SYSTEM FOR PROVIDING NEUROMODULATION, ESPECIALLY NEUROSTIMULATION

Abstract

A system for providing neuromodulation, especially neurostimulation, comprising a neuromodulator for neuromodulation of a cervical spinal cord, a dorsal neuromodulation array for providing neuromodulation to the cervical spinal cord, a sensor for measuring movements and/or muscle activations of a patient, a mapping module, which is configured and arranged for implementing a mapping between desired muscle activation patterns and neuromodulation parameters, and a processing module which is configured and arranged for processing signals provided by the sensor, further forwarding the processed signals to the mapping module and directing the mapping module output to the neuromodulator.

Claims

1. A system for providing neuromodulation, especially neurostimulation, comprising at least one neuromodulator for neuromodulation of a cervical spinal cord, at least one dorsal neuromodulation array for providing neuromodulation to the cervical spinal cord, at least one sensor for measuring movements and/or muscle activations of a patient, at least one mapping module, which is configured and arranged for implementing a mapping between desired muscle activation patterns and neuromodulation parameters, and at least one processing module which is configured and arranged for processing signals provided by the at least one sensor, further forwarding the processed signals to the at least one mapping module and directing a mapping module output to the neuromodulator.

2. The system according to claim 1, wherein the at least one neuromodulator is a neurostimulator.

3. The system according to claim 1, wherein the neurostimulator is configured and arranged to provide neurostimulation in at least three channels independently.

4. The system according to claim 2, wherein the neurostimulator is configured and arranged to provide neurostimulation in a frequency range from 0-1000 Hz.

5. The system according to claim 1, wherein the at least one neuromodulator is at least partially implanted, especially fully implanted.

6. The system according to claim 1, wherein the at least one dorsal neuromodulation array is a dorsal neurostimulation array.

7. The system according to claim 1, wherein the at least one dorsal neuromodulation array has a plurality of active sites E, wherein the active sites E are placed in a region of approx. 2-25 mm from a midline of the patient's spinal cord.

8. The system according to claim 1, wherein the neuromodulation is a neuromodulation for selective muscle activation.

9. A system wherein a mapping stage and an integration stage may be encompassed by a same computer/microcontroller and that one or both are external to a patient, partially implanted or fully implanted.

10. The system according to claim 1, wherein the neuromodulation is configured and arranged to target posterior roots.

11. The system according to claim 1, wherein the at least one dorsal neuromodulation array has a plurality of active sites E, wherein the active sites E are placed in a region of approx. 4-10 mm from a midline of the patient's spinal cord.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0045] Further details of the present disclosure shall now be described in greater detail in connection with the drawings.

[0046] It is shown in:

[0047] FIG. 1 a schematical overview of an embodiment of the system according to the present disclosure, with which the method according to the present disclosure can be performed;

[0048] FIG. 2 a muscle activation map during the natural performance of the task, performed with the embodiment of the system and method according to the present disclosure according to FIG. 1;

[0049] FIG. 3 recruitment curves and selectivity indices used in the embodiment of the system and method according to the present disclosure according to FIG. 1; and

[0050] FIG. 4 a task specific reconstruction of spinal segments spatiotemporal activation map.

DETAILED DESCRIPTION

[0051] FIG. 1 shows a schematical overview of an embodiment of the system 10 according to the present disclosure, with which the method according to the present disclosure can be performed.

[0052] The system 10 comprises a neuromodulator 12.

[0053] Alternatively, the system 10 could comprise more than one neuromodulator 12.

[0054] Further, the system 10 comprises a dorsal neuromodulation array 14.

[0055] Alternatively, the system 10 could comprise more than one dorsal neuromodulation arrays 14.

[0056] The system 10 further comprises a sensor 16.

[0057] In an alternative embodiment the system 10 could comprises more than one sensor 16.

[0058] The system 10 further comprises a mapping module 18.

[0059] Alternatively, the system 10 could comprise more than one mapping module 18.

[0060] The system 10 further comprises a processing module 20.

[0061] The neuromodulator 12 is connected to the dorsal neuromodulation array 14.

[0062] The sensor 16 is connected to the processing module 20.

[0063] The processing module 20 is connected to the mapping module 18.

[0064] The processing module 20 is also connected to the neuromodulator 12.

[0065] The connection between the neuromodulator 12 and the dorsal neuromodulation array 14 is in the shown embodiment a direct and unidirectional connection.

[0066] However, also an indirect (i.e. with another component of the system 10 in between) and/or bidirectional connection would be generally possible.

[0067] The connection between the neuromodulator 12 and the dorsal neuromodulation array 14 is in the shown embodiment a cable-bound connection.

[0068] However, also a wireless connection, e.g. by a wireless network WSN, could be generally possible.

[0069] The connection between the sensor 16 and the processing module 20 is in the shown embodiment a direct and bidirectional connection.

[0070] However, also an indirect (i.e. with another component of the system 10 in between) and/or unidirectional connection would be generally possible.

[0071] The connection between the sensor 16 and the processing module 20 is in the shown embodiment established by a wireless network WSN.

[0072] However, also a cable-bound connection could be generally possible.

[0073] The connection between the processing module 20 and the mapping module 18 is in the shown embodiment a direct and bidirectional connection.

[0074] However, also an indirect (i.e. with another component of the system 10 in between) and/or unidirectional connection would be generally possible.

[0075] The connection between the processing module 20 and the mapping module 18 is in the shown embodiment established by a wireless network WSN.

[0076] However, also a cable-bound connection could be generally possible.

[0077] The connection between the processing module 20 and the neuromodulator 12 is in the shown embodiment a direct and bidirectional connection.

[0078] However, also an indirect (i.e. with another component of the system 10 in between) and/or unidirectional connection would be generally possible.

[0079] The connection between the processing module 20 and the neuromodulator 12 is in the shown embodiment established by a wireless network WSN.

[0080] However, also a cable-bound connection could be generally possible.

[0081] The neuromodulator 12 provides neuromodulation to the cervical spinal cord of a patient.

[0082] Further, the neuromodulation is provided by a dorsal neuromodulation array 14 to the cervical spinal cord of a patient.

[0083] In this embodiment, the neuromodulation is a neuromodulation for selective muscle activation.

[0084] In this embodiment, the neuromodulator 12 is a neurostimulator 12 and provides neurostimulation to the cervical spinal cord of a patient via the dorsal neuromodulation array 14, which is a dorsal neurostimulation array.

[0085] The neurostimulator 12 provides neurostimulation to posterior roots of a patient.

[0086] In general, the neuromodulator 12 could be partially implanted in the body of a patient.

[0087] However, the neuromodulator 12 also could be fully implanted in the body of a patient.

[0088] The neuromodulator 12 provides neuromodulation in a frequency range from 0-1000 Hz or, in some embodiments, between 10 and 1000 Hz.

[0089] A response to the provided neuromodulation is measured by the sensor 16.

[0090] In this embodiment, the response to the provided neuromodulation is/are muscle activations of a patient.

[0091] In an alternative embodiment, the response to the provided neuromodulation are additionally and/or alternatively movements of a patient.

[0092] In general, every other type of physiological response to the neuromodulation could be measured by the sensor 16.

[0093] The mapping module 18 implements a mapping between desired muscle activation patterns and neuromodulation parameters.

[0094] The processing module 20 processes signals provided by the sensor 16, forwards the processed signals to the mapping module 18 and directs the mapping module output to the neuromodulator 12.

[0095] Not shown in FIG. 1 is that the neurostimulator 12 provides neurostimulation in at least three channels independently and simultaneously.

[0096] However, also fewer than three channels may be chosen at any given time.

[0097] Not shown in FIG. 1 is that the dorsal neuromodulation array 14 could have a plurality of active sites E, wherein the active sites E are placed in a region of approx. 2-25 mm, especially 4-10 mm from the midline of the patient's spinal cord.

[0098] Not shown in FIG. 1 is that at least one sensor 16 could measure brain signals.

[0099] Not shown in FIG. 1 is that brain signals could be measured by extracellular recording.

[0100] Not shown in FIG. 1 is that extracellular recording of brain signals may include single-unit recording, multi-unit recording, field potential recording and/or amperometry.

[0101] In general, not shown in FIG. 1 is that the at least one sensor 16 could be or could comprise an inertial measurement unit (IMU), an optical sensor, a camera, a piezo element, a velocity sensor, an accelerometer, a magnet sensor, a torque sensor, a pressure sensor, a force sensor, a displacement sensor, a contact sensor, an EEG measurement unit, an EMG measurement unit, a goniometer, a magnetic field sensor, a hall sensor and/or a gyroscope and/or a motion tracking video camera, or a infra-red camera.

[0102] Not shown in FIG. 1 is that two or more sensors 16 could form a sensor network 16.

[0103] Further not shown in FIG. 1 is that the mapping stage and the integration stage may be encompassed by the same computer/microcontroller and that the one or both are external to the patient partially implanted or fully implanted.

[0104] Not shown in FIG. 1 is that the sensor 16 could also be directly connected to the mapping module 18 and/or the neuromodulator 12 and/or the dorsal neuromodulation array.

[0105] Also not shown in FIG. 1 is that the dorsal neuromodulation array 14 could be directly connected to the mapping module 18 and/or the processing module 20.

[0106] Not shown in FIG. 1 is that the mapping module 18 may be embodied as a microcontroller or as a computer.

[0107] Not shown in FIG. 1 is that the processing module 20 may be embodied as a microcontroller or as a computer.

[0108] Further not shown in FIG. 1 is that the processing module 20 may be embodied as an implantable processor.

[0109] These connections could be unidirectional and/or bidirectional connections, wireless connections and/or cable-bound connections.

[0110] As shown in FIG. 2, FIG. 3, and FIG. 4 the feasibility of the above paradigm has been demonstrated in an animal model.

[0111] As already mentioned above, the system and the respective method according to the disclosure is an epidural stimulation control system for recovery of reaching and grasping movements after paralysis.

[0112] This is accomplished by first characterizing the muscle activity patterns in healthy subjects, e.g. as done by the inventors in two Macaca Fascicularis during three-dimensional reaching and grasping tasks. Then the stimulation specificity that can be achieved using cervical EES applied by a customized multi-site electrode is described. The stimulation parameters are selected automatically to achieve a desired muscle activation pattern. Finally, it is shown that the present disclosure can elicit the same patterns of muscle activation as observed in natural arm movements using EES delivered from only a few stimulation sites. The system elicits the same activation patterns of arm muscles as observed during natural three-dimensional reaching by utilizing the one-to-all connectivity between Ia proprioceptive afferents and motoneurons. The mechanism recruits these Ia afferents in the cervical roots, which monosynaptically and specifically recruit motoneurons.

System Description

[0113] The system 10 is composed by the following components:

[0114] A fully implantable neuromodulator 12, i.e. a neurostimulator 12, able to stimulate at least 3 channels independently and simultaneously at different frequencies ranging from 10 Hz to 1000 Hz.

[0115] A dorsal neuromodulation array 14, i.e. a dorsal epidural array 14 with lateralized active sites to target the posterior roots of the cervical spinal cord with active sites placed anywhere between 4-10 mm from the midline.

[0116] A sensor 16 system measuring arm kinematics, muscle activity or brain signals.

[0117] A microcontroller that implements a mapping between desired muscle activation patterns and stimulation parameters.

[0118] An implantable processor that integrates the signals from the sensor 16 system, feeds these signals to the microcontroller and directs the microcontroller output to the implantable neurostimulator 12 which in turn delivers stimulation of the posterior roots via the dorsal epidural array 14.

Innovation Step

[0119] Certain innovation steps described here are described as follows:

[0120] Unlike previous approaches in connection with EES for motor recovery of the lower limbs, the goal of the present disclosure is to be selective for muscle activations rather than motoneurons in the spinal segments. In the cervical spinal cord, now it is possible to achieve far greater stimulation specificity than in the lumbar spinal cord due to the more spatially separated posterior roots of the cervical spinal cord. Thus, the algorithm for the system and method according to the present disclosure can be designed to target distinct muscle groups, which was not possible for the lumbar EES system. The neuromodulation is unlike in previous approaches a dorsal neuromodulation, which targets posterior roots.

[0121] A novel stimulation algorithm is described for choosing stimulation parameters in real time. The algorithm depends on stimulation at one or multiple sites in order to achieve any desired muscle activation map. However, the choice of the sites is limited such that stimulation sites with overlapping muscle activations should not be chosen simultaneously. It is defined a metric for assessing this activation overlap and use it to automate stimulation site selection.

Methods

[0122] To design the stimulation protocol, the inventors conducted two experiments. First, they studied the three-dimensional reaching and grasping behavior of intact macaques to verify that a consistent spatiotemporal pattern of motor unit activation existed (Experiments 1A and 1B). Second, they tested our ability to use EES applied at the cervical level to selectively recruit groups of functionally activated muscles (Experiment 2). Using the results from all experiments, they developed a novel stimulation protocol to activate the motor units observed in natural reaching and grasping. Then, the stimulation results were compared with those of recorded muscle activity.

Animals and Behavioral Experiments

[0123] The experimental protocol was approved by the cantonal (Fribourg) and federal (Swiss) veterinary authorities (authorization No. 2014_42E_FR). Experiments were performed on two female macaque monkeys (Mk-CA and Mk-SA Macaca Fascicularis). Mk-CA performed a reach and grasp drawer task in Experiment 1A (Schmidlin E, et al., Behavioral assessment of manual dexterity in non-human primates. J. Vis. Exp. JoVE 3258 (2011); Chatagny P, et al., Distinction between hand dominance and hand preference in primates: A behavioral investigation of manual dexterity in nonhuman primates (macaques) and human subjects. 3 (2013)) while Mk-SA performed, a three-dimensional reaching task in Experiment 1B. The Experiment 1A task consisted of reaching and opening a drawer, which contained a food reward. Mk-CA was trained to rest the hand on a pad while waiting for a starting cue. Once ready, the subject reached with the left arm to the drawer knob, opened the drawer and extracted the reward (food pellet) contained inside. The Experiment 1B task required Mk-SA to reach, grasp, pull and release a sensorized object presented by a robotic arm (KUKA LBR iiwa, KUKA CEE GmbH). In both monkeys, intramuscular electromyographic (EMG) activity from n=8 muscles was recorded during task execution: deltoid (DEL), biceps (BIC), triceps (TRI), extensor digitorum communis (EDC), wrist extensor (WRE), flexor digitorum superficialis (FDS), wrist flexor (WRF), abductor pollicis brevis (APB)).

[0124] For behavioral experiments, three movement phases were automatically extracted: arm extension, grasping and pulling, and movement end. The arm extension phase was defined to be from movement onset to target (drawer knob or sensorized object) acquisition. Pulling lasted from target grasp until either complete opening of drawer (Exp. 1A) or until release of the sensorized object (Exp. 1B). Movement end was defined by the picking of the reward in Experiment 1A or release of the sensorized object in Experiment 1B. Then projected the recorded EMG activity was projected to the anatomical location of the corresponding motoneurons in the spinal segments based on published studies (Jenny A, et al., Principles of motor organization of the monkey cervical spinal cord. J. Neurosci. 3, 567 (1983)) in order to reconstruct the spatiotemporal activation patterns of arm and hand motoneurons during each of these phases.

Recruitment Curve Experiments

[0125] Prior to Experiment 2, both monkeys underwent a surgical procedure in which a customized spinal implant, based on previously reported technology (Minev I R, et al., Electronic dura mater for long-term multimodal neural interfaces. Science 347, 159 (2015)), was surgically inserted into the epidural space of the cervical spinal cord. To gain access to the site, a laminectomy was executed at the vertebral level T1. Each spinal implant comprised n=7 independent electrodes E (which are here E1, E2, E3, E4, E5, E6, E8), or “active sites” through which current could be applied, and one “inactive” site E7. During Experiment 2, single pulse stimulation at different current amplitudes was delivered from each electrode of the spinal implant while intramuscular EMG activity was recorded from all 8 muscles for both monkeys. Then, the level of muscle recruitment over the stimulation current amplitudes for each active site were calculated separately and the recruitment curves were computed.

Results

Identification of Spinal Cord Activation Patterns During Movement (Experiment 1A and 1B)

[0126] The spatiotemporal maps of arm and hand muscle activation during reaching and grasping movements revealed a well-defined pattern of activation over movement phases (cf. FIG. 2, muscle activation map during the performance of a task). During the reach phase, the most active muscles are those of the deltoid and triceps muscles. These are used to stabilize and extend the arm towards the target. The transition to the grasp and pull phase was associated with the activation of extensor hand muscles together with flexor muscles of the forearm. An activation of the deltoid muscle then ensued for the rest of the pull phase. This activation pattern presented a series of well separated sequential muscle activation hotspots during the task. It was concluded that the reach and grasp tasks in Experiments 1A and 1B could be represented by a well-defined temporal pattern of muscle recruitment.

Muscle Responses Induced by Cervical EES

[0127] The recruitment curves computed for the different active sites highlighted a good spatial selectivity of the epidural implant, cf. FIG. 3 showing muscle recruitment curves and selectivity indices of three epidural electrode active sites E1, E2 and E6 (from top to bottom) for Mk-Ca and Mk-Sa. These active sites enabled the selective activation of arm/hand flexors, arm/hand extensors and digit muscles, respectively. However, to elicit these activations, the active sites of the epidural implant were placed directly over the dorsal columns of the cervical spinal cord. This means that the width of the clinical electrode implant must span the width of the dorsal horns of the spinal cord (on average ˜7 mm from midline) (Sherman J L, et al., Measurements of the normal cervical spinal cord on MR imaging. Am. J. Neuroradiol. 11, 369-372 (1990); Kirazh Ö, et al., Anatomy of the spinal dorsal root entry zone: its clinical significance. Acta Neurochir. (Wien) 156, 2351-2358 (2014)). It was concluded that EES of the cervical spinal cord may be able to access specific groups of functionally activated muscles.

Design of Stimulation Control System

[0128] Based on the results of our first experiment, it was found that it is possible to characterize the spatiotemporal motoneuron activations during reaching and grasping movements. These activation patterns are not fixed during movement execution, suggesting that EES stimulation parameters will need to be adjusted in real time throughout the movement duration. This hypothesis was supported by a pilot experiment in one monkey (Mk-Sa) in which continuous stimulation was delivered. This stimulation resulted in movement perturbation and inaccurate reaching. Thus, the proposed stimulation protocol was implemented in two steps: the calibration step and the online stimulation selection step.

[0129] The calibration step requires the identification of a mapping between stimulation active sites and the muscles activated for varying stimulation amplitudes. Thus, it is important to define accurate recruitment curves to construct this map. This was accomplished using the results obtained in Experiment 2 in which the optimal stimulation pins to evoke specific muscles were deduced.

[0130] In clinical application, single pulse EES at different amplitudes would be delivered during electrode implantation from each active site to map motoneuronal activations.

[0131] Thus the following algorithm and method can be performed by means of the system 10:

[0132] During movement execution, the intended muscular activation μ.sup.d is continuously decoded from cortical signals or other biosignals.

[0133] Then, based on the recruitment curve mapping, the active site e and its relative amplitude a.sub.i.sup.e are chosen in order to produce a motoneuronal activation μ(e, a.sub.i.sup.e) that would fit at best the decoded activation μ.sup.d:

[00001]$\begin{array}{cc}e,a_i^{e}s.t.\underset{e,a_i^e}{\min }\left({\mu }^d-\mu \left(e,a_i^e\right)\right)& \mathrm{Equation}1\end{array}$

[0134] In order to account for coordinated multi-muscular activations, the selection of two concurrent active sites e.sup.1 and e.sup.2 is allowed under the following conditions:

[00002]$\begin{array}{cc}\arccos \left(\frac{\mu \left(e^1,a_i^{e^1}\right).Math.\mu \left(e^2,a_j^{e^2}\right)}{.Math.\mu \left(e^1,a_i^{e^1}\right).Math..Math..Math.\mu \left(e^2,a_j^{e^2}\right).Math.}\right)>N^o.& \mathrm{Equation}2\\ \left({\mu }^d-\mu \left(e^1+e^2,a_i^{e^1}+a_j^{e^2}\right)\right)<\left({\mu }^d-\mu \left(e^3,a_k^{e^3}\right)\right),\forall e^3,a_k^{e^3}.& \mathrm{Equation}3\end{array}$

[0135] The condition described in Eq. 2 prevents that two active sites eliciting responses from the same muscles would be concurrently activated up to a predefined threshold of N degrees, which can be set for each individual; Eq. 3 provides that the selection of a single electrode is generally preferred over a combination of electrodes to reduce stimulation complexity.

[0136] These criteria can be scaled to as many active sites as desired (i.e. more than two active sites may be selected).

[0137] To implement the stimulation algorithm of the system 10, the implantable pulse generator (IPG) used must be able to deliver stimulation on at least three sites simultaneously.

[0138] In addition, the algorithm performs the updates to stimulation parameters at a rate of 10-100 Hz, though it can function also outside of these frequency ranges, depending on the complexity of the movement.

[0139] Thus, the IPG must also be able to switch active sites and update all stimulation parameters (i.e. amplitude, frequency, pulse with, etc.) with a rate of at least 10 Hz.

[0140] From Experiments 1A and 1B, it is possible to deduce the time-varying map of muscle activations during natural reaching and pulling.

[0141] The goal during task performance is to recreate these same activation sequences using EES.

[0142] From the activation map, a template set of muscle activations during the reach and pull phases was constructed to guide the stimulation parameter selection. To translate to a clinical setting, this template may be constructed using the unaffected arm or from averaged activities across a population of users and fine-tuned to the particular user before stimulator usage.

[0143] During real-time usage, reach and pull phases will be detected using either brain signals (by electroencephalography (EEG)) or other muscle/movement signals (by electromyography (EMG) of unaffected areas/inertial measurement units or other sensors).

[0144] This will then be passed to the implantable processor to finally activate the template electrode contacts to stimulate the correct muscles for the appropriate movement phase.

[0145] FIG. 4 shows a task specific reconstruction of muscle envelope spatiotemporal activation map.

[0146] Top: Task specific spatiotemporal activation map obtained from muscle activation recording during behavioral experiments.

[0147] Middle: The task-specific spatiotemporal activation was reproduced selecting the muscle activation obtained from single pulse EES delivered from the electrode active sites.

[0148] Bottom: The active sites used (here, E1, E2, E3, E5, E6, E8), and their current amplitudes were chosen as the ones mimicking at best the activation pattern obtained during task performance. E4 was not used due to its lack of effective muscle recruitment.

[0149] Thus, FIG. 4 illustrates an example of the performance of the described stimulation approach, where the muscle activation recorded during the performance of our 3D reach and grasp task is used as a proxy for the decoded motoneuronal activation μ.sup.d. In the stimulation selection step the implantable processor chooses the optimal stimulation parameters to elicit desired movement patterns. The resulting map showed a succession of spatially defined hotspots that were qualitatively similar to those found in the original task-specific map.

[0150] Note that the example control and estimation routines included herein can be used with various neuromodulation and/or neurostimulation system configurations. The control methods and routines disclosed herein may be stored as executable instructions in non-transitory memory and may be carried out by a control unit such as a microcontroller (or a computer) in combination with the at least one neuromodulator 12, dorsal neuromodulation array 14, sensor 16, mapping module 18, processing module 20 and/or other system 10 hardware. The specific routines described herein may represent one or more of any number of processing strategies such as event-driven, interrupt-driven, multi-tasking, multi-threading, and the like. As such, various actions, operations, and/or functions illustrated may be performed in the sequence illustrated, in parallel, or in some cases omitted. Likewise, the order of processing is not necessarily required to achieve the features of the example embodiments described herein but is provided for ease of illustration and description. One or more of the illustrated actions, operations and/or functions may be repeatedly performed depending on the particular strategy being used. Further, the described actions, operations and/or functions may graphically represent code to be programmed into non-transitory memory of a computer readable storage medium in a control unit (e.g. a microcontroller) of the system, where the described actions are carried out by executing the instructions in a system including the various hardware components in combination with an electronic control unit.

[0151] According to the present disclosure a method is disclosed, the method characterized in that the method is performed with the system as defined in this disclosure, especially as defined of any of claims 1-9.

[0152] The following aspects are explicitly disclosed according to the present disclosure:

[0153] Aspect 1: A method for providing neuromodulation, especially neurostimulation, comprising the steps of

[0154] providing neuromodulation to the cervical spinal cord of a patient,

[0155] measuring movements and/or muscle activations of a patient,

[0156] implementing a mapping between desired muscle activation patterns and neuromodulation patterns, and

[0157] processing signals indicating movements and/or muscle activations, further forwarding the processed signals and directing the implemented mapping between desired muscle activation patterns and neuromodulation patterns for providing neuromodulation.

[0158] Aspect 2: The method according to aspect 1, characterized in that the neuromodulation is a neurostimulation.

[0159] Aspect 3: The method according to aspect 1 or aspect 2, characterized in that the neurostimulation is provided in at least three channels independently.

[0160] Aspect 4: The method according to one of the preceding aspects, characterized in that the neurostimulation is provided in a frequency range from 0-1000 Hz, or between 10-1000 Hz.

[0161] Aspect 5: The method according to one of the preceding aspects, characterized in that the neuromodulation is a neuromodulation for selective muscle activation.

[0162] Aspect 6: The method according to the preceding aspects, characterized in that the neuromodulation is configured and arranged to target posterior roots.

[0163] The following claims particularly point out certain combinations and sub-combinations regarded as novel and non-obvious. These claims may refer to “an” element or “a first” element or the equivalent thereof. Such claims should be understood to include incorporation of one or more such elements, neither requiring nor excluding two or more such elements. Other combinations and sub-combinations of the disclosed features, functions, elements, and/or properties may be claimed through amendment of the present claims or through presentation of new claims in this or a related application. Such claims, whether broader, narrower, equal, or different in scope to the original claims, also are regarded as included within the subject matter of the present disclosure.

REFERENCES

[0164] 10 System for neuromodulation/neuro stimulation [0165] 12 Neuromodulator/neurostimulator [0166] 14 Dorsal neuromodulation array/dorsal epidural array [0167] 16 Sensor [0168] 18 Mapping module [0169] 20 Processing module [0170] EX Electrode X/active site X [0171] E1 Electrode 1/active site 1 [0172] E2 Electrode 2/active site 2 [0173] E3 Electrode 3/active site 3 [0174] E5 Electrode 5/active site 5 [0175] E6 Electrode 6/active site 6 [0176] E8 Electrode 8/active site 8 [0177] APB Abductor pollicis brevis [0178] BIC Biceps [0179] DEL Deltoid [0180] EDC Extensor digitorum communis [0181] FDS Flexor digitorum superficialis [0182] TRI Triceps [0183] WRE Wrist extensor [0184] WRF Wrist flexor