Method of treating symptom by population coding of nerve information

Abstract

To provide a method related to patterned intensity modulation of a nerve tissue, in general.SOLUTION: A specific embodiment provides a method of treating a symptom by preparing an electrode and modulating a parameter of stimulation given by the electrode. The modulated parameter of stimulation is associated with intensity of stimulation and varies according to an input parameter of stimulation or time. As the input parameter of stimulation, an individual waveform (that is, Ψ or psi) may be selected. The waveform may vary for each pulse.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method for treating medical conditions by providing patterned intensity modulation of neural tissue. [Background technology]

[0002] Sensory perception is a complex process that involves many circuits and dimensions in neural pathways before it reaches conscious perception. Sensory loss is one of the more serious consequences of upper limb amputation. Natural sensation is essential for prosthetic limbs. Sensation is important for precise motor control, personal self-awareness, perception, and emotional communication. To date, no practical long-term sensory replacement has been demonstrated. Amputee people must rely on visual and auditory feedback from the device's motors to control the prosthesis. The prosthesis is perceived by the user as a foreign tool extending from the body, not as a part of the user's own body. In addition to improving control of the prosthesis, sensory perception is also important for providing a sense of embodiment and reducing phantom pain. Summary of the Invention [Problem to be solved by the invention]

[0003] As previously discussed, sensory perception is a complex process involving many circuits and dimensions in neural pathways before conscious perception can be achieved. By subjecting a subject's sensory pathways to stimulation that directly describes the sensation of the stimulus, insight into the processing of these circuits can be gained. Although described here in the context of sensory systems, embodiments of the present invention may be applied to any neural input to higher nervous system processing. [Means for solving the problem]

[0004] The present invention generally relates to patterned intensity modulation of neural tissue. Certain embodiments provide methods for treating medical conditions by providing electrodes and modulating parameters of the stimulation delivered by the electrodes. The parameter of stimulation that is modulated relates to the intensity of the stimulation and is varied based on input parameters and time. Intensity refers to any parameter for a particular pulse shape that affects the number of nerve fibers excited by the stimulation pulse. In general, parameters that affect intensity vary in a pattern or as a function that is not scalar or identical to the distorted waveform caused by the input parameters. The intensity of the stimulation may vary either as a function or a predefined pattern that is dependent on the input, variable, and / or time. In addition to the intensity of the stimulation, the selection of the individual pulse shape (i.e., Ψ, psi) may also vary from pulse to pulse. The patterned intensity mode of stimulation is referred to herein as "population-based encoding" of neural information, where the intensity and pulse shape determine which of the patient's axons are excited. By varying the recruitment of the population in a patterned manner, information is introduced into the nervous system as a variation of the "population" code. Non-limiting examples of nervous tissue that may be modulated include nerve axons or fibers, including peripheral nerves, such as autonomic and somatic nerves. Autonomic nerves include parasympathetic and sympathetic nerves. Nervous tissue may be axons of the central nervous system, such as axons of the brain or spinal cord or portions thereof.

[0005] In one embodiment, the present invention provides a method for introducing more information into a patient's nervous system via a population code. The method includes connecting a therapy delivery device to the patient's nervous tissue and activating the therapy delivery device. The method further includes varying parameters of the stimulation and fields within the patient's nervous tissue to change the population of excited fibers of the nervous tissue.

[0006] In another embodiment, the present invention provides a method of treating a medical condition in a patient, the method comprising connecting a treatment delivery device to neural tissue of the patient having the medical condition and activating the treatment delivery device, the method further comprising modulating a parameter of a stimulation provided by the treatment delivery device, the stimulation parameter relating to an intensity of the stimulation being varied based on an input parameter and a time of the stimulation.

[0007] In another embodiment, the present invention provides a method of improving sensation in an individual in need of improved sensation. The method includes connecting an electrode to neural tissue of a patient and activating the electrode. The method further includes modulating a parameter of a stimulation provided by the electrode. The stimulation parameter is related to an intensity of the stimulation and varies based on an input parameter and a duration of the stimulation. [Brief description of the drawings]

[0008] [Figure 1]FIG. 1 shows the stability and selectivity of a cuff electrode system implanted according to an embodiment of the present invention. FIG. 1A shows a schematic of three cuffs with 20 channels implanted in a subject's forearm: one 4-contact spiral cuff on the radial nerve of the forearm, and two 8-contact FINE on the median and ulnar nerves. The electrode leads were subcutaneously routed up the upper arm and connected to open helix percutaneous leads via Letechepia-type connectors. Single channel, charge-balanced monopolar stimulation was delivered using a universal external control unit (UECU) neurostimulator. FIG. 1B shows a schematic of typical sensory sites at threshold stimulation levels 3 weeks postoperatively. The cuff electrodes were highly selective, with each contact (M1-8, U1-8, R1-4) producing a unique site or unique sensation. The ulnar sites had the greatest overlap at threshold, but could be discriminated at suprathreshold responses. The regions drawn outside the hand outline indicate the extent of the sensory region surrounding the curvature of the fingers. As the drawings are based on line drawings by the subjects, some deviations in location may be due to subject error. Figure 1C shows a schematic of overlapping threshold regions for channels M2, M3, M4, M5, and M8 repeated over 3 and 10 weeks postoperatively, which represented consistent regional perception. These regions were prioritized as promising for sensory feedback of grasping. Location was relatively stable regardless of the stimulation waveform used. Figure 1D shows a graph of normalized mean charge density for all channels on the cuff, with the median (blue), ulnar (green), and radial (red) cuffs represented by solid lines. Shaded regions represent 95% confidence intervals, indicating no threshold elevation. Thresholds were determined by an unbiased stepwise search method. Frequency was kept constant at 20 Hz. The graphs in Figure 1E show that there were no significant changes in thresholds over time for midline channels M3, M4, and M5 from follow-up up to and beyond 68 weeks (p=0.053, 0.587, and 0.773, respectively). [Diagram 2]Figure 2 shows the waveform patterns. Figure 2A shows the pattern of classical square-wave, charge-balanced, anodal-preceding stimulation. Parameters such as pulse intensity (PA), pulse width (PW), interpulse interval (IPI), and frequency (f) are kept constant in most neurostimulation systems. Figure 2B shows a schematic diagram of a subject's sensory area. In general, PA and PW modulated the sensory area of ​​recruitment at suprathreshold levels with constant parameter stimulation. M5 showed channel-specific recruitment patterns as PW increased from 24 μs to 60 μs. M3 showed that recruitment of the sensory area was also achieved by increasing PA from 1.1 mA to 2.0 mA. These recruitment patterns were consistent with the patterns of finger innervation. Figure 2C shows the pattern of bursts at variable frequencies, where 2–10 prepulses of 250–500 Hz were applied prior to constant frequency (ftap) stimulation, resulting in a tapping sensation. Constant stimulation alone produced paresthesia. FIG. 2D (top plot) is an example of the largest modulation, where a sinusoidal (1 Hz) PW envelope was used to produce a natural sensation of pulsed pressure. FIG. 2D (bottom plot) is the resulting stimulation waveform, in this case with an interpulse interval (IPI) of 0.1 s (10 Hz). In a typical stimulation trial, an IPI of 0.01 s (100 Hz) was used according to the exemplary method of the present invention. [Diagram 3]Figure 3 shows the largest modulation, with a sinusoidal PW envelope. Figure 3A shows that at threshold (Bth), a pulsed pressure was felt in the thick circled area (M3, M4, M8, blue). Increasing the PW to a second threshold (Btingle) added a pulsed illusion, which typically overlapped with the pressure area but was larger in area. Further increases in PW expanded the area of ​​illusion while the area of ​​pressure remained the same. A moving sensation of touch, like someone lightly brushing the skin with their finger, was reported, with the same orientation consistently for a given set of stimulation parameters (R1, R4, red). Figure 3B shows a graph of the psychometric ratings of PWmax, showing a clear correlation between PW and perceived intensity. Subjects were given five levels of PWmax (100, 114, 131, 150, 167 μs), with each level presented 3–6 times in random order. Figure 3C shows a graph of the natural sensory threshold window measured for each channel of the midline cuff. Pressure generated at Bth (green) was accompanied by paresthesia at Btingle (black line, yellow) and was overwhelmed by paresthesia at BMask (red). The largest PW window in a channel was observed when PA was minimal. Higher levels of stimulation at M6 were avoided due to pain responses. [Figure 4]Figure 4 represents small-scale offset (SSO) modulation. Figure 4A shows a typical example of SSO modulation using a sinusoidal (1 Hz) PW with offset stimulation on M4 (solid line, red). PWpk-pk=90-95 μs was the minimum level of stimulation that produced a constant pressure sensation. For comparison, the threshold of pulsed pressure at the maximum magnitude of modulation is also shown (dotted line, blue). Figure 4B shows that the PWmin-max window that produced a constant pressure sensation was strongly influenced by PA, which changed both the size and the position of the window. Frequency was found to have only a small effect on the window, but not on the intensity (D). There was no response when PA was 0.5 mA. The data suggest that above 0.8 mA PA the window of sustained pressure sensation is reduced. Figure 4C shows that contralateral pressure corresponding to the indicated frequency can control the intensity of the constant pressure sensation. Subjects were presented with SSO modulation on channel M4 with IPIs of 50, 20, 10, 5, and 2 ms (20, 50, 100, 200, and 500 Hz) and instructed to indicate the perceived pressure with the contralateral hand. The intensity of the perceived constant pressure ranged from 0 to 500 grams (less than 1 pound). [Diagram 5] FIG. 5 depicts a functional task with sensory feedback. FIG. 5A is a photograph showing that without the sensory feedback system activated, the subject was unable to exert adequate grip force during the challenging task of picking and removing the stems of cherries. FIG. 5B is a photograph showing that with sensory feedback activated, the subject was able to tell when contact was made and was able to pick the cherries without damaging them. FIG. 3C is a graph showing the performance in the blindfolded and sighted conditions with and without sensory feedback during the cherry task, with a clear improvement in functional ability even in the sighted condition. FIG. 3D is a graph showing that the responses to the embodiment questionnaire regarding the ratings of embodiment were significantly consistent with the responses regarding the control when sensory feedback was on, a significant improvement from the functional task experience without sensory feedback. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] In the present disclosure, a condition may be referred to as being "treated." This does not necessarily mean that the patient's condition is cured, but may also mean that the symptoms are improved or minimized. A "condition" may include any biological function, disease, or injury in which it is desired to improve the patient's condition. A "patient" refers to a mammal, preferably a human, having an undesirable condition. A "treatment delivery device" includes any device capable of stimulating the patient's neural tissue, such as, but not limited to, electrodes / leads and / or catheters / drug pumps. Furthermore, when referring to an element described herein, the articles "a," "an," and "the" refer to one or more of that element, unless otherwise specified. Furthermore, the word "or" means "and / or" unless otherwise specified.

[0010] When an element is said to be "on," "mounted," "connected," "coupled," "in contact," or "in communication with" another element, it is understood that the element may be directly on, mounted, connected, coupled, in contact, or in communication with the other element, respectively, or there may be intervening elements between them. In contrast, when an element is said to be "directly on," "directly mounted," "directly connected," "directly coupled," "in direct contact," or "in direct communication with" another element, there are no intervening elements present.

[0011] Embodiments of the present invention provide a neural interface that produces stable, reproducible, and sustained natural tactile sensations in humans with upper limb amputees. The neural interface can produce multiple independent sensations across the entire phantom limb. Stimulation parameters are also provided. For example, by varying the parameters of the time-varying pattern of stimulation intensity, complex, multiple, natural sensations can be produced at a single site. This parameter variation can produce a variety of tactile sensations in an amputee subject, such as tapping, steady pressure, moving light touch, and vibration. The area and intensity of sensations can be controlled by the intensity and frequency of the stimulation, respectively. According to embodiments of the present invention, sensations at the fingertips of the prosthetic hand and between the splayed fingers allow the subject to perform difficult maneuvers that would not be possible otherwise. The sensations allow the subject to more naturally integrate the prosthetic hand as a part of their own body.

[0012] Also, embodiments of the present invention provide a neural interface for sustained restoration of tactile sensation. The neural interface may be used for prosthetic limbs. One embodiment of the device of the present invention includes a nerve cuff with multiple independent stimulation channels. Each cuff may be implanted around a peripheral nerve without penetrating the neuroprotective tissue.

[0013] As shown diagrammatically in Figure 1A and described in more detail below, three nerve cuff electrodes were placed around the nerves in the forearm of one amputee with a wrist disarticulation and the upper arm of a second amputee with a below-elbow amputation. The first subject maintained the implants for over 18 months, and the second subject for over 11 months. The interface technologies chosen were a self-sizing spiral nerve cuff electrode and an eight-channel flat interface nerve electrode (FINE). The FINE holds the nerve in an oval shape to facilitate selective stimulation. The stability and selectivity of the cuffs over a year allowed for stimulation waveforms that allowed the restoration of natural touch without the "tingly" sensation or paresthesia that often accompanies electrical stimulation. Sensory feedback improved the patient's ability to perform functional tasks and increased the patient's sense of somatization of the prosthesis. Other trials used direct peripheral nerve stimulation to produce sensation. Although single-channel nerve cuff electrodes have been able to produce sensory hand sensations for about 40 years, the predominant response to stimulation was fist clenching and paresthesia. In recent years, intrafibrous electrodes implanted within the nerve have been able to produce some degree of tactile sensation, but these are only a short-term solution, as thresholds continue to rise and complete loss of function can occur as early as 10 days. Approximately 30-50% of stimulation channels in intrafibrous electrode studies have been accompanied by paresthesia.

[0014] Alternatives to direct nerve stimulation have been explored to restore sensation to amputees. Sensory substitutes, typically vibration at distant sites on the body surface, have been applied to provide information about lost sensations such as pressure at the fingertips of a prosthetic limb. The substitute must be interpreted in terms of both mode (pressure encoded by vibration) and location (events at the fingertips encoded by sensations on the chest). Sensory substitutes have currently been used as the main alternative means of providing sensory feedback, and in limited settings have been shown to improve prosthetic control. However, this technology has not been widely adopted, possibly because the discomfort may outweigh the benefits if vibration is applied for too long, and because the substitute requires mental effort to interpret. The substitute rarely feels as "natural" as the lost original sensation.

[0015] Another alternative technique is cortical stimulation, where electrodes are inserted into the sensory cortex of the brain. For example, in one embodiment, patchy sensations distributed throughout the phantom hand may be created and controlled.

[0016] The multi-channel nerve cuff electrodes described herein produce sustained, stable, selective, restored sensation in humans. Such electrodes and associated devices provide consistent, complex, natural tactile sensation at multiple, patchy sites.

[0017] <The peripheral interface is selective and stable> As shown diagrammatically in Figure 1A, monopolar electrical nerve stimulation was applied to one of 20 channels implanted around the median, radial, and ulnar nerves of each subject during each week of the study. The threshold for sensation was determined by gradually increasing the stimulation intensity until the subject indicated that they felt something. For each supraliminal stimulation, the subject verbally described the sensation and hand-drawn the location where they perceived it. As shown diagrammatically in Figure 1B, stimulation of 19 of the 20 available channels produced sensations at unique locations on the sensory limb, demonstrating the high degree of selectivity achieved by the multichannel cuff. As shown in Figure 1C, the location of sensation was reproducible and stable. Sensation occurred at multiple locations, including the tips of the thumb and index finger, which was highly desirable because the prosthetic hand exerts forces primarily at these locations during functional tasks. All of the perceived locations were consistent with the innervation patterns of the median, ulnar, and radial nerves in which the corresponding electrodes were implanted, suggesting that the sensory pathways from the amputated limb and the subject's perception were not affected by cortical remapping, supporting the findings of previous short-term studies.

[0018] In addition to site stability, the minimum amount of stimulation required to produce sensation, i.e., threshold, was also stable. From week 2 to week 8, sensory thresholds were 95.5 ± 42.5 (n = 59), 70.7 ± 59.2 (n = 50), and 40.7 ± 12.4 nC (n = 24) for the median, ulnar, and radial nerves, respectively. Linear regressions over the 8-week period for each channel showed that thresholds remained unchanged (18 / 19) or decreased (1 / 19), strongly suggesting that the stimulation, nerve, and electrode were stable (p < 0.05, Figure 1D). Threshold tracking for a subset of channels was further extended beyond week 68, showing a trend toward continued stability (Figure 1E).

[0019] No muscle contractions were observed during stimulation, which was expected since all cuffs were implanted distal to the motor nerve branches of the remaining muscles. Furthermore, stimulation artifacts did not interfere with myoelectric control. The subjects in this study were highly active myoelectric prosthesis users who had repeatedly injured their prostheses. Thus, embodiments of the present invention provide a selective and reliable neural interface for sensory feedback in active individuals, regardless of the ongoing strain the subjects place on their body and prostheses.

[0020] <Stimulation with time-invariant parameters produces parasthesia> As illustrated in FIG. 2A, classical neurostimulation involves a train of identical, charge-balanced electrical pulses characterized by pulse intensity (PA), pulse width (PW), and pulse repetition frequency (f) or interpulse interval (IPI=1 / f). Traditionally, these three parameters are invariant over time and their values ​​are fixed as follows: PA(Φ i )=PA0, PW(Φ i )=PW0, and IPI(Φ i )=IPI0=1 / f0. Several variations of PW0, PA0, and f0 were explored and the resulting sensations were mapped. When the constant pulse train was given, subjects consistently reported an unnatural paresthesia over a 10-month period, describing it as "electrical" in 96% of 151 trials. Paresthesia was the predominant sensation at all frequencies tested, ranging from 1 Hz to 1000 Hz. There was a weak correlation between frequency and the intensity and size of the perceived sensation. As shown in Figure 2B, increasing PA0 or PW0 monotonically increased the intensity and / or the sensation area. Without wishing to be bound by theory, this may be the result of the recruitment of other sensory nerve fibers that coexist with the same dermatome in the peripheral nerve. In rare cases, a pressure sensation accompanied the paresthesia, but this was masked by the tingling sensation. As previously reported, the paresthesia did not resolve and turn into a natural sensation throughout the duration of stimulation, up to 60 seconds.

[0021] Although it was promising that the percepts evoked by constant-parameter stimulation were selective and stable, the lack of a natural sensation across multiple somatosensory modalities remained a major limitation, so we presented a train of different stimuli whose parameters varied as a function of time.

[0022] <Time-varying pulse width (PW(Φ i ,t) stimulation produces a natural pressure perception. [The largest modulations are perceived as pulses.] Thus, although variation in the stimulus is important for perception, tapping is not useful for pressure perception. The pulse width should be slow (f mod The pulse width was modulated with a sinusoidal envelope (β = 1 Hz) so that the pulse width ranged from 0 μs to β μs (see Fig. 2D and Methods below). In response to such stimulation, the subject reported a natural pulsating pressure sensation at 1 Hz, which he described as "like putting my finger here and feeling my own pulse or heartbeat," and made the gesture of placing his finger on the carotid artery in his neck. The peak pulse width was adjusted to a perceptible level (β th ), he described the sensation as being "very lightly" and repeatedly pressed on the spatially distributed areas of the skin, as shown in Figure 3A. For visual confirmation, the subject was asked to tap on the healthy limb in sync, and the frequency of the tapping was increased to the modulation frequency (f mod ) As β increased, the sensation of a stronger pulse of pressure was felt, but the frequency was f mod To determine whether the perceived intensity varied with pulse width, subjects were asked to rate their perception at five different values ​​of peak pulse width β. As shown in Figure 3B, subjects' verbal ratings of perceived intensity were significantly correlated with β (p<0.05, R 2= 0.85, n = 22). As long as the pulse width was varied according to this pattern, the sensation was described as "natural" at all sensory sites on the hand. At the fingertips, the pulsating sensation was described as similar to pressing against the tip of a ballpoint pen. The sensory modalities perceived across all 19 active channels were pulsating pressure (86.1%), moving light touch (7.3%), and tapping (7.3%) across all trials. Responses on all channels to stimulation with modulated pulse widths during one experimental session and across multiple sessions are shown in Tables 1 and 2. Table 1 shows examples of sensory modalities during one experimental session. Each channel was specific to the modality. Pressure was as if the area was being pressed with a finger unless otherwise stated. At M6, increasing the stimulation intensity changed to a sensation like a pin being stuck in a vein. U4 produced occasional sensations, which were likely located at the edge of the nerve or at a site off the nerve. Table 2 shows the average channel response for each cuff with the largest magnitude modulation, and the value of n represents the number of unique spontaneous responses per channel and per experimental trial. When PW was varied sinusoidally, a natural sensation without tingling was obtained in all channels. The sum of each column may not be 100% because some findings were perceived multiple times.

[0023] [Table 1] Summary of results (average channel response) for maximum magnitude sinusoidal (1Hz) PW modulation [Table 2]

[0024] Next, the range of β is β that brings about natural perception. th <Β<Β tingle It was decided that. tingle When the β was increased, the subjects reported a slight tingling sensation in addition to the natural sensation, as shown in Figure 3A. mask ) > Βmask > B tingle When this occurred, the paraesthesia became dominant over the natural perception, as shown in Figure 3C.

[0025] [Small offset (SSO) modulation feels like a constant pressure.] We further refined the stimulus waveform to produce the perception of a constant pressure rather than a pulsating pressure. We found that by increasing the lower pulse width from 0 μs to the threshold level during sinusoidal pulse width modulation, subjects felt a less pulsating and more prolonged pressure. During very small modulations, subjects felt a sustained pressure that they described as "the most natural feeling" and "like someone had their fingers over my hand." The typical size of the modulated pulses (PW pk-pk ) was surprisingly small at 5 μs. The IPI0 of the stimulation pulse was 0.01 s, and the frequency of the modulation envelope (f mod ) was 1Hz. PW offset is the amount of B needed to produce a natural pulsating sensation. th The pressure was set to approximately 90% of the normal pressure (see Fig. 4A and Methods below). As shown in Table 3, a constant pressure was obtained in all channels (M2, M3, M4, M5, U1) at all sites on the palm side. PW offset When set to an appropriate level, PW pk-pk However, pulses of 1 to 10 μs produced a sustained sensation of pressure. pk-pk or PW offset When was too large or too small, the resulting sensation was either a paresthesia or a natural sensation produced by the maximum magnitude of modulation.

[0026] [Table 3]

[0027] PW required to produce a certain pressure pk-pk The size and range of the window are determined by the stimulus channel, PA0, PW, as shown in Figure 4B. offsetTo confirm that the natural sensation was the result of modulating the stimulation, we performed PW pk-pk When PW was decreased toward 0, the perception changed again to paresthesia. pk-pk When the pressure was increased, the sensation returned to a constant sensation of pressure. offset As the pulse width increased, subjects reported increasing intensity. tingle Above this threshold, pressure became associated with paresthesia. Thus, there exists a window for producing natural sensations that depends on the time-varying properties of the stimulus and the intensity of the stimulus.

[0028] [The frequency of stimulation controls the intensity of the natural sensation.] From microneurography studies, it is generally believed that the firing rate of axons encodes the intensity of pressure. Without wishing to be bound by theory, it is hypothesized that since the sensation of constant pressure can be achieved with a fixed frequency, the perception of the intensity of constant pressure could be modulated by the pulse repetition frequency (f=l / IPI). Subjects were presented with an initial stimulation with a pulse rate of IPI=0.02 s and were told that its perceived intensity was "5". Subjects then scored the subsequent sensations in comparison to the initial sensation. The lightest intensity (score 1) occurred for the longer IPI (0.2 s) and was described by subjects as feeling like "just having fingers on skin". The greatest intensity (score 13) occurred for the shortest IPI (0.002 s) and was reported as a "gripping" pressure. In another experiment, subjects were instructed to press a pressure sensor with the contralateral limb that matched the shape and location of the sensation perceived on the phantom limb. As shown in Figure 4C , there was a direct association between IPI and matched pressure sensation ( p < 0.05, R = 0.70, n = 25).

[0029] Sensory feedback improves functional ability The ultimate goal of restoring sensation is to improve the functional use of the prosthesis, increase the subject's confidence in using the prosthesis, and provide a sense of embodiment of the prosthesis. Sensation is essential for the performance of tasks that require precise control of grasping pressure. To test whether sensation improves control, subjects were asked to grasp a cherry with their prosthetic hand while pulling off the stem with their healthy hand. As shown in Figures 5A and 5B, subjects had to apply enough pressure to hold the cherry firmly in order to remove the stem without applying excessive pressure that would damage or crush the cherry. Subjects were tested with and without sensory feedback using a prosthetic hand equipped with force-sensitive resistive sensors on the fingertips. In both half of the tests, subjects could see the cherry being grasped and hear the motor sound of the prosthesis, while in the other half, vision and hearing were 100% blocked.

[0030] Without sensory feedback, subjects were able to remove 43% and 77% of the cherries with and without occlusion, respectively. With sensory feedback, subjects were able to remove 93% and 100% of the cherries with and without occlusion, respectively, as shown in Figure 4C. By performing tests at two rates, it was found that sensory feedback significantly improved performance with occlusion (p<0.001). Sensation is essential for controlled grasping without vision. With sensation, performance was significantly improved (p<0.005), and was actually maximized without blindfolding (100% accuracy). Thus, sensation improves performance even when the user can see their prosthesis. Without sensation, prostheses are often only used for crude tasks such as pushing or holding.

[0031] In addition, subjects expressed increased confidence in their functional abilities and greater somatization when sensory feedback was available. A one-tailed t-test of subjects' self-reported confidence during the functionality trials showed a significant increase in confidence (p = 0.0305). A modified prosthetic somatization survey (3, 17) was administered after performing the functional tasks with and without sensory feedback. There was a significant correlation (p = 0.002) between the improved scoring responses and somatization of the prosthesis with sensory feedback, suggesting that subjects had an increased sense of somatization, as shown in Figure 4D. This result supports the idea that sensation allows for more sensitive activities, improved bilateral activity, and therefore a more normal appearance of the prosthesis, integration of the prosthesis with self-image, and increased confidence.

[0032] [Consideration] The method of the present invention relates to a system of continuous, multi-site, multi-sensory sensory feedback for improved functional ability and quality of life. Perception is controlled by varying the electrical stimulation regime. It has been shown that the spatial extent, intensity, and quality of sensation can be controlled by independently controlling the pulse width, pulse intensity, and stimulation frequency, as well as the pattern of change of these parameters. Electrodes can achieve this level of control at multiple sites on a single nerve, allowing complete and independent control at 19 different sites on the hand using only three implanted cuffs with a total of 20 electrical contacts. These results remained stable, reliable, and reproducible over a period of 18 months, with no signs of likely change thereafter. The system is durable for the subjects' active and normal daily activities.

[0033] Tasks requiring precise prosthetic control, such as removing cherries from stems, which could not be performed reliably even with sight without sensation, could be performed without sensory feedback. Sensation reduces the visual and attentional requirements typically required to use myoelectric prostheses. Sensation reduces the visual and attentional requirements typically required to use myoelectric prostheses. Sensation increases confidence in one's ability to perform the task, as well as the ability to feel the pressure exerted upon an object. Furthermore, subjects report that it feels like they are actually gripping the object, rather than simply using a tool to grasp it.

[0034] Unlike sophisticated and often complicated experiments using non-human primates to indirectly estimate the effects of sensory stimulation, subjects were able to directly report the perceptions evoked by the stimulation. This immediate and detailed feedback enabled a new implementation of time-varying patterned stimulation for the control of perceptual quality.

[0035] Without wishing to be bound by any theory of the control mechanism of sensory perception, it is conceivable that a time-varying train of stimuli would recruit spatially distinct populations at different times, resulting in heterogeneous excitation patterns. For example, for the SSO modulation scheme, at the lowest pulse width, only a small population of axons would fire actively above threshold. Since the pulse width never falls below this level, this population would always be excited at a stimulation frequency that is consistently SA-type pattern. Only at larger pulse widths would it be possible to stimulate a more distant population of neurons. This population would show a more phasic or pulsatile excitation pattern, similar to the typical pattern for RA / PC fibers. It is unlikely that either population is composed exclusively of SA or RA / PC fibers. Thus, the resulting excitation pattern is not strictly natural. However, the results presented here would suggest that the thalamic relay and upstream processing in the primary sensory cortex columns are robust enough to correct errors in the patterns that generate perception. Normal brain processing is highly tolerant of abnormal patterns, categorizing them according to the best matching previous sensory experience. Optical illusions are an example of this property of the brain's categorization function. We hypothesize that as long as a stimulus persistently activates a sufficient population of SA fibres and transiently excites a sufficient population of RA / PC fibres, the brain will interpret the stimulus as naturally resulting from pressure applied by a phantom hand.

[0036] The subject's comments described herein were particularly compelling. Although the subject strongly preferred to have sensory feedback, he stated that it was natural and did not require additional interpretation as with other sensory substitution methods. Whenever the subject was asked about performing a task with sensory feedback turned on, he stated that he "knew he had it" whatever object was being tested. The subject repeatedly expressed a desire for a fully implantable, take-home system that would provide a permanent sensation ("I want that, I'd love that"). With sensation turned on, the subject's perceived hand and prosthesis were in near perfect harmony. However, with sensation turned off, the prosthesis was perceived by the subject as an extension of his own hand. The subject reported several episodes of sudden phantom pain per month, but stated that he had not experienced any since being implanted with the system 18 months prior.

[0037] Regarding the selectivity of the extraneural electrodes, almost all contacts produced sensations at unique, well-defined locations. Subjects were able to distinguish between sensations at multiple locations independently of each other, even when two locations were stimulated within the same cuff. Peripheral nerves have been shown to be somatotopically organized, so multichannel cuff electrodes can also produce sustained somatotopically selective results, as reported in acute trials with intrafascicular electrodes. Reports from acute trials with intrafascicular electrodes on two nerves at similar locations showed sensations in 9 of 32 channels (28%). In contrast, in the examples of the method of the present application, 19 of 20 channels (95%) produced patchy sensations. Intrafascicular electrodes implemented with a conventional constant pattern stimulation scheme were found to produce both paresthesia and some degree of tactile sensation. Intensity was modulated by stimulation frequency, as in the methods described herein. Sensation quality was controlled by introducing a time-varying stimulation implementation. By varying the stimulation pattern, a single stimulation channel was used to produce sensations of constant pressure, pulsating pressure, flutter, tapping, vibration, or paresthesia, all occurring at the same sensory site on the hand. Over 1.5 years after implantation, these results remain stable and show no signs of imminent change.

[0038] <Conclusion> The embodiments described herein provide a sustained and stable neuromodulation system in humans that produces multimodal sensations at multiple locations in the sensory hand that significantly improves task performance using prosthetic limbs. Time-varying modulation of stimulation controls the quality of the perceived sensation, and the frequency of stimulation produces a natural, graded sensory intensity. This is a fundamental change in stimulation architectures that can be applied in the fields of neural interfaces and sensory modulation, especially when it comes to sensory-related events such as pain.

[0039] Additionally, although sensory embodiments are described herein, this collective encoding using patterned stimulation intensities can be applied to any suitable afferent or autonomic nervous system inputs to other nervous systems, such as somatosensory perception, autonomic control of homeostasis, and pain perception. Other pathologies or functions include, for example, modulation of pain by TENS or spinal cord stimulation, or by any suitable means of nerve stimulation; taste; modulation of swallowing by stimulation of sensory afferents in the oropharyngeal pathway; gastric reflux; autonomic regulation, such as regulation of blood pressure by baroreceptor stimulation, insulin regulation by pancreatic nerve stimulation, and appetite regulation by stimulation of the enteric nervous system; restoration of hearing by cochlear nerve stimulation or cochlear implantation; sexual sensation and / or sexual enhancement; and / or regulation of incontinence, urination, and other urogenital functions. Method embodiments may be used for, but are not limited to, spinal cord stimulation, peripheral nerve stimulation, mechanical stimulation, and / or deep brain stimulation.

[0040] Embodiments of the present invention may also be applied outside of medicine: for example, the methods of the present invention could provide sensory input in haptic interfaces, and could also be used in robotics, gaming technology and other human-machine interfaces. EXAMPLES

[0041] Materials and Methods <Research design> Without wishing to be bound by theory, it is believed that direct nerve stimulation of the residual upper limb using selective, non-penetrating peripheral nerve cuff electrodes can produce graded sensation and proprioception in multiple regions of the sensory hand. After initial data suggested the existence of a stable nerve stimulation platform with multiple sensory regions, our specific goal was to determine a stimulation waveform that would produce a natural sensation of touch without paresthesia.

[0042] Criteria for recruitment as human subjects included unilateral upper limb amputees, age 21 years or older, and current or prescribed use of a myoelectric prosthesis. Potential subjects were excluded if they had poor health (untreated diabetes, chronic skin ulcers, history of untreated or ongoing infection) or severe, persistent pain in the residual or phantom limb.

[0043] <Method> Three electrodes were implanted by a surgeon in an outpatient surgical procedure. A multi-contact nerve cuff electrode was placed in the residual limb of a 46-year-old man who had undergone wrist disarticulation due to an occupational injury. The subject had been post-amputation for 18 months at the time of implantation and had been using a myoelectric prosthesis regularly for 7 months. An eight-contact flat interface nerve electrode (FINE) was implanted on the median nerve and the ulnar nerve, respectively, and a four-contact spiral electrode was implanted on the radial nerve. The opening size of the FINE was 10 mm wide and 1.5 mm long for both the median and radial nerves. The inner diameter of the spiral electrode for the radial nerve was 4 mm. The size of the electrodes was selected based on the anatomical knowledge of peripheral nerves in human cadavers and confirmed by the surgeon at the time of surgery. As shown in Figure 1, all electrodes were located in the mid-forearm and connected to percutaneous leads exiting from the upper arm. The implanted components (cuff electrodes, percutaneous leads, and connectors) were manufactured by Ardiem Medical (Indiana, PA) and sterilized with ethylene oxide by Ethox International (Buffalo, NY). Subjects were discharged on the same day of surgery.

[0044] The implanted electrodes were allowed to stabilize for three weeks before stimulation experiments were performed. During this recovery period, the subject reported no discomfort at the implant site and no abnormal phantom sensations. Experimental stimulation was then administered for up to ten seconds through each contact in weekly sessions. In all trials, subjects were blinded to the stimulation. After stimulation, subjects described any sensations they felt and drew their location on the schematic diagram of their hand. No-stimulation trials were randomly interspersed to eliminate subject bias.

[0045] <Experimental equipment> The stimulation system includes a computer that controls the stimulation parameters and sends commands to a single-board computer (Mathworks) running the xPC Target. The stimulator is a Universal External Control Unit (UECU) that is controlled through the xPC Target. Optical isolation between the wall-mounted equipment and the subject is provided by an isolator. To prevent overstimulation, the charge density is 50 μC / cm 2 Stimulation was limited to less than 50% of the duty cycle throughout all stimulation protocols.

[0046] The UECU was a custom-built stimulator manufactured by Ardiem Medical (Indiana, PA). It had 24 channels of current-controlled stimulator output with a maximum stimulation intensity of 5.6 mA, a maximum stimulation pulse width of 255 μs, and a compatible voltage of 50 V. The UECU delivered unipolar, biphasic, charge-balanced, anode-first stimulation pulses with a common cathode. Cathode feedback was provided through a 2” × 4” surface electrode on the dorsal aspect of the upper arm, just adjacent to the elbow.

[0047] <General Framework of Electrical Stimulation> The general stimulation waveform (Λ) is a train of pulses (Ψ) separated by interpulse intervals (IPI). For each pulse shape, pulse parameters (Δ) are selected to produce nerve excitation. To induce tactile perception, pulse parameters Δ and IPI are functions of external input measured over time (Φ(t)) and time (t). Perception is the result of processing by higher-order neural circuits and is influenced by both the excitation of peripheral receptors and the pattern of change in excitation. To induce natural perception, peripheral nerve stimulation is expected to depend on excitation changes, so Δ and IPI are defined as functions of the desired tactile perception and time, i.e., Δ(Φ(t),t) and IPI(Φ(t),t) (Equation 1). JPEG2024174872000005.jpg27168

[0048] This example shows the psychometric relationship between modulation of stimulation parameters and perception, where Ψ is a charge-balanced bipolar square pulse with parameters Δ={PA(Φ,t),PW(Φ,t)} and IPI(t), where PA is the pulse intensity and PW is the pulse width. The findings show that perception is controlled by the change of a parameter over time, which is called time-varying modulation. Variations of each of the stimulation parameters were systematically investigated according to the following conditions:

[0049] <Stimulation with time-invariant parameters> JPEG2024174872000006.jpg14156

[0050] <Time-varying frequency (IPI(Φ i ,t)) to stimulate> The burst pattern is defined as follows: JPEG2024174872000007.jpg29167where N is the number of burst pulses, which typically ranges from 2 to 10 (Figure 2C).

[0051] <Time-varying pulse width (PW(Φ i ,t) to stimulate [Maximum modulation] In this set of trials, the stimulation parameters of Equation 1 were slowly varied according to: JPEG2024174872000008.jpg26169a and b are parameters that control the magnitude of the pulse width modulation. In these trials, the pulse interval was kept constant at IPI0 = 0.01 s. The pulse width was slow (f mod The pulse width was modulated by a sinusoidal envelope of b = a = B / 2 (B = 1 Hz), where B is the peak value of the varying pulse width. Thus, the pulse width was varied from 0 μs to B μs, as shown in Figure 2D.

[0052] Small offset modulation (SSO) was defined by Equation 2, where a=(PW max -P.W. min ) / 2=PWpk-pk / 2, and b = PW offset (A typical PW pk-pk The value of is 5μs, IPI0 is 0.01s, f mod = 1Hz. PW offset is the amount of β required to produce the natural pulsating sensation shown in Figure 4A. th The setting was approximately 90%.

[0053] <Threshold detection method> The threshold was determined using the Single-Interval Adjustment Matrix, an unbiased adaptive staircase method. Parameters were set to 50% as the target capacity, which would result in true stimulation 50% of the time. The threshold search was completed after 12 or 16 reversals. Stimuli were delivered for 1 s and repeated if the subject wished. The stimulation frequency was kept constant at 20 Hz. To prevent sensory overstimulation, the stimulation pulse intensity (PA) and pulse width (PW) were increased in 10 steps of 0.1 mA each until the rough threshold was determined. The PA was then held constant at one step (0.1 mA) below the rough threshold, and an adaptive staircase method was used to refine the PW threshold with a resolution of 1 μs.

[0054] <Pressure matching on the other side> SSO modulation was delivered on channel M4. Subjects were presented with a 5-second stimulus for each trial and were then instructed to match the pressure sensation by pressing on the manipulator with the contralateral hand for 5 seconds. The final 2 seconds of matched pressure data were averaged for each trial. The manipulator was shaped to mimic the sensation to be perceived and was pressed with the contralateral palm at the same location as the sensation to be perceived. The manipulator was a balsa wood shape with a rounded flat end and approximately 1 / 2" diameter. The manipulator was attached to the end of a Tekscan (South Boston, MA) A201 Flexiforce sensor (0-1 lb range). The sensor was calibrated and sampled at 10 Hz by a data acquisition system.

[0055] <Functionality test> To test for improved sensory control, subjects were asked to grasp a cherry with their prosthetic hand while plucking the stem with their healthy hand. Pressure on the prosthetic hand was measured using thin Flexiforce sensors (A201 type) attached to the tips of the thumb and index finger of the prosthetic hand. Although subjects reported that they performed this task primarily by relying on pressure sensations at their fingertips, they also wore flexion sensors to provide feedback for opening the prosthesis. The sensory feedback stimuli used were intensity-controlled SSOs with linear mapping of sensor force across a range of pulse repetition frequencies from 10 to 125 Hz. Subjects were given no sensory feedback for 60 trials, and 30 trials in which they were given sensory feedback. In both cases, half of the trials were performed in a way that allowed the subjects to see the cherry being picked and hear the motor sounds of the prosthesis, while the other half were performed with 100% vision and hearing occlusion. Subjects wore sleep masks to occlude vision and noise-reducing earmuffs over earphones playing white noise for hearing occlusion. After each trial, subjects were asked to rate their confidence in their overall ability without knowing the results of the test.

[0056] To assess the subjects' level of somatization during the functional task with sensory feedback, subjects were given a modified somatization questionnaire immediately after the functionality test. This questionnaire contained nine responses that subjects could answer on a scale of -3 to +3, corresponding to "not at all true" to "very true," respectively. Three of the responses were related to somatization, while six were control responses related to task compatibility and suggestibility. The order of responses was randomized. The questionnaire was modified by changing the phrase "researcher's touch" to "object touch [manipulated during the functional task]." Functionality tests for both subjects were compiled and the results of the somatization questionnaire with "sensory feedback off" were compared with the results of the somatization questionnaire with "sensory feedback on."

[0057] <Statistical analysis> Unless otherwise stated, risk rates were determined with a statistical power of a = 0.05.

[0058] The above-mentioned description and examples are merely illustrative of the present invention and are not intended to be limiting thereof. Each of the disclosed aspects and embodiments of the present invention may be considered alone or in combination with other aspects, embodiments, or variations of the present invention. In addition, although certain features of an embodiment of the present invention may only be described in a particular drawing, such features may be incorporated into other embodiments shown in other drawings as long as they fall within the scope of the present invention. In addition, unless otherwise specified, no steps in the method of the present invention are limited to be performed in a particular order. Those skilled in the art may conceive modifications of the embodiments of the present invention that are consistent with the spirit and scope of the present invention, and such modifications are within the scope of the present invention. In addition, all documents cited herein are incorporated by reference in their entirety.

Claims

1. A nerve stimulator; a first multi-channel nerve cuff electrode; a second multi-channel nerve cuff electrode; and a third multi-channel nerve cuff electrode, the neurostimulator is configured to generate sensory stimulation that evokes a natural sense of touch in the individual; the first multi-channel nerve cuff electrode is configured for placement around a median nerve and is connected to the nerve stimulation device to selectively deliver the sensory stimulation; the second multi-channel nerve cuff electrode configured for placement around the ulnar nerve and connected to the nerve stimulator device for selectively delivering the sensory stimulation; the third multi-channel nerve cuff electrode having at least one third electrical contact configured for placement on a radial nerve and connected to the nerve stimulation device for selectively delivering the sensory stimulus; the sensory stimulus comprises a plurality of pulses having parameters including a pulse intensity (PA), a pulse width (PW), and an interpulse interval (IPI), the PW varying over time; The system, wherein the delivery of the sensory stimulus via at least one of the first multi-channel nerve cuff electrode, the second multi-channel nerve cuff electrode, or the third multi-channel nerve cuff electrode generates a sensory stimulus that evokes the natural sensation of touch in the individual as a tactile sensation having an intensity that occurs at at least one of the following points: at least one fingertip, the tip of the thumb, the palm, the back of the wrist, or the back of the hand.

2. 2. The system of claim 1, wherein the first multi-channel nerve cuff electrode includes eight first electrical contacts each configured to electrically communicate with a unique site on the median nerve.

3. 3. The system of claim 2, wherein the eight first electrical contacts of the first multi-channel nerve cuff electrode are configured to be positioned on the median nerve such that the sensory stimulation induces the natural sensation of touch in at least one of the at least one fingertip, the tip of the thumb, and the palm.

4. 2. The system of claim 1, wherein the second multi-channel nerve cuff electrode includes eight second electrical contacts each configured to electrically communicate with a unique site on the ulnar nerve.

5. 5. The system of claim 4, wherein the eight second electrical contacts of the second multi-channel nerve cuff electrode are configured to be positioned on the ulnar nerve such that the sensory stimulation induces the natural sensation of touch in at least one of a portion of a little finger, an outer portion of a palm, and an outer portion of a wrist.

6. 2. The system of claim 1, wherein the third multi-channel nerve cuff electrode includes four third electrical contacts each configured to electrically communicate with a unique site on the radial nerve.

7. 7. The system of claim 6, wherein the four third electrical contacts of the third multi-channel nerve cuff electrode are configured to be positioned on the radial nerve such that the sensory stimulation evokes the natural sensation of touch on at least a portion of the back of the hand.

8. The system of claim 1 further comprising the device including at least one sensor associated with operation of the device.

9. 10. The system of claim 8, wherein the device is a prosthetic device including at least one prosthetic hand, and the at least one sensor is at least one pressure sensor disposed on the prosthetic hand to detect a pressure measurement occurring between the pressure sensor and an object, the pressure measurement being indicative of contact of the prosthetic hand with the object.

10. The stimulation device receives the pressure measurements and generates a stimulation signal based on the pressure measurements. modulating said parameters of the sensory stimulation; a multi-channel nerve cuff electrode, or the at least one of the third multi-channel nerve cuff electrodes; 10. The method of claim 9, further configured to select at least one of the sensory stimuli to deliver the sensory stimulus. The system described in

Images