the stimulation frequency by amplitude and frequency. [0025] FIGS. 8A-8C: Representative data from experimental recordings of receptive field targeted stimulation. FIG. 8A includes a representative experimental setup for recording responses to receptive field targeted stimulation of different receptive field areas (tibial versus common peroneal nerve). FIG. 8B includes representative data from neurons recorded at different positions from receptive field targeted stimulation of different nerve branches. FIG. 8C shows the mean response of all neurons to receptive field targeted stimulation at 50 Hz.

[0026] FIGS. 9A-9D: Representative data from model responses and experimental recordings of receptive field targeted dual frequency stimulation. FIG. 9A represents targeted dual frequency stimulation in the computational model. FIG. 9B represents average model WDR neuron responses to dual frequency stimulation relative to 50 Hz stimulation of the center zone alone. FIG. 9C includes a representative experimental setup for recording responses to receptive field targeted dual frequency stimulation. FIG. 9D includes representative data demonstrating the difference between stimulation at 50 Hz alone on the tibial nerve and all paired frequencies.

[0027] FIGS. 10A-10E: Representative data obtained using the distributed model for optimization with different frequencies for each zone. FIG. 10A includes data from combinations of unique frequencies to each node based on the zone. FIG. 10B includes optimal zone 1 frequency for each pair of frequencies in zones 2 and zones 3 to project the solution onto two-dimensional frequency space. FIG. 10C includes percent of the baseline response of the model WDR neurons across all three zones. FIG. 10D provides a representation of power consumption for each optimal frequency combination assuming a uniform pulse width. FIG. 10E includes optimized scores combining efficacy (FIG. 10C) and efficiency (FIG. 10D) for each combination of surround frequencies with the optimal zone 1 frequency.

DETAILED DESCRIPTION

[0028] Section headings as used in this section and the entire disclosure herein are merely for organizational purposes and are not intended to be limiting.

1. Definitions

10029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In case of conflict, the present document, including definitions, will control. Preferred methods and materials are described below, although methods and materials similar or equivalent to those described herein can be used in practice or testing of the present disclosure. All publications, patent applications, patents and other references mentioned herein are incorporated by reference in their entirety. The materials, methods, and examples disclosed herein are illustrative only and not intended to be limiting.

[0030] The terms “comprise(s),” “include(s),” “having,” “has,” “can,” “contain(s),” and variants thereof, as used herein, are intended to be open-ended transitional phrases, terms, or words that do not preclude the possibility of additional acts or structures. The singular forms “a,” “and” and “the” include plural references unless the context clearly dictates otherwise. The present disclosure also contemplates other embodiments “comprising,” “consisting of’ and “consisting essentially of,” the embodiments or elements presented herein, whether explicitly set forth or not.

|0031] For the recitation of numeric ranges herein, each intervening number there between with the same degree of precision is explicitly contemplated. For example, for the range of 6- 9, the numbers 7 and 8 are contemplated in addition to 6 and 9, and for the range 6.0-7.0, the number 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, and 7.0 are explicitly contemplated. Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise- indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. For example, if a concentration range is stated as 1% to 50%, it is intended that values such as 2% to 40%, 10% to 30%, or 1% to 3%, etc., are expressly enumerated in this specification. These are only examples of what is specifically intended, and all possible combinations of numerical values between and including the lowest value and the highest value enumerated are to be considered to be expressly stated in this disclosure.

|0032] “Pain” generally refers to the basic bodily sensation induced by a noxious stimulus, received by naked nerve endings, characterized by physical discomfort (e.g., pricking, throbbing, aching, etc.) and typically leading to an evasive action by the individual. As used herein, the term pain also includes chronic and acute neuropathic pain. The term “chronic neuropathic pain” refers to a complex, chronic pain state that is usually accompanied by tissue injury wherein the nerve fibers themselves may be damaged, dysfunctional or injured. These damaged nerve fibers send incorrect signals to other pain centers. The impact of nerve fiber injury includes a change in nerve function both at the site of injury and areas around the injury. The term “acute neuropathic pain” refers to self-limiting pain that serves a protective biological function by acting as a warning of on-going tissue damage. Acute neuropathic pain is typically a symptom of a disease process experienced in or around the injured or diseased tissue.

[0033] “Subject” and “patient” as used herein interchangeably refers to any vertebrate, including, but not limited to, a mammal (e.g., cow, pig, camel, llama, horse, goat, rabbit, sheep, hamsters, guinea pig, cat, dog, rat, and mouse, a non-human primate (e.g., a monkey, such as a cynomolgus or rhesus monkey, chimpanzee, etc.) and a human). In some embodiments, the subject may be a human or a non-human. In one embodiment, the subject is a human. The subject or patient may be undergoing various forms of treatment.

[0034] “Treat,” “treating” or “treatment” are each used interchangeably herein to describe reversing, alleviating, or inhibiting the progress of a disease and/or injury, or one or more symptoms of such disease, to which such term applies. Depending on the condition of the subject, the term also refers to preventing a disease, and includes preventing the onset of a disease, or preventing the symptoms associated with a disease. A treatment may be either performed in an acute or chronic way. The term also refers to reducing the severity of a disease or symptoms associated with such disease prior to affliction with the disease. Such prevention or reduction of the severity of a disease prior to affliction refers to administration of a treatment to a subject that is not at the time of administration afflicted with the disease. “Preventing” also refers to preventing the recurrence of a disease or of one or more symptoms associated with such disease.

[0035] “Therapy” and/or “therapy regimen” generally refer to the clinical intervention made in response to a disease, disorder or physiological condition manifested by a patient or to which a patient may be susceptible. The aim of treatment includes the alleviation or prevention of symptoms, slowing or stopping the progression or worsening of a disease, disorder, or condition and/or the remission of the disease, disorder or condition. In some embodiments, the treatment comprises the treatment, alleviation, and/or lessening of pain.

[0036] Unless otherwise defined herein, scientific and technical terms used in connection with the present disclosure shall have the meanings that are commonly understood by those of ordinary skill in the art. For example, any nomenclatures used in connection with, and techniques of, cell and tissue culture, molecular biology, neurobiology, microbiology, genetics, electrical stimulation, neural stimulation, neural modulation, and neural prosthesis described herein are those that are well known and commonly used in the art. The meaning and scope of the terms should be clear; in the event, however of any latent ambiguity, definitions provided herein take precedent over any dictionary or extrinsic definition. Further, unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular.

2. Spatially Optimized Electrical Stimulation Patterns

[0037] Classic experiments on neural circuits in the spinal cord showed that large diameter afferent (sensory) fibers that respond to touch (A[3 afferent fibers) activate inhibitory mechanisms in the dorsal horn. Smaller A5 and C afferent fibers respond to pain and excite dorsal horn projection neurons. However, beyond these simplistic “labeled lines” there are also heterogeneous excitatory and inhibitory connections between different regions of the peripheral receptive fields determining the sensitivity of afferents. Given the complexity of dorsal horn pain circuits, the myriad stimulation parameters in the spatial and temporal domains, and the heterogeneity of peripheral receptive fields it is highly unlikely that the present ad hoc empirical testing of a limited number of stimulation parameter combinations identifies optimal or even suitable parameters to reduce maximally pain for patients. Computational modeling enables the testing parameters across a much wider range than is practical in a clinical setting.

[0038] Thus, embodiments of the present disclosure include methods for evaluating different neuromodulation parameter settings combined with targeting specific fiber populations. In accordance with these methods, computational models of dorsal column axons and dorsal horn neurons were developed and validated as a novel approach to test various electrical parameter combinations to identify combinations that reduce pain in a subject.

[0039] As described further herein, a computational model was developed to test the effects of spinal cord stimulation (SCS) on spinal cord pain networks. The model comprises wide-dynamic range (WDR) neurons that transmit pain signals to the brain as well as excitatory (EX) and inhibitory (IN) interneurons. The activity of model WDR neurons is a validated proxy for the level of pain; there is a significant link between firing rate of WDR neurons and pain ratings, and changes in WDR firing rates during SCS parallel behavioral effects on pain. Furthermore, the firing rate of WDR neurons, or level of pain, is dependent on the pulse repetition frequency of SCS (FIG. 1). Previous computational work and preclinical animal studies demonstrated reduced WDR firing rates from two simultaneously applied frequencies. However, this model did not incorporate or account for the distinct and heterogeneous receptive field connections within each of the three receptive field zones, and these connections are critical for determining the effects of SCS. [0040] In contrast, the computational models of the present disclosure, take into account that the distinct responses to inputs from different receptive field zones would enable the engagement of separate neural mechanisms in the dorsal horn of the spinal cord to reduce activity in pain transmitting WDR neurons. Further, the computational models of the present disclosure recognized that the delivery of unique and appropriate stimulation frequencies to each zone suppresses activity in pain-transmitting WDR neurons. In accordance with this, embodiments of the present disclosure are based on a novel approach to exploit this recognition by identifying the appropriate stimulation frequencies to deliver to each zone. This innovative approach combines selection of stimulation electrode geometries to activate neurons in each of the receptive field zones and selection of stimulation pulse parameters, (e.g., pulse repetition frequency) to deliver effective stimulation in each of the receptive field zones to reduce pain.

|0041] Receptive fields in the somatosensory system exhibit center- surround architecture (FIG. 2). Sensory inputs in the center of a receptive field will elicit excitation from the corresponding neurons in the spinal cord while inhibiting responses in the surrounding receptive fields. The novel computational model of the present disclosure (e.g., the distributed model) accounted for center-surround architecture through different nodes representing different receptive field areas (FIG. 3). In the model, for example, zone 1 represents the dorsal horn circuit that receives excitatory afferent inputs from the center peripheral receptive area. Zone 1 center projection neurons will be excited by either a touch or a bee sting in the corresponding peripheral receptive field. However, pain inputs from A5 and C fibers may target immediately surrounding nodes as well as the center node. This architecture leads to two separate zones in the surround receptive field, zones 2 and 3. Both zones 2 and 3 are inhibited by zone 1 after excitation of A(3 fibers, but zone 2 is subsequently excited by excitation of A5 and C fibers. For example, in zone 2, rubbing the area surrounding a bee sting will inhibit pain, but pinching the zone 2 area can amplify the original pain response. This distributed model was validated by comparing it with experimental responses to dorsal column and peripheral electrical stimulation in different zones (FIG. 4). The timing and magnitude of the model responses matched experimental responses.

[0042] Previous research that used model-based analysis of spinal cord stimulation inputs represented the effects of SCS on dorsal column axons by defined spike times arriving at the dorsal horn neural network. However, the dorsal column axons may not consistently follow the applied stimulation pulses. That is, the axons may not generate an action potential in response to every stimulation pulse; therefore, the action potential firing rate and pattern may differ from the applied stimulation pulse rate and pattern. In the present disclosure, the response of dorsal column axons to different stimulation parameters was calculated, including amplitudes and repetition rates of applied stimulation pulses, using a validated computational model of dorsal column axons. The resulting action potential rates and patterns were then used as inputs to the dorsal horn neural networks. Computational models of dorsal column axons were used, coupled with an anatomical finite element model of the spinal cord to quantify the effect of different combinations of pulse width, frequency, and amplitude. The dorsal column axon model is based on a previous myelinated axon model with some modifications to represent the responses of dorsal column axons to different parameters of SCS.

[0043] Experiments were conducted to examine how model dorsal column nerve fibers with a range of diameters found in dorsal columns (e.g., 2.2 pm - 6.5 pm for rat dorsal columns) responded to the electric fields generated by an electrode placed above the dorsal columns. The resulting rate and pattern of action potentials were calculated in each of the model nerve fibers in a population of model nerve fibers with different diameters and positions within the dorsal columns for applied spinal cord stimulation parameters with different pulse repetition frequencies, pulse durations, and pulse amplitudes (FIG. 5A). In this model, 100 pA represents the motor threshold. Maps of dorsal column axons in the lumbar spinal cord indicate that they exhibit somatotopic arrangements based on the nerve root of origin. Dorsal column nerve fiber positions corresponding to different receptive field areas were represented at the T13 vertebral level, but this is not meant to be limiting, as the relative location of fibers from different receptive field areas should be similar for different vertebral levels (FIG. 5B). Action potential firing of model nerve fibers was quantified in response to ranges of pulse repetition frequencies and pulse duration combinations (FIG. 5C).

|0044] Several representations of populations of dorsal column axons were generated by randomly seeding 60 model nerve fibers with diameters sampled from a normal distribution (p = 4.4, o = l) (FIG. 5D). The response of dorsal horn neurons was then calculated using the distributed model described herein because it accurately represents different receptive field areas (FIG. 2) to the rate and pattern of action potential inputs arriving from each of the model dorsal column nerve fibers in each of the populations in response to specific SCS parameters.

|0045] As described herein, it was recognized that the delivery of unique and appropriate stimulation frequencies to each receptive field zone (e.g., represented by distinct groups of model dorsal column nerve fibers) would enable maximal suppression of activity in paintransmitting WDR neurons. The novel approach provided in the present disclosure was developed in order to exploit this recognition by identifying the appropriate stimulation frequencies to deliver to each zone. This requires appropriate electrode geometry and position to activate targeted groups of dorsal column nerve fibers within each zone. Thus, this approach combines selection of stimulation electrode geometries and positions to activate neurons in each of the receptive field zones and selection of stimulation pulse parameters (e.g., pulse repetition frequency) to deliver in each of the receptive field zones. Results of the present disclosure demonstrate that the proper titration of amplitude, frequency, and pulse width led to maximum suppression of the WDR neuron.

[0046] A spatial shift in the electrode position, or equivalently, an alteration in the electrode geometry (e.g., the combination of active contacts) was modeled by altering the relative activation of model dorsal column nerve fibers representing each of the three zones. In this example, the center model nerve fibers were shifted to represent surround, and the surround model nerve fibers were shifted to represent zone 1 (FIG. 6A). In a clinical setting, this is synonymous with a shift in the current directed to different electrode contacts on the multiple contact electrode lead or a physical repositioning of the electrode lead. This selection of electrode geometry and position increased the suppression of activity in pain transmitting model WDR neurons (e.g., increased therapeutic efficacy across multiple frequencies). The greatest suppression of model WDR neuron activity (e.g., greatest therapeutic effect) occurred at the lowest amplitude that activated dorsal column fibers (e.g., threshold current, 30-40 pA; FIGS. 6B-6C). Model neuron responses were for 90 Hz/225 ps stimulation.

[0047] The computational modeling results were validated by recording responses of rat dorsal horn neurons to 90 Hz/225 AP-electrical stimulation (A[3-ES). Electrodes with 16- or 32- contacts were used to record multiple single units within the dorsal horn. Putatively excitatory (pEX) neurons were identified based on their waveform shape. Motor threshold was identified, and stimulation was performed at amplitudes normalized to the motor threshold (FIG. 6D). Results demonstrated that responses of pEX neurons to mechanical and electrical stimulation was maximized based on the recording location where the electrode was placed (FIG. 6E). Neurons recorded at the L3 spinal entry point were the most excited by peripheral mechanical stimulation, but these neurons were also the most inhibited by electrical stimulation, and maximally inhibited by stimulation at 40% MT (FIGS. 6G-6H). Neurons recorded at the L4 spinal level were both inhibited and excited by AJ3-ES and neurons recorded at the L5 spinal root were predominantly excited by AJ3-ES. These responses are consistent with the spatial distribution of inhibition observed in the model. However, maximum inhibition from A -ES also overlapped with maximum excitation from peripheral crush inputs, indicating that spatial targeting is an important factor in maximizing the efficacy of stimulation (FIG. 61).

[0048] Experiments were also conducted to investigate the role of stimulation pulse width and frequency on model WDR responses. Different stimulation parameters were tested while preserving the same neural dose by adjusting the pulse width with stimulation frequency according to a prior clinical study (see, Paz-Solis et al. 2021). Results demonstrated a nonmonotonic relationship between efficacy and stimulation frequency/pulse width (FIGS. 7A- 7B). The most effective stimulation parameters were quantified for each of the 25 different cases and found that stimulation frequencies between 50 and 90 Hz were most effective. Results demonstrated that increasing stimulation frequency decreased the percent of the target population that was activated (FIG. 7C), which is consistent with a reported decrease in ECAP amplitude as stimulation frequency increased in a prior clinical study. At threshold current, the model dorsal column nerve fibers consistently followed the stimulation frequency approximately 50% of the time for pulse repetition frequencies between 30 - 90 Hz (FIG. 7D). This demonstrates the importance of including validated model nerve fibers to calculate the response of dorsal column axons to different amplitudes and frequencies of SCS, and thus derive the inputs to the dorsal horn networks from the action potential firing times, rather than if the action potential firing times occur in response to every stimulation pulse.

[0049] This analysis was extended to also examine the effect of stimulation location. Specifically, stimulation was delivered either on the tibial nerve or the common peroneal nerve at 60% motor threshold and recorded the responses of dorsal horn pEX neurons (FIG. 8A). Similar to previous findings, these results demonstrated a significant inhibitory effect on neurons in the E3 spinal level from stimulation in the peroneal nerve (FIGS. 8B-8C). However, these results also indicated that net inhibition extended to the E4 spinal level with tibial nerve stimulation, perhaps because the extent of the tibial nerve is larger and more caudal than the peroneal nerve. These results further demonstrated the importance of spatial targeting for stimulation and how broadening the stimulation activation site can lead to broader inhibition with correct parameter settings.

[0050] Given the importance of targeting stimulation and of stimulation parameters, as demonstrated herein, experiments were conducted to investigate how pairs of stimulation frequencies affect model WDR neuron and recorded pEX neuron outputs. In this model, pairs of frequency inputs to zone 1 and zone 2 were tested and the response was compared to 50 Hz alone in zone 1 (FIG. 9A). The effect of paired frequencies within the default model state was tested (e.g., the healthy condition) and after doubling the number of pain fibers (A5 and C) that were active in surround zones. In the model, results demonstrated that high frequency stimulation in zone 2 supplemented the therapeutic effect of 50 Hz alone in the default model state (FIG. 9B). In the pain state, combinations of lower center frequencies and higher surround frequencies tended to be more effective than 50 Hz center stimulation alone. To validate the computational modeling results, the tibial and peroneal branches of the sciatic nerve were stimulated, and quantified pEX neuron responses to paired frequency stimulation were compared to 50 Hz stimulation of the tibial branch alone (FIG. 9C). These recordings were limited to the L4 and L3 spinal roots, the areas where inhibition was previously observed from single frequency stimulation (FIG. 8). Dual frequency stimulation generally increased pEX neuron activity more than 50 Hz alone for neurons recorded in the L4 spinal root, perhaps because activation of the peroneal nerve did not have a large effect on this area. However, neurons were more inhibited by pairs of frequencies than 50 Hz alone with surround frequencies between 30-90 Hz. Neuropathic pain was stimulated by adding bicuculline, a GABAA receptor antagonist, or VU 0463271, a KCC2 inhibitor. Both drugs increased mechanical allodynia following application. Like the model response, results demonstrated that combinations of center frequencies with high surround frequencies (50 - 150 Hz) were more effective than 50 Hz alone. These results demonstrate the potential benefits of multifrequency stimulation in combination with spatial targeting for neuropathic pain.

[0051] Additionally, based on the efficacy of dual frequency stimulation, these results were extended to examine the effects of three frequency combinations in the model. Possible combinations of SCS pulse repetition frequencies were tested between 0 and 150 Hz in increments of 5 Hz (FIG. 10A) delivered to each of the three different receptive field zones. Frequency combinations with high optimized scores were identified; these combinations specified the appropriate stimulation pulse repetition frequency to deliver to each of the three receptive field zones to achieve optimal suppression of model WDR neuron activity (e.g., pain relief). The optimized score for each frequency combination was determined using a fitness function that can incorporate various aspects of the model response. First, we included the pain score as one element of the fitness function in the model (where the pain score is related to a change in firing rate and or pattern of one or more neurons in the model relative to the baseline firing rate). Second, we included the efficiency of the pattern (where the efficiency is proportional to the average frequency of stimulation and is important for evaluating the impact on battery life or recharge interval of implantable pulse generations). Third, we included the network correlation in the model (where the network correlation is calculated as the sum of individual correlations of filtered instantaneous firing rates of pairs of neurons in the model). Fourth, we included the relative frequency pain score (where we calculated the effect of combinations of stimulation frequencies on the pain score compared to single frequency stimulation alone). Each term in the fitness function of the optimization score has a weighting coefficient so we can choose how much each of the factors will influence optimized scores. For example, frequency combinations were identified with high optimized scores by quantifying the firing rate of model WDR neurons across all three zones to measure efficacy (FIG. 10C), and SCS pulse repetition frequency was used as a measure of power consumption (FIG. 10D). In general, higher frequencies appeared better at suppressing WDR firing but these frequencies also had lower efficiency.

[0052] As provided herein, appropriately selected combinations of different frequencies delivered to each of the different receptive field zones can result in greater suppression of model WDR neuron activity and have a positive therapeutic effect. This example also demonstrates that reducing stimulation amplitude may have a similar effect as combining stimulation frequencies. Furthermore, appropriate combinations of pulse repetition frequency and stimulus pulse duration, selected using the described model-based approach and directed to reduce the firing rate of model WDR neurons, can be effective at reducing WDR firing rates (e.g., producing therapeutic reductions in pain) at low stimulation pulse amplitudes (40 pA in this example) that activate only a very small proportion of dorsal column axons (between 10 and 25% of the target population in this example). The use of low stimulation pulse amplitudes is beneficial as it reduces the overall energy required for stimulation, thereby increasing battery life, or reducing requirements to recharge batteries. Eow stimulation amplitudes also reduce side effects associated with stimulation, such as the sensation of paresthesia, which may be bothersome to some patients.

[0053] In accordance with the embodiments of the present disclosure, the selection of appropriate electrode position, electrode geometry, stimulation pulse amplitude, and stimulation pulse duration, which can collectively determine the population of dorsal column nerve fiber activated in each zone, combined with the selection of the appropriate pulse repetition frequency in each receptive field zone, with both selections being enabled by the use of the models and methods described herein, enabled maximum suppression of activity in pain transmitting WDR neurons and thereby produced maximal therapeutic reduction of pain (FIG. 10D).

3. Methods and Treatment

[0054] In accordance with the above description, embodiments of the present disclosure include methods of identifying an optimized electrical stimulation pattern for delivering electrical stimulation to a subject for pain reduction. In some embodiments, the method includes selecting an electrode geometry and/or electrode position, and selecting at least one stimulation pulse parameter, and delivering electrical stimulation to a subject based on these parameters. In some embodiments, the method includes evaluating the electrode geometry and/or the electrode position and the stimulation pulse parameter based on neural activity in at least one target zone using a computational model of a neuronal network (e.g., computational models of dorsal column axons and dorsal horn neurons). In some embodiments, the method includes identifying at least one spatially optimized candidate stimulation pattern that is capable of reducing pain based on the evaluation performed using the computation model.

[0055] In some embodiments, neural activity in target zone(s) includes a reduction in the activity of in pain-transmitting neurons. As described further herein, the target zone(s) can include a receptive field zone in a somatosensory system that exhibits center-surround architecture (FIG. 2). In some embodiments, the spatially optimized candidate stimulation pattern activates neural activity in a target zone that is in the center of a receptive field. In some embodiments, the spatially optimized candidate stimulation pattern activates neural activity in a target zone that is in an area of surround inhibition. In some embodiments, at least one target zone can include multiple different target zones. In some embodiments, the target zone comprises two or more target zones. In some embodiments, the target zone comprises three or more target zones. In some embodiments, the target zone comprises four or more target zones. In some embodiments, the target zone comprises five or more target zones. In some embodiments, the target zone is located in the spinal cord. In other embodiments, the target zone is located in the peripheral nervous system.

[0056] In some embodiments, the method includes adjusting and/or selecting one or more electrode parameters used to deliver a spatially optimized electrical stimulation pattern for reducing pain. In some embodiments, the electrode parameters can include electrode geometry, and selecting an electrode geometry includes selecting or altering which electrode contacts are active in a given electrode lead or array (e.g., selecting which electrodes receive current and which do not with respect to neural activation). In some embodiments, the electrode parameters can include the physical position of the electrode in a subject. In some embodiments, selecting the electrode position comprises selecting or altering a physical position of the electrode lead in the subject (e.g., FIG. 6A-6I), such that different neurons within a target zone are activated.

[0057] In some embodiments, the method also includes adjusting or selecting a pulse stimulation parameter used to deliver a spatially optimized electrical stimulation pattern for reducing pain. In some embodiments, the pulse stimulation parameter is selected or adjusted independently of selecting or adjusting electrode position and/or geometry. In other embodiments, the pulse stimulation parameter and the electrode position and/or geometry are selected or adjusted together as part of evaluating neural activity in a target zone(s). In some embodiments, selecting at least one stimulation pulse parameter includes selecting or altering one or more of pulse repetition frequency, pulse amplitude, pulse duration, pulse shape, temporal pattern, and any combinations thereof. As would be recognized by one of skill in the art based on the present disclosure, other pulse parameters may also be selected or adjusted based on various factors that include, but not limited to, the target zone, the type of electrical stimulation being delivered, the neuromodulation system being used, and the like.

[0058] In some embodiments, the stimulation pulse parameter that is being selected or adjusted as part of identifying a therapeutic spatially optimized candidate stimulation pattern is pulse repetition frequency. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 350 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 300 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 250 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 200 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 150 Hz. In some embodiments, the average pulse repetition frequency ranges from about 10 Hz to about 100 Hz. In some embodiments, the average pulse repetition frequency ranges from about 50 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 100 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 150 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 200 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 250 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 300 Hz to about 400 Hz. In some embodiments, the average pulse repetition frequency ranges from about 50 Hz to about 350 Hz. In some embodiments, the average pulse repetition frequency ranges from about 100 Hz to about 300 Hz. In some embodiments, the average pulse repetition frequency ranges from about 100 Hz to about 200 Hz. In some embodiments, the average pulse repetition frequency ranges from about 200 Hz to about 300 Hz.

[0059] In some embodiments, the stimulation pulse parameter that is being selected or adjusted as part of identifying a therapeutic spatially optimized candidate stimulation pattern is pulse amplitude. In some embodiments, the pulse amplitude ranges from about 1 mA to about 20 mA. In some embodiments, the pulse amplitude ranges from about 5 mA to about 20 mA. In some embodiments, the pulse amplitude ranges from about 10 mA to about 20 mA. In some embodiments, the pulse amplitude ranges from about 15 mA to about 20 mA. In some embodiments, the pulse amplitude ranges from about 1 mA to about 15 mA. In some embodiments, the pulse amplitude ranges from about 1 mA to about 10 mA. In some embodiments, the pulse amplitude ranges from about 1 mA to about 5 mA. In some embodiments, the pulse amplitude ranges from about 5 mA to about 15 mA. In some embodiments, the pulse amplitude ranges from about 5 mA to about 10 mA. In some embodiments, the pulse amplitude ranges from about 10 mA to about 15 mA.

[0060] In some embodiments, the stimulation pulse parameter that is being selected or adjusted as part of identifying a therapeutic spatially optimized candidate stimulation pattern is pulse width. In some embodiments, the pulse width ranges from about 30 ps to about 600 ps. In some embodiments, the pulse width ranges from about 30 ps to about 550 ps. In some embodiments, the pulse width ranges from about 30 ps to about 500 ps. In some embodiments, the pulse width ranges from about 30 ps to about 450 ps. In some embodiments, the pulse width ranges from about 30 ps to about 400 ps. In some embodiments, the pulse width ranges from about 30 ps to about 350 ps. In some embodiments, the pulse width ranges from about 30 ps to about 300 ps. In some embodiments, the pulse width ranges from about 30 ps to about 250 ps. In some embodiments, the pulse width ranges from about 30 ps to about 200 ps. In some embodiments, the pulse width ranges from about 30 ps to about 150 ps. In some embodiments, the pulse width ranges from about 30 ps to about 100 ps. In some embodiments, the pulse width ranges from about 50 ps to about 600 ps. In some embodiments, the pulse width ranges from about 100 ps to about 600 ps. In some embodiments, the pulse width ranges from about 150 ps to about 600 ps. In some embodiments, the pulse width ranges from about 200 ps to about 600 ps. In some embodiments, the pulse width ranges from about 250 ps to about 600 ps. In some embodiments, the pulse width ranges from about 300 ps to about 600 ps. In some embodiments, the pulse width ranges from about 350 ps to about 600 ps. In some embodiments, the pulse width ranges from about 400 ps to about 600 ps. In some embodiments, the pulse width ranges from about 450 ps to about 600 ps. In some embodiments, the pulse width ranges from about 500 ps to about 600 ps. In some embodiments, the pulse width ranges from about 550 ps to about 600 ps. In some embodiments, the pulse width ranges from about 50 ps to about 500 ps. In some embodiments, the pulse width ranges from about 100 ps to about 400 ps. In some embodiments, the pulse width ranges from about 200 ps to about 300 ps. In some embodiments, the pulse width ranges from about 100 ps to about 300 ps. In some embodiments, the pulse width ranges from about 200 ps to about 400 ps. In some embodiments, the pulse width ranges from about 300 ps to about 500 ps. In some embodiments, the pulse width ranges from about 400 ps to about 600 ps.

[0061] In some embodiments, the stimulation pulse parameter that is being selected or adjusted as part of identifying a therapeutic spatially optimized candidate stimulation pattern is temporal pattern. In some embodiments, the temporal pattern comprises a random, stochastic, periodic, and/or bursting pattern, or any combinations thereof. Notwithstanding these embodiments, the methods and systems for delivering temporal patterns of electrical stimulation to a subject in need thereof described in U.S. Patent Appln. Serial Nos. 14/774,156, 14/774,160, and 15/806,686 are herein incorporated by reference in their entireties and for all purposes.

[0062] In some embodiments, the computational model of the neuronal network that is used to evaluate and identify a spatially optimized candidate stimulation pattern capable of reducing pain is based on the activity of a wide dynamic range (WDR) neuron. In some embodiments, the activity of the WDR neuron in the computational model is a proxy for pain, as described further herein. In accordance with these embodiments, the computational model of the neural network comprises input responses from computational models of dorsal column axons to the spatially optimized candidate stimulation patterns. In some embodiments, the dorsal column axons input responses are calculated based on diameter and position of the dorsal column axons. In some embodiments, the position of the dorsal column axons corresponds to different receptive field zones. In some embodiments, the spatially optimized candidate stimulation patterns activate specific populations of dorsal column axons. [0063] In some embodiments, the computational model of the neural network comprises a network zone(s) that includes heterogeneous inhibitory and excitatory neural connections. In some embodiments, the computational model of the neural network comprises two or more network zones that include heterogeneous inhibitory and excitatory neural connections. In some embodiments, the computational model of the neural network comprises three or more network zones comprising heterogeneous inhibitory and excitatory neural connections.

[0064] In some embodiments, the method further includes programming a neuromodulation device to deliver a spatially optimized candidate stimulation pattern to a subject. In some embodiments, the method further includes delivering a spatially optimized candidate stimulation pattern to the subject to reduce pain. In some embodiments, the at least one spatially optimized candidate stimulation pattern is delivered to an area of the spinal cord. In other embodiments, the at least one spatially optimized candidate stimulation pattern is delivered to an area of the peripheral nervous system.

4. Neuromodulation Systems

[0065] Embodiments of the present disclosure also include methods and systems for delivering electrical stimulation to a subject to reduce pain. In accordance with these embodiments, the method includes programming a neuromodulation device to deliver at least one spatially optimized candidate stimulation pattern to a subject. In some embodiments, the spatially optimized candidate stimulation pattern is optimized with respect to electrode geometry and/or electrode position and at least one stimulation pulse parameter, as described above. The method also includes using a neuromodulation device to deliver the spatially optimized candidate stimulation pattern to the subject to reduce pain.

[0066] In accordance with these methods, the present disclosure also provides systems for delivering spatially optimized electrical stimulation to a subject to reduce pain. In accordance with these embodiments, the system includes an electrode sized and configured for implantation in proximity to neural tissue, and a pulse generator coupled to the electrode. In some embodiments, the pulse generator includes a power source with a battery and a microprocessor coupled to the battery, and the pulse generator is configured to generate electrical signals for delivering a spatially optimized candidate stimulation pattern to the subject to reduce pain.

[0067] For example, the system can include a neuromodulation device (e.g., SCS device), an electrical connection lead, and at least one electrode or electrode array operatively positioned to stimulate target neural tissue in a subject that is experiencing neuropathic pain. The electrode or electrode array can be positioned at the site of nerves that are the target of stimulation (e.g., along the spinal cord), or positioned in any suitable location that allows for the delivery of electrical stimulation to the targeted neural tissue.

[0068] In some embodiments, the system includes a pulse generator coupled to the electrode. The pulse generator can include a power source comprising a battery and a microprocessor coupled to the battery, and the pulse generator is generally configured to generate electrical signals for delivering a spatially optimized electrical stimulation pattern (e.g., optimized electrode position/geometry and one or more stimulation pulse parameters), as described further herein. In some embodiments, the system further includes a controller comprising hardware, software, firmware, or combinations thereof for implementing functionality described herein. For example, the controller can be implemented by one or more processors and memory. The controller can be operatively connected to the pulse generator to facilitate the generation of electrical signals for delivering the spatially optimized electrical stimulation pattern to targeted neurological tissue in a subject. The output signals may be received by the connection lead and carried to the electrode or electrode array for the delivery of electrical stimulation to targeted neurological tissue. The system can include a power source, such as a battery, for supplying power to the controller and the pulse generator.

[0069] In some embodiments, the system also includes an external computing device that is not implanted within the subject. The computing device can communicate with a neuromodulation device (e.g., SCS device) or system via any suitable communication link (e.g., a wired, wireless, or optical communication link). The communication link may also facility battery recharge. A clinician may interact with a user interface of the computing device for programming the output of the implanted pulse generator, including the electrodes that are active, the stimulation pulse amplitude, the stimulation pulse duration, the stimulation pattern (including pulse repetition frequency), and the like, applied via each electrode contact to each sub-population. In accordance with these embodiments, systems and methods of the present disclosure can be used to deliver spatially optimized electrical stimulation patterns, as described herein, to reduce pain in a subject.

[0070] In some embodiments, systems and methods of the present disclosure can be implemented as an algorithm within a pulse generator device. An on-board controller can deliver multiple frequencies and patterns of electrical stimulation through different output channels to different contacts on the spinal cord stimulation electrode. By virtue of stimulation through multiple contacts, different populations of axons (e.g., sub-populations of dorsal column nerve fibers) traversing the dorsal column may be activated at different frequencies and in different patterns, resulting in greater suppression of the neurons responsible for transmitting nociceptive information to the brain. Values of the stimulation frequencies and patterns of stimulation and the electrodes through which these frequencies and patterns are delivered can be input by either a physician or a patient through a user interface. Alternatively, the device can be pre-programmed with specific combinations of frequencies and spatially optimized patterns to use. The applied frequencies and patterns may or may not be offset from each other at the start of stimulation.

[0071] In some embodiments, computer readable program instructions for carrying out operations of the present disclosure, including programming the pulse generator to output a spatially optimized electrical stimulation pattern, can be assembler instructions, instruction- set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like, and conventional procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions may execute entirely on the user’s computer, partly on the user’s computer, as a stand-alone software package, partly on the user’s computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user’s computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

5. Examples

[0072] It will be readily apparent to those skilled in the art that other suitable modifications and adaptations of the methods of the present disclosure described herein are readily applicable and appreciable, and may be made using suitable equivalents without departing from the scope of the present disclosure or the aspects and embodiments disclosed herein. Having now described the present disclosure in detail, the same will be more clearly understood by reference to the following examples, which are merely intended only to illustrate some aspects and embodiments of the disclosure, and should not be viewed as limiting to the scope of the disclosure. The disclosures of all journal references, U.S. patents, and publications referred to herein are hereby incorporated by reference in their entireties.

[0073] The present disclosure has multiple aspects, illustrated by the following nonlimiting examples.

Example 1

[0074] FIGS. 3A-3B include representative diagrams of the architecture of distributed biophysically-based network model of afferent signal processing in the dorsal horn and responses to SCS. Synaptic connections for each neuron in the model are shown in FIG. 3A(i). Network architecture within the model for a single node is demonstrated in FIG. 3A(ii). In this example, primary afferents (Ap, A5, and C) transmit inputs to local inhibitory (IN) and excitatory (EX) interneurons and a wide-dynamic range (WDR) projections neuron. SCS is represented as an input to the dorsal columns that propagates antidromically along the A fibers. The output of the node is the firing rate of the WDR neuron (modified from Zhang et al., 2014b). FIG. 3 A(iii) includes a distributed multinodal model architecture. Each circle represents a node of the model from B with unique inputs. The center node (Zone 1) receives excitatory inputs from zone 2 nodes and inhibitory inputs from both zone 2 and zone 3 nodes. All connections between nodes are from interneurons to WDR neurons. The model is extended to eliminate edge effects for the center node.

[0075] FIG. 3B includes a representative diagram of default model inputs and outputs in neuropathic pain condition for zone 1 model neurons. The top raster shows a 50 Hz SCS input along the Ap fibers, but the SCS input is not actually input to the IN neuron for this simulation. The voltage traces of the output neurons are shown on the right. Scale bar, 100 ms and 30 mV.

Example 2

[0076] FIGS. 4A-4C include a representative design of a computational model for simulating dorsal column axon activity. FIG. 4A provides a model based on a prior mammalian myelinated axon fiber model, using the distributed model for optimization with different frequencies for each zone. FIG. 4B provides a model that uses a double cable structure with nodal dynamics governed by sodium channels (Naf and Nap), slow potassium (Ks), and linear leakage (Lk). FIG. 4C provides a dorsal column fiber model, which includes several modifications to the MRG model. This modified model can represent activity commonly seen in dorsal column axons. Blue dashes represent responses to stimulation and red dashes represent additional spikes in the model.

Example 3

[0077] FIGS. 5A-5D include representative data modeling the response of dorsal column axons to different parameter settings. FIG. 5A provides a representative range of firing rates for axons activated by 50 Hz/300 ps stimulation waveform. Concentric rings represent different diameter axons in each position. FIG. 5B provides representative data of the organization of axons into different receptive field zones (Zl, Z2, and Z3) based on position. FIG. 5C includes a representative frequency-pulse width curve used to sample parameters. And FIG. 5D includes data from an example fiber population. The top row of FIG. 5D demonstrates the response of axons to different amplitudes of stimulation, with size representing diameter and color representing firing rate. In the bottom row, color is representative of the receptive field zone.

Example 4

|0078] FIGS. 6A-6I include representative data from dorsal horn responses to stimulation. FIGS. 6A-6D are model responses, while FIGS. 6E-6I are experimental responses. More specifically, FIG. 6A demonstrates that spatial targeting in the model was simulated by altering the target population that was activated at each amplitude. With center targeting (brown), axons from the center of the peripheral receptive field were positioned in the most medial and dorsal positions within the dorsal columns. With surround targeting (purple), axons from surround were in the most medial and dorsal positions and targeting in between center and surround (green) mixed center and surround axons. FIG. 6B includes data from responses of model WDR neurons with each stimulation target. Model PT is estimated as 50% of the MT. FIG. 6C includes data of raw changes in WDR firing rate at 20, 40, 60, and 80% of estimated MT. Lines with stars represent significant changes in the population response between positions (ANOVA, post-hoc Tukey’s test, p<0.05). FIG. 6D provides model changes in DC axon recruitment with surround targeting (purple). The difference between surround and center axon recruitment is maximized at 40-50% MT, below the estimated PT. In FIG. 6E, individual DH neuron responses were sorted into three groups based on the location where they were recorded. FIG. 6F provides normalized changes in pEX neuron activity split up by recording position. FIG. 6G includes data of raw changes in pEX neuron activity split up by recording position. FIG. 6H provides pEX neurons classified as responders by recording location. FIG. 61 provides percent of pEX neurons excited compared to spontaneous activity at each recording position for peripheral brush, crush, and A -ES.

Example 5

[0079] FIGS. 7A-7D include representative data using the distributed model for optimization with different amplitude/frequency/targeting combinations. FIG. 7A includes representative response of 25 sample populations to changing frequency and amplitude. Eight lines represent samples of dorsal column populations and dark lines represent the population median. Error bars represent the 25^th and 75^th percentiles. FIG. 7B includes the same data as FIG. 7A, but response of all neurons were sorted by neuron response amplitude. FIG. 7C includes the percent of the target population activated by amplitude and frequency. Error bars represent the standard deviation across 10 sample populations. FIG. 7D includes the percent of model nerve fibers firing faithfully with the stimulation frequency by amplitude and frequency.

Example 6

|0080] FIGS. 8A-8C include representative data from experimental recordings of receptive field targeted stimulation. FIG. 8A represents an experimental setup for recording responses to receptive field targeted stimulation of different receptive field areas (tibial versus common peroneal nerve). FIG. 8B includes representative data from neurons recorded at different positions from receptive field targeted stimulation of the tibial and common peroneal branches of the sciatic nerve. FIG. 8C includes representative data of the response of all neurons to receptive field targeted stimulation at 50 Hz.

Example 7

[0081] FIGS. 9A-9D include representative model responses and experimental recordings of receptive field targeted dual frequency stimulation. FIG. 9A represents targeted dual frequency stimulation in the computational model. Individual frequencies were applied to zones 1 and 2. Zone 2 has two nodes but only 50% activation of fibers in each node. FIG. 9B represents average model WDR neuron responses to dual frequency stimulation relative to 50 Hz stimulation of the center zone alone. The left plot shows response of the default model state and right plot shows response of the model with increased number of pain fibers activated. FIG. 9C includes a representative experimental setup for recording responses to receptive field targeted dual frequency stimulation. The tibial nerve and the common peroneal (CP) nerve were stimulated with separate cuff electrodes and recorded putatively excitatory (pEX) neuron responses in the dorsal horn. Responses were evaluated in positions 2 and 3 based on the regions where net inhibition from stimulation was observed (see also FIG. 8). FIG. 9D includes representative data demonstrating the difference between stimulation at 50 Hz alone on the tibial nerve and all paired frequencies. As in the model case, responses are normalized to the greatest change between paired frequencies and 50 Hz alone.

Example 8

[0082] FIGS. 10A-10E include representative data obtained using the distributed model for optimization with different frequencies for each zone. There are two zone 2 nodes, two zone 3 nodes, and one zone 1 node. Node 1 receives inputs to 100% of the input fibers while zones 2 and 3 receive inputs to 50% each. Optimized score across frequency combinations for all three zones. The best combination is shown in blue with their frequencies labeled in order by zone number. FIG. 10A includes data from combinations of unique frequencies to each node based on the zone. FIG. 10B includes optimal zone 1 frequency for each pair of frequencies in zones 2 and zones 3 to project the solution onto two-dimensional frequency space. FIG. 10C includes percent of the baseline response of the model WDR neurons across all three zones. FIG. 10D provides a representation of power consumption for each optimal frequency combination assuming a uniform pulse width. FIG. 10E includes optimized scores combining efficacy (FIG. 10C) and efficiency (FIG. 10D) for each combination of surround frequencies with the optimal zone 1 frequency. The efficiency score was calculated as the sum of all frequencies applied to zones 1, 2, and 3. The efficacy score was calculated as the mean reduction in WDR firing rate in zones 1 , 2, and 3 compared to the baseline firing rate with no stimulation input. Both efficiency and efficacy scores followed normal distributions across all frequency combinations, so the optimized scores were calculated by combining the z-score for efficiency with the z-score for efficacy. The z-score for efficacy was weighted twice as high as the z-score fore efficiency.

Claims

26 CLAIMS What is claimed is:

1. A method of identifying an optimized electrical stimulation pattern for delivering electrical stimulation to a subject for pain reduction, the method comprising: selecting an electrode geometry and/or electrode position; selecting at least one stimulation pulse parameter; evaluating the electrode geometry and/or the electrode position and the at least one stimulation pulse parameter based on neural activity in at least one target zone using a computational model of a neuronal network; and identifying at least one spatially optimized candidate stimulation pattern capable of reducing pain.

2. The method according to claim 1 , wherein the neural activity comprises a reduction in activity of pain-transmitting neurons.

3. The method according to claim 1 or claim 2, wherein the at least one target zone comprises a receptive field zone.

4. The method according to any of claims 1 to 3, wherein the spatially optimized candidate stimulation pattern activates a target zone comprising a center of a receptive field and/or a target zone comprising an area of surround inhibition.

5. The method according to any of claims 1 to 4, wherein the at least one target zone comprises three or more target zones.

6. The method according to any of claims 1 to 5, wherein at least one target zone is located in the spinal cord.

7. The method according to any of claims 1 to 6, wherein at least one target zone is located in the peripheral nervous system.

8. The method according to any of claims 1 to 7, wherein selecting the electrode geometry comprises selecting or altering which electrode contacts are active.

9. The method according to any of claims 1 to 8, wherein selecting the electrode position comprises selecting or altering a physical position of the electrode lead in the subject.

10. The method according to any of claims 1 to 9, wherein selecting at least one stimulation pulse parameter comprises selecting or altering pulse repetition frequency, pulse amplitude, pulse duration, pulse shape, temporal pattern, and any combinations thereof.

11. The according to claim 10, wherein the average pulse repetition frequency ranges from about 10 Hz to about 400 Hz.

12. The according to claim 10, wherein the pulse amplitude ranges from about 1 mA to about 20 mA.

13. The according to claim 10, wherein the pulse width ranges from about 30 ps to about 600 ps.

14. The according to claim 10, wherein the temporal pattern comprises a random, stochastic, periodic, and/or bursting pattern.

15. The method according to any of claims 1 to 14, wherein the computational model of the neuronal network simulates activity of a wide dynamic range (WDR) neuron.

16. The method according to claim 15, wherein the activity of the WDR neuron in the computational model is a proxy for pain.

17. The method according to any of claims 1 to 16, wherein the computational model of the neural network comprises input responses from computational models of dorsal column axons to the spatially optimized candidate stimulation patterns.

18. The method according to claim 17, wherein the dorsal column axons input responses are calculated based on diameter and position of the dorsal column axons, wherein the position corresponds to different receptive field zones.

19. The method according to claim 18, wherein the spatially optimized candidate stimulation patterns activate specific populations of dorsal column axons.

20. The method according to any of claims 1 to 19, wherein the computational model of the neural network comprises three network zones comprising heterogeneous inhibitory and excitatory neural connections.

21. The method according to any of claims 1 to 20, wherein the method further comprises programming a neuromodulation device to deliver the at least one spatially optimized candidate stimulation pattern to a subject.

22. The method according to any of claims 1 to 20, wherein the method further comprises delivering the at least one spatially optimized candidate stimulation pattern to the subject to reduce pain.

23. The method according to claim 22, wherein the at least one spatially optimized candidate stimulation pattern is delivered to an area of the spinal cord.

24. The method according to claim 22, wherein the at least one spatially optimized candidate stimulation pattern is delivered to an area of the peripheral nervous system.

25. A method using electrical stimulation to reduce pain in a subject, the method comprising: programming a neuromodulation device to deliver at least one spatially optimized candidate stimulation pattern to a subject, wherein the at least one spatially optimized candidate stimulation pattern is optimized with respect to electrode geometry and/or electrode position and at least one stimulation pulse parameter; and delivering the at least one spatially optimized candidate stimulation pattern to the subject to reduce pain.

26. A system for using electrical stimulation to reduce pain in a subject, the system comprising: an electrode sized and configured for implantation in proximity to neural tissue; and 29 a pulse generator coupled to the electrode, the pulse generator including a power source comprising a battery and a microprocessor coupled to the battery, wherein the pulse generator is configured to generate electrical signals for delivering a spatially optimized candidate stimulation pattern to the subject to reduce pain.