Methods and systems for closed loop vagus nerve stimulation triggering

Abstract

Provided herein are methods of treating a subject having or at risk of having a dysfunction including a method of treating a subject having or at risk of having a dysfunction comprising establishing a set of task outcomes for at least one task, the set of task outcomes including one or more patterns of progressive activity, a progression threshold, and one or more patterns of antagonist activity, monitoring a subject's activity performing the at least one task, including monitoring the one or more patterns of progressive activity, comparing the subject's monitored activity to the set of task outcomes, and triggering a stimulation of a nerve when performance of the at least one task, comprises the monitored one or more patterns of progressive activity that meets the progression threshold and does not involve the one or more patterns of antagonist activity.

Description

METHODS AND SYSTEMS FOR CLOSED LOOP VAGUS NERVE STIMULATIONTRIGGERINGCROSS-REFERENCE

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 513,795, filed July 14, 2023, which application is incorporated herein by reference.STATEMENT AS TO FEDERALLY SPONSORED RESEARCH

[0002] Aspects of the present disclosure were made with the support of the United States government grant No. N66001-17-2-4011, awarded by Defense Advanced Research Projects Agency. The government has certain rights in the present disclosure.BACKGROUND

[0003] The present disclosure relates generally to the fields of rehabilitation and medicine. More particularly, it concerns systems, devices, and methods for the application of vagus nerve stimulation.

[0004] There are approximately 800,000 strokes each year in the U.S. Many survivors are left with long-term upper limb hemiparesis, which can lead to disability. There is a clear and present need to develop interventional strategies to reduce this disability. Recently, a strategy based on delivering bursts of vagus nerve stimulation (VNS) concurrent with rehabilitation received FDA approval for the treatment of upper extremity motor deficits associated with chronic ischemic stroke. In the common implementation of this therapy, a therapist monitors physical rehabilitation and presses a button to trigger VNS when the participant is moving their upper extremity during the performance of exercises. Three clinical trials, including a Phase 3 pivotal study, demonstrate that VNS paired with rehabilitation significantly improves recovery of upper limb motor function compared to equivalent rehabilitation without VNS in individuals with chronic stroke. While these findings are encouraging, this approach leaves room for improvement. The development of optimized strategies to apply VNS therapy holds significant promise to leverage this technology to promote post-stroke recovery as well as other subject dysfunctions such as spinal cord injury (SCI), multiple sclerosis (MS), Cerebral palsy (CP), Nerve Injury, traumatic brain injury (TBI).

[0005] Supervised rehabilitation with VNS can generate long-lasting, clinically significant improvements in stroke subjects. However, subjects that receive VNS with rehabilitation can still exhibit residual deficits. Maximizing the clinical impact of VNS for stroke recovery may depend on the selection of a therapeutic process that can properly address many unique patterns of activity that take place during rehab. Clinical evidence supports this notion. Whereas subjectsdemonstrated significant increases when VNS was triggered by a therapist observing a pattern of activity, these same subjects failed to demonstrate any further gains when receiving unsupervised VNS that was not explicitly paired with a pattern of activity, even when stimulation was delivered for over a year.SUMMARY

[0006] Improved systems, devices, and methods for the application of VNS are disclosed herein. In particular, a method of treating a subject having or at risk of having a dysfunction comprising establishing a set of task outcomes for at least one task, the set of task outcomes including one or more patterns of progressive activity, a progression threshold, and one or more patterns of antagonist activity, monitoring a subject’s activity performing the at least one task, including monitoring the one or more patterns of progressive activity, comparing the subject’s monitored activity to the set of task outcomes, and triggering a stimulation of a nerve when performance of the at least one task, comprises the monitored one or more patterns of progressive activity that meets the progression threshold and does not involve the one or more patterns of antagonist activity.

[0007] The task may comprise a performance of a rehabilitative exercise, a performance of a game, a motor response, a sensory response, a physiological response, a cognitive response, or an activity of daily living.

[0008] The nerve may be a cranial nerve, the vagus nerve, or the 10thcranial nerve. The nerve may be any nerve related to a dysfunction. The dysfunction may be a motor disability, a sensory disability, a dysfunction of the ankle, knee, hip, thumb, finger / fingers, wrist, elbow, shoulder or trunk, a motor sensory dysfunction, a result of a stroke, multiple sclerosis, surgical procedures, traumatic brain injury, spinal cord injury, PTSD, aphasia, a result of a peripheral nerve or limb injury, a walking dysfunction, or ankle dorsiflexion.

[0009] The stimulation may occur at a frequency once every five seconds or longer. The stimulation may occur after at least a refractory period of a subject’s movement. The stimulation may not occur at or near an onset of the subject’s activity. The stimulation may be applied for a pre-determined time to invoke the subject’s Hebbian process. The stimulation may be limited to no more than once every 5 seconds. The stimulation may be not performed when a compensatory activity is detected.

[0010] The task outcome may include one or more of a task completion beyond the progression threshold, an activity, a sequence of activities, a pattern of activities, an activity repetition, a sequence repetition, a pattern repetition, a peak force, a peak range of motion, a peakacceleration, a peak velocity, a peak angular velocity, a peak angular acceleration, or electromyogram activity,

[0011] The antagonist activity may one or more of a compensatory movement, a non-sequential movement, a muscle tremor, a muscle spasticity, or a failure to activate a target muscle or a muscle group, or activation of muscles opposing the task outcomes, the at least one task, or one or more patterns of progressive activity.

[0012] The set of task outcomes may include 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 75, 100, 500, or 1000 tasks.

[0013] The monitoring may be done using one or more sensors in contact with the subject. The monitoring may be performed using a controller. The monitoring may be performed using a camera. The monitoring may be performed using one or more EMG sensors. The monitoring may be performed using one or more EEG sensors.

[0014] The method may comprise assessing a stop condition and instructing the subject to stop the subject’s activity when the stop condition is met. The method may comprise adjusting a duration of stimulation. The method may comprise adjusting the progression threshold for stimulation. The progression threshold may include a threshold duration. The method may comprise adjusting the threshold duration. The progression threshold may be established from the set of task outcomes. The progression threshold may be at least at the 50thpercentile, 75thpercentile, 80thpercentile, 85111percentile, 90111percentile, 95111percentile, 98thpercentile or 99111percentile of the set of task outcomes.

[0015] The method may comprise updating the set of task outcomes based on the monitoring of the subject’s activity performing the at least one task. The method may comprise adjusting an intensity of the stimulation within a repetition of the performance of the at least one task. The method may comprise adjusting an intensity of the stimulation based on activation of the subject’s recurrent laryngeal nerve. The method may comprise adjusting the progression threshold such that the subject receives a predetermined number of nerve stimulations per unit time. The predetermined number of nerve stimulations per unit time may be about 200 simulations per day.

[0016] The method may comprise adjusting the progression threshold based on a buffer of task outcomes. The method may comprise measuring at least one characteristic of the subject’s response to the at least one task and adjusting the progression threshold based on the at least one characteristic of the subject’s response to the at least one task. The measured at least one characteristic may be an inter-stimulation interval in response to a speech task. The speech task may be a sequence production task for conversational speech.

[0017] The method may comprise preprocessing activity data. Preprocessing activity data may comprise averaging or smoothing the subject’s activity performing the at least one task.Smoothing the subject’s activity performing the at least one task may comprise removing outliers or noise. Smoothing the subject’s activity performing the at least one task may comprise applying a moving average to a buffer of the monitoring of the subject’s activity performing the at least one task. The buffer may include a predetermined time of about 300 ms. Preprocessing activity data may comprise extracting one or more principal components from the subject’s activity. The one or more principal components from the subject’s activity may include attributes of displacement, velocity, acceleration, a Euclidean distance from the subject’s activity to the at least one task, one or more of changes in direction, intermittent differences in range, and intermittent differences in pauses. The one or more principal components from the subject’s activity may include anatomical features of the subject such as their body composition, muscles, joints, tendons, flexibility, range of motion among other subject- specific factors such as missing digits or severity of condition.

[0018] The method may comprise determining when a continuous improvement of the subject’s activity performing the at least one task occurs, and triggering a stimulation based on the continuous improvement. The continuous improvement of the subject’s activity performing the at least one task may be based on a quantitative measure of the subject’s function or task outcome.

[0019] Another aspect of the present disclosure provides a method of treating a subject having or at risk of having a dysfunction may comprise selecting a task including at least one associated activity, at least one task outcome, one or more patterns of antagonist activity, and a progression threshold, capturing a subject performing the task, determining when performance of the task meets the progression threshold, and stimulating the subject’s nerve when the performance of the task meets the progression threshold and does not involve the one or more patterns of antagonist activity.

[0020] Another aspect of the present disclosure provides a rehabilitation system for treating a subject having or at risk of having a dysfunction may comprise a computer processing device comprising a processor and a non-volatile storage medium with instructions for the processor to perform the method of any one of the methods herein.

[0021] Another aspect of the present disclosure provides a rehabilitation system comprising one or more computer processors and computer memory coupled thereto. The computer memory may comprise machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein.

[0022] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrativeembodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure.Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.INCORPORATION BY REFERENCE

[0023] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and / or take precedence over any such contradictory material.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present disclosure will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the present disclosure are utilized, and the accompanying drawings (also “Figure” and “FIG.” herein), of which:

[0025] FIG. 1 A shows an illustrative example of a method of delivering VNS concurrent with rehabilitative exercises captured using a rehabilitation system according to some implementations.

[0026] FIG. IB shows examples of unidirectional games, unidirectional movements and subject interfaces, and corresponding movement signals according to some implementations.

[0027] FIG. 1C shows examples of bidirectional games, bidirectional movements and user interfaces, and corresponding movement signals according to some implementations.

[0028] FIG. ID shows examples of game-based exercises and corresponding activity signals according to some implementations.

[0029] FIG. IE shows examples of repetition-based exercises and corresponding movement signals after processing by a static threshold algorithm according to some implementations.

[0030] FIG. IF shows examples of repetition-based exercises and corresponding movement signals after processing by a dynamic algorithm according to some implementations.

[0031] FIGS. 2A-2C are flow charts showing examples of a graphical description of triggering methods including a dynamic algorithm, a static threshold algorithm, and a periodic algorithm respectfully.

[0032] FIGS. 3A-3C are movement graphs showing examples of triggering during rehabilitative exercises with each triggering method according to an example.

[0033] FIGS. 4A-4B are flow charts showing methods for conditioning a subject based on a pattern of activity according to some implementations.

[0034] FIG. 5A is a flow chart showing an illustrative example of a method of selecting a pattern of activity, capturing activity data, preprocessing the activity data, and performing a triggering method on the preprocessed activity data to provide a therapeutic stimulation according to some implementations .

[0035] FIG. 5B is a flow chart showing an illustrative example of a method of streaming sensor data, preprocessing the sensor data, determining when a minimum inter-stimulation interval (1S1) has occurred, and triggering a therapeutic stimulation according to some implementations.

[0036] FIG. 6A shows a schematic depicting a process of using a camera sensor and computer vision to convert video data into continuous activity (movement / position data) according to some implementations.

[0037] FIG. 6B is a flow chart showing an illustrative example of a method of visually monitoring motions by a subject, streaming sensor data based on the subject motions, preprocessing the sensor data, determining when a minimum ISI has occurred, and pairing triggering a therapeutic stimulation with peak sensor input according to some implementations.

[0038] FIG. 7A shows a graph including a number of sessions with paired triggers with each triggering method according to an example.

[0039] FIG. 7B shows a graph including an increase in paired movement using the dynamic algorithm as compared to manual stimulation according to an example.

[0040] FIG. 7C shows a graph including a triggering rate in stimulations per minute with each triggering method according to an example.

[0041] FIG. 7D shows a graph including a percent of maximum sensor input during exercise with each triggering method according to an example.

[0042] FIG. 8 shows a graph including a triggering rate as a function of percent of max movement according to an example.

[0043] FIG. 9 shows a graph including a schematic depicting a subject’s activity above a progression threshold and antagonist activity below an antagonist threshold.

[0044] FIG. 10 shows a schematic depicting a successful trigger request according to an example.

[0045] FIG. 1 1 shows a schematic depicting a failed trigger request due to an antagonistic pattern of activity according to an example.

[0046] FIG. 12 shows a schematic depicting a successful trigger request followed by failed trigger due to 5 second timeout according to an example.

[0047] FIG. 13 shows a perspective view of examples of controllers with distinct handles to isolate specific neural pathways for motor and sensory functions used by the rehabilitation system according to some implementations.DETAILED DESCRIPTION

[0048] While various embodiments of the present disclosure have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the scope of the present disclosure. It should be understood that various alternatives to the embodiments of the present disclosure described herein may be employed.

[0049] Whenever the term “at least,” “greater than,” or “greater than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “at least,” “greater than” or “greater than or equal to” applies to each of the numerical values in that series of numerical values. For example, greater than or equal to 1 , 2, or 3 is equivalent to greater than or equal to 1 , greater than or equal to 2, or greater than or equal to 3.

[0050] Whenever the term “no more than,” “less than,” or “less than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “no more than,” “less than,” or “less than or equal to” applies to each of the numerical values in that series of numerical values. For example, less than or equal to 3, 2, or 1 is equivalent to less than or equal to 3, less than or equal to 2, or less than or equal to 1.

[0051] Certain inventive embodiments herein contemplate numerical ranges. When ranges are present, the ranges include the range endpoints. Additionally, every sub range and value within the range is present as if explicitly written out. The term “about” or “approximately” may mean within an acceptable error range for the particular value, which will depend in part on how the value is measured or determined, e.g., the limitations of the rehabilitation system. For example, “about” may mean within 1 or more than 1 standard deviation, per the practice in the art. Alternatively, “about” may mean a range of up to 20%, up to 10%, up to 5%, or up to 1% of a given value. Where particular values are described in the application and claims, unless otherwise stated the term “about” meaning within an acceptable error range for the particular value may be assumed. Throughout this application, the term “about” is used to indicate that a value may include the inherent variation of error for the device, the method being employed to determine the value, or the variation that exists among study subjects.

[0052] As used herein the specification, “a” or “an” may mean one or more. As used herein in the claim(s), when used in conjunction with the word “comprising,” the words “a” or “an” may mean one or more than one.

[0053] The use of the term “or” in the claims is used to mean “and / or” unless explicitly indicated to refer to alternatives only or the alternatives are mutually exclusive, although the disclosure supports a definition that refers to only alternatives and “and / or.” As used herein “another” may mean at least a second or more.

[0054] As used herein, the term “lesion” refers to any pathological or traumatic discontinuity of tissue or loss of function of a part thereof. For example, lesions include any injury associated with the spinal cord, for example, but not limited to contusions, compression injuries, etc.

[0055] The terms "administer", "administering", "administration", and the like, as used herein, refer to the methods that are used to enable delivery of agents or compositions to the desired site of biological action. In particular embodiments, administering refers to the delivery of an electrical impulse to the vagus nerve.

[0056] The terms "treat", "treating" or "treatment", and other grammatical equivalents as used herein, include alleviating, inhibiting or reducing symptoms, reducing or inhibiting severity of, reducing incidence of, prophylactic treatment of reducing or inhibiting recurrence of, preventing, delaying onset of, delaying recurrence of, abating or ameliorating a disease or condition symptoms, ameliorating the underlying metabolic causes of symptoms, inhibiting the disease or condition, e.g., arresting the development of the disease or condition, relieving the disease or condition, causing regression of the disease or condition, relieving a condition caused by the disease or condition, or stopping the symptoms of the disease or condition. The terms further include achieving a therapeutic benefit. By therapeutic benefit is meant eradication or amelioration of the underlying disorder being treated, and / or the eradication or amelioration of one or more of the physiological symptoms associated with the underlying disorder such that an improvement is observed in the individual.

[0057] The term “dysfunction” is defined as loss of motor function or sensory function as a result of a neurological injury or condition. The inability to perform motor or sensory tasks that were previously performed in daily life.

[0058] As used herein, “antagonistic or compensatory pattern of activity” refers to activation of nerves or muscles that elicit activity that are not the target of the therapy or inhibit progression of the therapy. A compensatory movement may be non-targeted motor activity or signal performed to accomplish the rehab task. In an aspect, the compensatory movement should not be reinforced or paired with the stimulation or VNS.

[0059] A “task” or “rehabilitative exercise” refers to any task intentionally paired with stimulation to recover lost functions.

[0060] As used herein, the term “sensor” refers to a device which detects or measures a property and produces an output signal. Exemplary sensors include neural electrodes including implantable and non-implantable electrodes such as an electroencephalogram (EEG) system, muscular electrodes such as for detecting an electromyogram (EMG), as well as movement or motion detectors such as a camera, accelerometers (IMG), and body worn sensors.

[0061] In exemplary embodiments, the systems and methods for treating dysfunction provided herein encompass treating dysfunction by administering a therapeutically effective amount of VNS based on performance of a game interaction, rehabilitative exercise, or activities of daily living. Embodiments of the present disclosure provide methods for treating an individual having a dysfunction, such as a motor disability or sensory disability and upper extremity motor deficits associated with chronic ischemic stroke. Examples of a dysfunction include a motor sensory dysfunction, a result of a stroke, multiple sclerosis, surgical procedures, traumatic brain injuries, PTSD, aphasia, cancer, a result of a peripheral nerve or limb injury, a result of a spinal cord injury (SCI), a walking dysfunction, or ankle dorsiflexion. The dysfunction may comprise but is not limited to neurological conditions, motor disability or sensory disability, including dysfunction at the ankle, knee, hip, thumb, finger / fingers, wrist, elbow, shoulder or trunk. In an example, the dysfunction may be a walking dysfunction, such as ankle dorsiflexion. In an example, the dysfunction may include difficulty in swallowing or speech.

[0062] In an aspect, VNS may be administered in an amount effective to ameliorate, eliminate or prevent one or more symptoms of the dysfunction, such as motor deficits. As used herein, “one or more symptoms” may include objectively measurable parameters, such as improvement in the subject’s motor and sensory function, and subjectively measurable parameters, such as subject well-being, subject perception of improvement in motor and sensory function, perception of lessening of pain or discomfort associated with the dysfunction or ischemic stroke.

[0063] In an exemplary embodiment, a rehabilitation system may include a computer or control system, a rehabilitation application (also referred to as the “RePlay App”) operating on the control system, and one or more controllers in conjunction with a neurostimulator in communication with the control system. In an example, the RePlay App or control system may include a game system for interfacing with a subject interface configured to sense a subject’s activity. In an example, the RePlay App or control system may include an exercise interface for performing an exercise motion invoking volitional movement.

[0064] The methods of the present disclosure may rely on modulated electrical stimulation or neurostimulation of the vagus nerve. Such electrical stimulation can be achieved by a variety ofdifferent methods known in the art. In an aspect, the neurostimulator is configured to provide VNS to a subject in coordination with the RePlay App or control system. In an exemplary embodiment, the rehabilitation system may include a neurostimulator configured to provide neurostimulation to one or more nerves, such as the subject’s vagus nerve. The rehabilitation system may include either an implantable neurostimulator or external neurostimulator.

[0065] The present methods can be guided by a therapist or performed by a subject going through daily activities or identified tasks. In particular aspects, the present methods are subject- guided, such as in a home setting. Parameters that are measured (e.g., software monitoring of game play) can include but are not limited to force, range of motion, torque, and angular velocity over a session. In an aspect, the task may be adjusted to increase measured inputs. Stimulation will occur when the measured input is in a top predetermined percentile of a previous buffer of trials.

[0066] In some implementations, the rehabilitation system may include one or more controllers with distinct handles to isolate specific neural pathways for motor and sensory functions. In an example, the controllers may include handheld sensors, a touchscreen, a keyboard or other types of user interfaces or objects. (See FIG. 13). In an example, the controllers may use isolating sensor input types to control the game. In an example, the controllers may use an isometric sensor input to control the game. The isometric input measures force or torque, with little or no movement. In an example, the controllers may use an isotonic sensor input to control the game. The isotonic devices measure torque without a change in angle of the device. In an aspect, the isotonic devices are configured to measure rotation or movement without little or no force.

[0067] In an aspect, the controllers can be used to monitor quantitative performance such as peak force, torque, angular rotation, velocity, acceleration, electromyograms, and angular velocity. Examples of measurements from sensors on the controllers include rotation angle (°), pressing and pinching force (g), movement distance, and 3-axis accelerometer, gyrometer, magnetometer. Examples of measurements from a keyboard include typing speed, keystroke pressure, and finger spread. Examples of measurements from a touch screen include speed of finger movements across the screen, pressure, number of digits on the screen, spreading of two or more digits.

[0068] In an example, a controller may include a FitMi handheld motion controller (Flint Rehab, California) to perform rehabilitative exercises while playing games on an Android tablet. The FitMi controller is a rubberized puck including several sensors such as a 3-axis accelerometer / gyrometer, magnetometer, and a force sensor. In an example, a game may further use a tablet’s touchscreen or a physical keyboard for gameplay in addition to the FitMi controller.

[0069] In an aspect, the controllers can be used to mimic a subject’s tasks or rehabilitative exercise. Examples of tasks or rehabilitative exercise may include: Reach and grasp; Lift objectsfrom table; Circumduction and bimanual tasks (mainly involving wrist and distal joints); Stacking objects; Slide credit card in slot; Turning on and off light switch; Squeezing objects; Writing; Typing; Stirring liquid in a bowl (bimanual); Dial a cell phone (bimanual); Fold towels or clothes (bimanual); Wear a belt; Tying shoelaces; Eating; Brushing teeth; Combing hair. For a wrist flexion, a camera may be used to model the movement as a wire frame (e.g., bones with joints) and compare the movement to past attempts and to optimal (e.g., normal) movement in order to find the best movements that the subject can generate. Movements, such as walking, grasping or tying, may be quantified as location, direction, speed, and angle of each joint as a function of time. For speech production, vocalizations might be compared to previous sounds and normal speech sounds produced by others. Vocal movements might be quantified based on the intensity, duration, pitch, formant structure (vowels), formant transitions (consonants), voiceonset time, and other standard methods of quantifying speech sounds. Some examples of typical motor therapies may be actions such as: squeezing a dynamometer, turning on / off a light switch, using a lock and key, opening and closing a door by twisting or depressing different doorknobs, flipping cards, coins and other objects over, placing light and heavy objects at different heights, moving pegs to hole and remove pegs from hole, lifting a shopping basket / briefcase, drawing geometric shapes, dressing, typing, reaching and grasping light and heavy objects, grasping and lifting different (size, shape, and texture) objects, doing a precision grasp, writing, drawing connect the dots, opening and closing a jar or medication bottle, lifting an empty and full cup / glass, using feeding utensils, cutting food, stirring liquids, scooping, pouring a glass of water with the paretic hand; or using the paretic hand to stabilize the glass and pouring with the good hand, picking an object and bring to target, using a spray can, cutting with scissors, or brushing teeth / hair.

[0070] In an exemplary embodiment, a method 100 of delivering VNS concurrent with a rehabilitative activity may include a step 110 of activity selection, a step 120 of activity capture, and a step 130 of selective triggering. (See FIG. 1A). In an exemplary embodiment, the activity selection may include a neural circuit selection and the activity capture may include a neural activity capture. In an exemplary embodiment, the activity selection may include a movement selection and the activity capture may include a movement capture.

[0071] According to some implementations, the step 110 of activity selection may include a subject selecting a rehabilitation game having at least one associated exercise or movement in a rehabilitation application (also referred to as “RePlay”) to isolate an activity signal. In an example, the activity signal may be isolated in a dimension of interest. In an example, a subject may select a Traffic Racer game 112 controlled with a range of motion handle exercise 114. In an aspect, the subject’s activity may include a motion of interacting with the controller, handle,or one or more muscles nerves, tendons, joints evoked with creating the motion. According to some implementations, the step 110 of activity selection may include establishing a set of task outcomes for at least one task. In an example, the set of task outcomes may include one or more patterns of progressive activity, a progression threshold, and / or one or more patterns of antagonist activity.

[0072] According to some implementations, the step 120 of activity capture may include performing preprocessing to captured activity data. In some implementations, the step of activity capture may include monitoring or capture activity data from multiple sensors simultaneously to determine when a task has been completed. In an aspect, preprocessing to captured activity data from multiple sensors may include prioritizing a subset of sensors over others.

[0073] In some implementations, the RePlay App is configured to perform preprocessing to captured activity data prior to selective triggering or applying a triggering method to sensed activity. In some implementations, the RePlay App is configured to monitor controller activity while the subject interacts with a rehabilitation system, and to continuously preprocess to extract a movement rate of change and to apply selective triggering or a triggering method. In some implementations, each type of controller may have its own preprocessing method. In an example, controllers may be programmed to detect a sequence of the subject’s activities such as wrist flexion followed by wrist pronation and then finger flexion or any order or combinations of activities. In an aspect, the controller may include one or more sensors configured to detect the sequence of the subject’s activities and provide a single output related to a task outcome.

[0074] According to some implementations, the rehabilitation game may include unidirectional games 112a and bidirectional games 112b. Examples of unidirectional games 112a may include but are not limited to a space runner game operated by a touch controller, a fruit ninja game operated by a swipe motion controller, and a typer shark game operated by a button or keyboard controller. Examples of bidirectional games 112b may include but are not limited to a breakout game using a controller configured to detect rolling, a fruit archery game using a controller configured to detect supination, and a traffic racer game using a controller configured to detect rotation.

[0075] According to some implementations, the step 120 of activity capture may include monitoring a subject’s activity performing the at least one task. According to some implementations, the step 120 of activity capture may include comparing the subject’s monitored activity to the set of task outcomes, the progression threshold, and the one or more patterns of antagonist activity. According to some implementations, the step 120 of activity capture may include capture of movements in absolute and / or relative degrees 122. In some embodiments, the movement capture may include capture of pronation and supination movements in absoluteand / or relative degrees during the game and / or exercise or rehabilitative exercise. In an aspect, pronation movement may be differentiated into simultaneous motions. In an example, a pronation movement of an ankle may include subtalar eversion, ankle dorsiflexion, and forefoot abduction.

[0076] According to some implementations, step 130 of selective triggering may include a step of determining an activity magnitude and applying a triggering method to the calculated movement magnitude. According to some implementations, the triggering method may include unidirectional triggering 132 and bidirectional triggering 135. According to some implementations, step 130 of selective triggering may include triggering a stimulation of a nerve when performance of the at least one task meets the progression threshold, and does not involve the one or more patterns of antagonist activity.

[0077] According to some implementations, the step 130 of selective triggering may include a stimulation of VNS therapy. VNS therapy is premised on the timing of VNS concurrent with a pattern of activity such as active neural pathways controlling muscles during a game movement or rehabilitative exercises. In an aspect, the pattern of activity is driven by engagement of motor networks in the central nervous system, and the concurrent VNS generates a rapid release of neuromodulators that facilitate synaptic plasticity in the active motor networks. Consequently, manipulations that degrade the timing between VNS and the occurrence of the target movement reduces the efficacy of this approach. Studies in animal models show that delaying VNS until after training results in significantly less recovery.

[0078] Moreover, even VNS delivered during rehabilitative exercises fails to be effective if it is not delivered concurrent with a progressive activity or movement or target of rehabilitative exercises. In an aspect, stimulation may be withheld when the progressive activity or movement occurs contemporaneously with antagonistic activity and or compensatory activity.

[0079] In an example, a best movement may be determined by an activation of pattern of activity. In an example, a best movement may be determined by an activation of one or more muscles to perform a specific task. In an example, a best movement may be determined by at least one modification of at least one associated exercise or movement.

[0080] In an aspect, the rehabilitation system is configured to provide precise timing of stimulation during specific subject’s activity patterns to maximize the benefits of the conditioning. In an aspect, a successful trigger request requires a minimum inter-stimulation interval. In an example, the inter-stimulation interval requires at least 5 seconds before allowing a selective trigger to occur. In an example, the inter-stimulation interval is between 5 and 8 seconds. In an example, the stimulation is configured to occur within about 2 seconds of completion of a completed task. In an example, the stimulation is configured to occur upondetection of the pattern of activity, force, range of motion in a top about 90% detected subject’s activity in a buffer of time (e.g., few minutes).

[0081] Clinical evidence also supports the importance of timing VNS with the target movements. Whereas VNS delivered by a therapist during movements enhances recovery of upper limb function, additional VNS may not provide further benefits when delivered during unsupervised exercises where stimulation did not explicitly coincide with movement.

[0082] Although the above steps show method 100 of conditioning a subject in accordance with many embodiments, a person of ordinary skill in the art will recognize many variations based on the teaching described herein. The steps may be completed in a different order. Steps may be added or deleted. Some of the steps may comprise sub-steps. Many of the steps may be repeated as often as beneficial to the method(s).

[0083] Turning to FIG. IB, examples of unidirectional games 112a, unidirectional movements and user interfaces 114a, corresponding movement signals 122a, triggering methods 124a, and therapy triggers 126 are shown according to some implementations. In an aspect, unidirectional triggering may be preferred when rehabilitation is focused on recovering range of motion or strength in a single dimension. In an example, unidirectional triggering may be set to positive only or negative only. By applying selective triggering 133 with the positive only setting, the trigger method is configured to produce triggers 134 when the subject’s activity surpasses the progression threshold for the given task (e.g., 95th percentile of supination movements).

[0084] Turning to FIG. 1C, examples of bidirectional games 112b, bidirectional movements and user interfaces 114b, corresponding activity signals 122b, triggering methods 124a-b, and therapy triggers 126 are shown according to some implementations. In an aspect, bidirectional triggering may be preferred when targeting general increases in range of motion or strength. The bidirectional triggering may provide flexibility to handle VNS timing in subjects with unbalanced deficits. In an example, by applying selective triggering 133 with the bidirectional triggering, the trigger method produces therapy triggers 126 when the subject’s activity surpasses either the progression threshold (e.g., 95th percentile) of either the positive threshold 133 or negative threshold 137. In an example, the positive threshold 133 or negative threshold 137 may be assigned to individual aspects of supination or pronation movements.

[0085] According to some implementations, the rehabilitation system may include game-based activity, movement, and repetition-based exercises. FIG. ID shows examples of game-based exercises and corresponding movement signals according to some implementations. In an example, a game-based movement may include a touch motion 132a, a swipe motion 132b, and a rotation motion 132c. FIGs. 1E-1F show examples of repetition-based exercises and corresponding movement signals according to some implementations. In an example, arehabilitative exercise may include a curling motion 142a, a shoulder abduction 142b, and a reach across motion 142c. FIG. IE shows an example of corresponding movement signals after processing by a static threshold algorithm according to some implementations. In an example, a triggering method 124 may be applied to FIG. IF shows an examplary corresponding movement signals after processing by a dynamic algorithm according to some implementations.

[0086] In an aspect, the RePlay App may be configured to continuously preprocess activity signals to extract one or more principal components from the subject’s activity. In an aspect, the principal component analysis may include spatial and / or temporal patterns of neural circuits. In an aspect, the principal component analysis may include spatial and / or temporal patterns of muscular components and movements. In an example, the subject’s activity may be differentiated in somatosensory and mechanosensory components. In an example, the subject’s activity may be differentiated in the subject’s spared neural network from their damaged neural network.

[0087] According to some implementations, movement signals may be captured in 3D / 4D space with timing components tracked. In an aspect, the movement signals may be differentiated by principal component analysis revealing attributes in displacement, velocity, acceleration, throughout the subject’s movements (e.g., flexion and pronation). In an aspect, the rehabilitative exercise may be differentiated by one or more of changes in direction, intermittent differences in range, and intermittent differences in pauses. In an aspect, the principal component analysis may include anatomical features of the subject such as their body composition, muscles, joints, tendons, flexibility, range of motion among other subject-specific factors such as missing digits or severity of condition.

[0088] Turning to FIG. 5A, an illustrative example is shown of a method 500 of using the rehabilitation system to treat a subject. In an example, the method may include a step 510 of preprocessing activity data and a step 520 of performing a trigger method on the preprocessed activity data according to some implementations.

[0089] In an example, preprocessing activity data may include selecting a pattern of activity, capturing a subject’s activity, and smoothing the captured activity. In an example, preprocessing activity data may include filtering activity data based on velocity or acceleration, performing a Fast Fourier Transform (FFT) to identify frequency components in the activity data signal. In an aspect, smoothing the captured activity may include applying a moving average to continuous data stream. In an aspect, smoothing the captured activity may include removing outliers or noise from the continuous data stream. In an example, the moving average may smooth collected data in a predetermined time. In an example, the predetermined time may be about 300 ms. According to some implementations, the predetermined time may be on the order of minutes, hours, days, and weeks. In an aspect, the predetermined time may be based on a previous session. In anexample, the moving average may smooth collected data based on a rate of change of a motion associated with the movement selected. In an aspect, smoothing the captured activity may result in calculating a magnitude of movement samples averaged over the predetermined time. In an aspect, smoothing the captured activity may include digitizing the continuous stream of analog activity data.

[0090] In an example, the step of performing a trigger method on the preprocessed activity data may include comparing a current activity with a buffer of previous activity samples and triggering based on the comparison when a triggering threshold has been met. In an example, the buffer may include up to around 3000 of the last activity samples. In an aspect, the buffer is designed to detect improvements and continuously adjust difficulty to push progressive recovery. If the buffer is too short, triggering may occur constantly. If the buffer is too long, triggering may adjust the progressive threshold much slower.

[0091] Turning to FIG. 5B, an illustrative example is shown a method 530 of using the rehabilitation system to treat a subject. In an example, the method may include a step 531 of having a subject interface with a controller, a step 532 of streaming sensor data, a step 533 of preprocessing the sensor data, a step 534 of calculating a magnitude of activity samples averaged over a predetermined time, a step 535 of comparing a current activity with a buffer of previous activity samples, a step 536 of determining when a minimum ISI has occurred, and a step 537 of triggering a therapeutic stimulation according to some implementations.

[0092] According to some implementations, the rehabilitation system may include one or more sensors configured to detect motion. Turning to FIGs. 6A-6B, illustrative examples are shown of a method 600, 602 of visually monitoring motions by a subject and triggering a therapeutic stimulation based on the visually monitored motions.

[0093] In an example, the rehabilitation system may include a camera sensor configured to provide computer vision to convert video data into continuous activity / position data and store a buffer of the activity / position data. (See FIG. 6A). In an example, the method 600 may include a step of selecting above average events and triggering a therapeutic stimulation. FIG. 6B shows an illustrative example of a method of visually monitoring motions by a subject, streaming sensor data based on the subject’s motions, preprocessing the sensor data, determining when a minimum ISI has occurred, and pairing triggering a therapeutic stimulation with peak sensor input. Raw video obtained from a camera may be converted into one or more metrics of activity or performance. In an example, metrics of activity may include neural data and motion data such as joint angle, speed, and variability. For example, FIGs. 6A-6B show a video camera signal being converted into a game-associated activity signal.

[0094] In some aspects, the present methods comprise providing a closed-loop triggering method capable of selective triggering of neurostimulation or VNS based on certain parameters being met such as a progressive pattern of activity (e.g., best sensor inputs) and therefore may increase a probability of a successful response, allowing for efficacious therapy beyond the clinic. The RePlay App continuously monitors input measures while the subject interacts with the programming (e.g., game or exercise). A triggering method is used to monitor a buffer of performance data (e.g., peak force, torque, angular rotation, velocity, acceleration, electromyograms, and or angular velocity) and stimulate on a top predetermined percent of metrics measured over the past N trials when the subject hits a target. The goal is to reward the ‘best’ neural activity driven sensor input measured during the activity or game. Stimulation occurs after a target is hit if they reached the ‘best’ criteria. In an example, stimulation is configured to occur upon completion of progressive activity or success (e.g., after neural activity ends for a given input) not at an onset or beginning of an activity, movement, or sensor input. In an example, stimulation is limited to no more than once every 5 seconds.

[0095] In some implementations, a successful pattern of activity may be defined as a series of sensor inputs in a particular sequence demonstrating a progression in therapeutic or rehabilitative status. For example, a subject performing activities of daily living may reach out with their arm, extend their wrist, pronate their wrist, and flex their fingers to grab an object. If that sequence is performed in order, it may reach a progression threshold and may result in a trigger request. In some implementations, a successful pattern of activity may include a sequence of activities as part of sensory feedback.

[0096] In some implementations, the pattern of activity may include neural activity detected by invasive or non-invasive electrodes. In an example, neural activity may include a series of neural signals independent of muscular activity. In an example, neural activity may include a series of neural signals in synchrony with muscular activity. In an example, neural activity may include EEG, intensity of brain response as determined by Functional near-infrared spectroscopy (fNIRS), Auditory evoked potentials (AEPs), or other types of sensors or transducers receptive to a subject’s nerves. In an example, the pattern of activity may be neural circuit activation of the subject as sensed by EEG or other neural electrodes. In some implementations, a pattern of activity may include muscular activity such as EMG or other types of sensors or transducers receptive to a subject’s muscles.

[0097] In particular aspects, the stimulation trigger does not occur faster than once every 5 seconds. In particular aspects, the stimulation trigger may be based on an outcome of a task, task outcome, or activity outcome, such as after successful completion of task (e.g., a video game task), hitting a goal of the top predetermined percent of the goal based on past performance overlast more than 5 seconds, and / or based on sensor data (e.g., peak force, peak range of motion, peak acceleration, peak velocity, and / or EMG activity patterns). The stimulation may be prevented based on compensatory activity (e.g., upper body moves, forearm activity, or activity of antagonistic or compensatory muscles etc.), features of the sensed activity (e.g., tremor, spasticity, velocity, and / or acceleration), detection of electrical activity of antagonistic or compensatory nerves / muscles (not movement / force), onset of movement, or failed trial in game or task. In some aspects, the stimulation trigger is paired with or just after functional electrical stimulation of muscles to cause movement. In particular aspects, one of more of these parameters needs to be met in order for the trigger request for stimulation to occur. Thus, the stimulation trigger is not limited to movement (e.g., the onset of movement), but and or on a set of criteria based on sensor input or a series of sensor inputs plus the task outcome or goal of the subject’s task.

[0098] In particular aspects, the present methods comprise withholding selective triggering upon detection of antagonistic or compensatory patterns of activity beyond an antagonist threshold. In an example, an antagonistic pattern of activity may include activation of nerves or muscles that elicit activity that are not the target of the therapy or inhibit progression of the therapy. In an aspect, a compensatory pattern of activity may be specific to an ailment or dysfunction of the subject. In an example, a subject may compensate for lack of a grip or pronation with a shoulder movement. In an example, a subject may compensate with torso movement to move their entire arm. In an example, a subject may compensate with tenodesis to grip an object by using forearm movement to push the hand or fingers.

[0099] In some implementations, a decrease in an antagonistic or compensatory pattern of activity, even beyond an antagonist threshold, may be considered a progressive activity. In an example, the progressive activity can be prioritized over the antagonistic or compensatory pattern of activity.

[0100] In an example, electrical activity of antagonistic muscles may be measured by EMG activity, such as a sleeve measuring EMG activity in the hand or arm. Thus, if there is an antagonistic or compensatory activity detected, such as by EMG activity or movement detected by a body worn accelerometer, then the stimulation trigger is not performed. For example, a trigger would occur for activation of muscle a or b but not for muscle c. If muscles a, b, and c were active, stimulation trigger would not occur. Other methods of detecting antagonistic activity may comprise EEG signals, accelerometer signals, acoustic signals, peripheral nerve signals, cranial nerve signals, cerebral or cerebellar cortical signals, signals from basal ganglia, cameras (e.g., motion capture cameras), and signals from other brain or spinal cord structures. Generally,any cell phone, tablet, or USB camera may be used for detecting movement, including antagonistic movement.

[0101] In particular aspects, the present methods may not comprise stimulation at a beginning or following a start of a movement. The present methods can provide adaptive stimulation based on past performance. Thus, the stimulation can be provided just following the “best” event or series of events detected. Continuous improvement is then rewarded based on quantitative measures of function. By continuously monitoring quantitative measures, stimulation is configured to occur when a quantitative measure in the top predetermined percent of the previous measurements over a given time period. Specifically, there is no selection of therapeutic tasks based on the therapeutic goals. Thus, the tasks can be adaptive, not repetitive or simply activities of daily living.

[0102] In an exemplary embodiment, a method of delivering neuromodulation concurrent with rehabilitative exercises may include invoking long-term changes in synaptic efficacy by release of neuromodulators in presence of eligibility traces. Eligibility traces are a mechanism by which neural circuits can be potentiated or depressed based on an outcome of an event or reward. In an aspect, eligibility traces may be formed by Hebbian processes resulting from spike timing dependent changes. In an example, release of neuromodulators within about 10 seconds of the neural activity may invoke long-term potentiation or depression based on the spike timing. In an example, neurostimulation of the vagus nerve at the end of motor, sensory, proprioceptive or cognitive tasks may be used to drive long-term potentiation and depression of neural circuits to enhance functional recovery.

[0103] Turning to FIG. 4A, a method of conditioning a subject based on a pattern of activity is shown according to some implementations. In some implementations, a conditioning method 400 is provided for treating a subject having a dysfunction comprising a step 402 of establishing a set of task outcomes including one or more patterns of progressive activity, a threshold parameter for success or progression threshold, and one or more patterns of antagonist activity, a step 404 of monitoring the subject’s activity performing at least one task, and a step 406 comparing the subject’s activity to the set of task outcomes, the progression threshold, and the one or more patterns of antagonist activity. In some implementations, the method may include a step 408 of determining when a performance of the at least one task meets the progression threshold. In an example, when the performance of the at least one task meets the threshold parameter (Y), the method advances to step 410 of determining when a performance of the at least one task involves one or more patterns of antagonist activity. In an example, when a performance of the at least one task does not involve one or more patterns of antagonist activity (N), the method advances to astep 412 of triggering a nerve stimulation. In an example, the nerve stimulation does not occur at onset of movement.

[0104] Turning to FIG. 4B, a method of conditioning a subject based on a pattern of activity is shown according to some implementations. In some implementations, a conditioning method 420 is provided to treat a subject, the method comprising a step 422 of selecting a task including at least one associated activity, at least one task outcome, one or more patterns of antagonist activity, and a progression threshold, a step 424 of capturing a subject performing the task, a step 426 of determining when performance of the task meets the progression threshold, a step 428 of stimulating the subject’s nerve when the performance of the task meets the progression threshold and does not involve the one or more patterns of antagonist activity.

[0105] In an example, the predetermined progressive activity task may be a game interaction, therapeutically effective amount of VNS based on performance of a game interaction, rehabilitative exercise, or activities of daily living. In an example, activity may be determined by a force measured in grams as sensed by a controller or sensor.

[0106] Although the above steps show methods 400 and 420 of conditioning a subject in accordance with many embodiments, a person of ordinary skill in the art will recognize many variations based on the teaching described herein. The steps may be completed in a different order. Steps may be added or deleted. Some of the steps may comprise sub-steps. Many of the steps may be repeated as often as beneficial to the method(s).

[0107] The following examples are included to demonstrate preferred embodiments of the disclosure. It should be appreciated by those of skill in the art that the techniques disclosed in the examples which follow represent techniques discovered by the inventor to function well in the practice of the disclosure, and thus can be considered to constitute preferred modes for its practice. However, those of skill in the art should, in light of the present disclosure, appreciate that many changes can be made in the specific embodiments which are disclosed and still obtain a like or similar result without departing from the spirit and scope of the disclosure.

[0108] In an aspect, a successful trigger request requires several conditions to occur. Turning to FIG. 10, a schematic depicting a successful trigger request 1000 is shown applying the methods discussed. In an aspect, a successful trigger request requires a minimum inter-stimulation interval 1010. In an example, the inter-stimulation interval requires at least 5 seconds before allowing a selective trigger to occur. In an example, the inter-stimulation interval is between 5 and 8 seconds. In an aspect, a successful trigger request requires a minimum level or pattern of activity 1020 which may be monitored on a continuous basis. In an aspect, a successful trigger request requires a minimum level or pattern of activity 1020 which may be monitored on a continuous basis.

[0109] In an example, the pattern of activity may be set an amount of force measure in grams on one or more controller interfaces. In an example, the pattern of activity may he set as a particular neural activity. In an example, the pattern of activity may have one or more progressive thresholds 1024 a-b relative to a baseline 1022.

[0110] In an example, the pattern of activity having been met may result in a positive activity indication 1026. In an aspect, a successful trigger request requires a predetermined activity outcome 1030 a-b. In an example, the predetermined activity outcome may be a pong game that requires the subject to move and maintain a paddle at a certain location to interface with a ball. When the paddle is maintained for the ball interaction, the predetermined activity outcome is a success 1030 a-b. In an example, the predetermined activity outcome may be based on any activity, game, controller, or input type. In an aspect, a successful trigger request requires absence of antagonistic or compensatory activity 1040. In an example, the antagonistic or compensatory activity may have one or more antagonist thresholds 1044 a-b relative to a baseline 1042. In an example, the antagonistic activity may include any activity that would counter act a progressive pattern of activity. For example, if a progressive pattern of activity would require isolation of neural or motor activity and yet the subject activated neural or motor activity beyond the isolation, this may constitute antagonistic or compensatory activity. Specific instances of subject activated neural or motor activity may be tailored to the subject’s dysfunction.

[0111] In an aspect, a successful trigger request may require a combination or synchrony 1050 of (i) the inter-stimulation interval 1010, (ii) minimum level or pattern of activity 1020, (iii) successful predetermined activity outcome 1030 a-b, and (iv) absence of antagonistic or compensatory activity 1040. In an aspect, the combination or synchrony 1050 of the successful trigger request results in triggering 1052 of the neurostimulation.

[0112] Turning to FIG. 11 , a schematic depicting a failed trigger request 1100 due to an antagonistic activation is shown. In an aspect, a successful trigger request requires a min and / or max inter- stimulation interval 1110. In an example, the minimum inter-stimulation interval requires at least 5 seconds before allowing a selective trigger to occur. In an aspect, a successful trigger request requires a minimum level or pattern of activity 1 120. In an example, the pattern of activity may be set an amount of force on one or more controller interfaces. In an example, the pattern of activity may be set as a particular neural activity. In an example, the pattern of activity may have one or more progressive thresholds 1124 a-b relative to a baseline 1122. In an example, the pattern of activity having been met may result in a positive activity indication 1126.

[0113] In an aspect, a successful trigger request requires a predetermined activity outcome 1 130 a-d. In an example, the predetermined activity outcome is a success 1030 a-d. In an aspect, a successful trigger request requires absence of antagonistic or compensatory activity 1140. In anexample, the antagonistic or compensatory activity may have one or more antagonist thresholds 1 144 a-b relative to a baseline 1 142. Tn an example, the antagonistic activation may be a muscle activation of the subject as sensed directly by a sensor, controller, or camera. In an example, the antagonistic activation may be a muscle activation of the subject as sensed indirectly by inferring from the monitored activity what neurological or muscular components were activated to achieve the activity. In an example, the antagonistic activation may be neural circuit activation of the subject as sensed by EEG or other neural electrodes.

[0114] In an aspect, a successful trigger request requires a combination or synchrony 1150 of (i) the inter-stimulation interval 1110, (ii) minimum level or pattern of activity 1120, (iii) successful predetermined activity outcome 1130 a-d, and (i v) absence of antagonistic or compensatory activity 1140. As shown here, the combination or synchrony 1150 of the successful trigger request does not result in triggering 1152 of the neurostimulation.

[0115] Turning to FIG. 12, a schematic depicting a successful trigger request 1200a followed by a failed trigger request 1200b is shown. In an aspect, a successful trigger request requires a minimum inter-stimulation interval 1210. In an aspect, a successful trigger request requires a minimum level or pattern of activity 1220. In an example, the pattern of activity may have one or more progressive thresholds 1224 a-b relative to a baseline 1222. In an example, the pattern of activity having been met may result in a positive activity indication 1226. Here, positive activity indication 1226a has an inter-stimulation interval less than at least 5 seconds than positive activity indication 1226b.

[0116] In an aspect, a successful trigger request requires a predetermined activity outcome 1230 a-c. In an example, the predetermined activity outcome is a success 1230 a-c. In an aspect, a successful trigger request requires absence of antagonistic or compensatory activity 1240. In an example, the antagonistic or compensatory activity may have one or more antagonist thresholds 1244 a-b relative to a baseline 1242. In an aspect, a successful trigger request requires a combination or synchrony 1250a-b of (i) the minimum inter-stimulation interval 1210, (ii) minimum level or pattern of activity 1220, (iii) successful predetermined activity outcome 1230 a-c, and (iv) absence of antagonistic or compensatory activity 1240. As shown here, the combination or synchrony 1250b of trigger request 1252b does not result in triggering 1252b of neurostimulation due to the minimum inter- stimulation interval 1210 not having been met from the combination or synchrony 1250a of the successful trigger request 1252a.

[0117] According to some implementations, when an extended inter-stimulation interval occurs (e.g., no stimulation has been provided in about 5 minutes), the methods described may adjust emphasis on one or more steps to allow a greater likelihood for triggering. For example, when an extended inter-stimulation interval occurs, either of the one or more progressive thresholds andthe one or more antagonist thresholds may vary to allow a greater likelihood for triggering. In an example, the one or more progressive thresholds associated with an activity may be reduced to allow a greater likelihood for triggering. In an example, an activity may require a combination of a task or activity that may be separated to encourage stimulation. In an example, an arm reach out with their arm, extend their wrist, pronate their wrist, and flex their fingers to grab an object may be separated into different tasks. In an example, the one or more antagonist thresholds associated with an activity may be raised to allow a greater likelihood for triggering. According to some implementations, a different triggering method may be applied to allow a greater likelihood for triggering. In an example, a bilateral triggering may replace a unilateral triggering method.

[0118] In an aspect, trigger methods may include different characteristics to achieve VNS triggering based on reliance on activity selection and inter-stimulation timing. While closed-loop VNS has been mostly limited to use in office supervised by a rehabilitative therapist, the dynamic algorithm may extend therapy at home. In a previous clinical trial, a therapist delivered VNS to subjects by pressing a button when the subject performed supervised in-office task-oriented rehabilitation during upper limb therapy. VNS produced a significant improvement in upper limb function. During a second phase of the same study, subjects received periodic VNS during unmonitored exercise sessions at home. This paradigm failed to increase upper limb recovery. Taken together, these findings show that VNS during supervised rehabilitation, where stimulation is delivered coincident with large movements is effective, whereas VNS during unsupervised exercises, in which stimulation is not explicitly paired, provides no additional benefits. Here, it was shown that periodic triggering during upper limb rehabilitation primarily produces stimulations during rest and only rarely during the largest movements. Based on this evidence, it is likely subjects did not improve with periodic triggering due to the temporal separation of task performance from the delivery of VNS. Thus, introducing the dynamic algorithm capable of triggering on specific sensor inputs may increase the probability of a successful response, allowing for efficacious therapy beyond the clinic.

[0119] Implementing methods to process unbalanced deficits can aide in handling bidirectional movements. In an aspect, a bidirectional movement may include any volitional control over muscles or tendons. Motor impairments after neurological injury commonly present as deficits that exist to a greater degree in one direction over another, such as a moderate impairment of forearm supination and severe impairment of forearm pronation in the same arm. These deficits appear as low levels of activity in one direction of the respective movement signal.

[0120] In an aspect, the dynamic algorithm can handle bilateral movements by maintaining two separate movement distributions, which provides automatic adjustment of two individualthresholds during bidirectional movements. Alternatively, a therapist could set the triggering method with a single adaptive threshold so that VNS is always paired with the clear sensor input in the weak direction. In an aspect, automatic consideration of bidirectional movement increases specificity of VNS pairings and does not rely on manual calculation of deficit imbalances.

[0121] In an aspect, the static threshold algorithm may compensate when different minimum activity thresholds are set for each direction, but extensive manual tuning would be needed.

[0122] In an aspect, the periodic algorithm may be unable to compensate for a deficit imbalance, which is helpful for handling bidirectional training. However, the periodic algorithm may be to identify clear movements while maintaining an effective triggering rate. (FIG. 2C). It is likely that more complex methods, such machine learning implementations, may be used for sensor input selection. In an aspect, the dynamic algorithm has minimal complexity to facilitate verification and validation testing for regulatory approval.

[0123] The triggering method can be used to pair VNS or other forms of neuromodulation with sensory and cognitive rehabilitation. For example, speech therapy designed to improve speech volume would use a microphone and filter the signal to extract the speech envelope so that the dynamic algorithm can trigger VNS when speech volume is above average. Speech therapy designed to help stroke subjects speak more quickly could the dynamic algorithm to trigger VNS by extracting out the rate of syllable level modulation. To promote recovery of sensory function, the dynamic algorithm could trigger VNS based on neural evoked potentials or behavioral reaction times. An algorithm could decode speech to text and trigger on correct speech responses or patterns.

[0124] Delivery of VNS during therapeutic exercises improves upper limb recovery in stroke subjects. In some aspects, the present methods provide an algorithm that can be used to pair VNS with the largest or best sensor inputs at a reliable interval to replicate and build on the manually paired VNS delivery paradigm that produced clinical benefits. Automatic closed-loop VNS can be achieved with an algorithm that dynamically modulates a minimum activity threshold based on previous sensor inputs or sequence of sensor inputs.

[0125] The vagus nerve (i.e., the tenth cranial nerve, paired left and right) is composed of motor and sensory fibers. The vagus nerve leaves the cranium, passes down the neck within the carotid sheath to the root of the neck, then passes to the chest and abdomen, where it contributes to the innervation of the viscera. A vagus nerve in a human consists of over 100,000 nerve fibers (i.e., axons), mostly organized into groups. The groups are contained within fascicles of varying sizes, which branch and converge along the nerve. Under normal physiological conditions, each fiber conducts electrical impulses only in one direction, which is defined to be the orthodromic direction, and which is opposite the antidromic direction. However, external electricalstimulation of the nerve may produce action potentials that propagate in orthodromic and antidromic directions. Besides efferent output fibers that convey signals to the various organs in the body from the central nervous system, the vagus nerve conveys sensory (afferent) information about the state of the body's organs back to the central nervous system. Some 80- 90% of the nerve fibers in the vagus nerve are afferent (sensory) nerves, communicating the state of the viscera to the central nervous system.

[0126] The largest nerve fibers within a left or right vagus nerve are approximately 20 pm in diameter and are heavily myelinated, whereas only the smallest nerve fibers of less than about 1 pm in diameter are completely unmyelinated. When the distal part of a nerve is electrically stimulated, a compound action potential may be recorded by an electrode located more proximally. A compound action potential contains several peaks or waves of activity that represent the summated response of multiple fibers having similar conduction velocities. The waves in a compound action potential represent different types of nerve fibers that are classified into corresponding functional categories, with approximate diameters as follows: A-alpha fibers (afferent or efferent fibers, 12-20 pm diameter), A-beta fibers (afferent or efferent fibers, 5-12 pm), A-gamma fibers (efferent fibers, 3-7 pm). A-delta fibers (afferent fibers, 2-5 pm), B fibers (1-3 pm) and C fibers (unmyelinated, 0.4-1.2 pm). The diameters of group A and group B fibers include the thickness of the myelin sheaths. It is well known that electrical stimulation activates the largest fibers at the lowest current levels assuming all other parameters are the same. Increasing current amplitude will activate A fibers first, B fibers second and finally C fibers.

[0127] The vagus (or vagal) afferent nerve fibers arise from cell bodies located in the vagal sensory ganglia. These ganglia take the form of swellings found in the cervical aspect of the vagus nerve just caudal to the skull. There are two such ganglia, termed the inferior and superior vagal ganglia. They are also called the nodose and jugular ganglia, respectively. The jugular (superior) ganglion is a small ganglion on the vagus nerve just as it passes through the jugular foramen at the base of the skull. The nodose (inferior) ganglion is a ganglion on the vagus nerve located in the height of the transverse process of the first cervical vertebra.

[0128] In an aspect, the neurostimulator can include one or more electrodes configures to excite or depress selected nerve fibers. Selected nerve fibers may be stimulated in different embodiments of methods that make use of the neurostimulator, including stimulation of the vagus nerve at a location in the subject’s neck. In an example, the neurostimulator can include a cuff electrode configured to be in contact with the subject’s vagus nerve. The vagus nerve at subject’s neck is situated within the carotid sheath, near the carotid artery and the interior jugular vein. The carotid sheath is located at the lateral boundary of the retopharyngeal space on each side of the neck and deep to the sternocleidomastoid muscle. The left vagus nerve is sometimesselected for stimulation because stimulation of the right vagus nerve may produce undesired effects on the heart, but depending on the application, the right vagus nerve or both right and left vagus nerves may be stimulated instead.

[0129] Electrical stimulation of a nerve involves the direct depolarization of axons. When electrical current passes through an electrode placed in close proximity to a nerve, the axons are depolarized, and electrical signals travel along the nerve fibers. The intensity of stimulation will determine what portion of the axons are activated. A low-intensity stimulation will activate those axons that are most sensitive, i.e., those having the lowest threshold for the generation of action potentials. A more intense stimulus will activate a greater percentage of the axons. Electrical stimulation may activate the largest fibers at the lowest current levels assuming all other parameters are the same. Increasing current amplitude will activate A fibers first, B fibers second and finally C fibers.

[0130] Many such therapeutic applications of electrical stimulation involve the surgical implantation of electrodes within a subject. In contrast, devices may be used to stimulate nerves by transmitting energy to nerves and tissue non-invasively.

[0131] The methods provided herein may comprise invasive (e.g., surgical implantation) or noninvasive (e.g., transcutaneous) devices. In particular aspects, noninvasive methods are used to administer VNS.

[0132] A medical procedure is defined as being non-invasive when no break in the skin (or other surface of the body, such as a wound bed) is created through use of the method, and when there is no contact with an internal body cavity beyond a body orifice (e.g., beyond the mouth or beyond the external auditory meatus of the ear). Such non-invasive procedures are distinguished from invasive procedures (including minimally invasive procedures) in that the invasive procedures insert a substance or device into or through the skin, or other surface of the body, such as a wound bed, or into an internal body cavity beyond a body orifice. For example, noninvasive stimulation of the cervical vagus nerve which involves stimulating specific afferent fibers of the vagus nerve to modulate brain function has been demonstrated in animal and human studies to treat a wide range of central nervous system disorders including headache, chronic and acute cluster and migraine, epilepsy, bronchoconstriction, anxiety disorders, depression, rhinitis, fibromyalgia, irritable bowel syndrome, PTSD, Alzheimer’s disease, and autism.

[0133] In some embodiments, neurostimulation may be administered by transcutaneous electrical stimulation of a nerve which is non-invasive because it involves attaching electrodes to the skin, or otherwise stimulating at or beyond the surface of the skin or using a form-fitting conductive garment, without breaking the skin. In contrast, percutaneous electrical stimulation of a nerve is minimally invasive because it involves the introduction of an electrode under the skin, vianeedle-puncture of the skin. Another form of non-invasive electrical stimulation is magnetic stimulation. Tt involves the induction, by a time-varying magnetic field, of electrical fields and current within tissue, in accordance with Faraday's law of induction. Magnetic stimulation is non- invasive because the magnetic field is produced by passing a time-varying current through a coil positioned outside the body. An electric field is induced at a distance, causing electric current to flow within electrically conducting bodily tissue. The electrical circuits for magnetic stimulators are generally complex and expensive and use a high current impulse generator that may produce discharge currents of 5,000 amps or more, which is passed through the stimulator coil to produce a magnetic pulse.

[0134] By way of example, such electrical stimulation can be achieved via the use of a neurostimulating device or neurostimulator which may be implanted within the subject's body.

[0135] Forms of neurostimulators or accessories thereof that can be employed in the methods disclosed herein are described in U.S. Patent Nos. 4,573,481 ; 4,702,254; 4,867,164; 4,920,979; 4,979,511 ; 5,025,807; 5,154,172; 5,179,950; 5,186,170; 5,215,089; 5,222,494; 5,235,980, 5,237,991 ; 5,251,634; 5,269,303; 5,304,206; and 5,351,394, and U.S. Patent Publication No. 2011 / 0276112. In particular aspects, the neurostimulators is an implantable pulse generator, such as the Vivistim system produced by MicroTransponder, Inc.

[0136] In some embodiments, the neurostimulator can include a non-invasive electrical stimulator configured to be applied to the subject’ s neck. In a preferred embodiment, the non- invasive electrical stimulator may comprise two electrodes that lie side-by-side within separate stimulator heads, wherein the electrodes are separated by electrically insulating material. Each electrode and the subject’s skin are connected electrically through an electrically conducting medium that extends from the skin to the electrode. The level of stimulation power may be adjusted with a wheel or other control feature that also serves as an on / off switch.

[0137] In an aspect, the neurostimulator can utilize a conventional microprocessor and other standard electrical and electronic components, and in the case of an implanted device, communicates with a programmer and / or monitor located externally to the subject’s body by asynchronous serial communication for controlling or indicating states of the device. Passwords, handshakes, and parity checks can be employed for data integrity. The neurostimulator also may include means for conserving energy, which is important in any battery-operated device, and especially where the device is implanted for medical treatment, and means for providing various safety functions, such as preventing accidental reset of the device.

[0138] In some embodiments, the neurostimulator can include an implantable pulse generator (IPG) configured to be implanted in the subject’s body in a pocket formed by the surgeon just below the skin in the chest in much the same manner as a cardiac pacemaker would be implanted.The IPG neurostimulator may include implantable stimulating electrodes together with a lead system for applying the output signal of the IPG to the subject’s vagus nerve. Components external to the subject’s body include a programming wand for telemetry of parameter changes to the IPG and monitoring signals from the generator, and a computer and associated software for adjustment of parameters and control of communication between the generator, the programming wand, and the computer. A stimulating nerve electrode set is conductively connected to the distal end of an insulated electrically conductive lead assembly attached at its proximal end to a connector. The electrode set can be a bipolar stimulating electrode of the type described in U.S. Patent No. 4,573,481. The electrode assembly is surgically implanted on the vagus nerve in the subject’s neck. The two electrodes are wrapped about the vagus nerve, and the assembly can be secured to the nerve by a spiral anchoring tether such as that disclosed in U.S. Patent No. 4,979,511. In an aspect, each lead may be secured while retaining the ability to flex with movement of the chest and neck by a suture connection to nearby tissue.

[0139] In conjunction with its microprocessor-based logic and control circuitry, the IPG can include a battery or set of batteries which can be of any reliable, long-lasting type conventionally employed for powering implantable medical electronic devices, such as those employed in implantable cardiac pacemakers or defibrillators. For example, the battery can be a single lithium thionyl chloride cell. The terminals of the cell are connected to the input side of a voltage regulator which smooths the battery output to produce a clean, steady output voltage, and provides enhancement thereof such as voltage multiplication or division if required. In some embodiments, the neurostimulator does not include a battery and is powered by near field communication, as well as to communicate with the implant.

[0140] In an example, the voltage regulator supplies power to the logic and control section, which may include a microprocessor and controls the programmable functions of the device. Among these programmable functions are output current, output signal frequency, output signal pulse width, output signal on-time, output signal off-time, daily treatment time for continuous or periodic modulation of vagal activity, and output signal-start delay time. Such programmability allows the output signal to be selectively crafted for application to the stimulating electrode set to obtain the desired modulation of vagal activity. Timing signals for the logic and control functions of the generator are provided by a crystal oscillator.

[0141] A built-in antenna enables communication between the IPG and the external electronics, including both programming and monitoring devices, to permit the device to receive programming signals for parameter changes, and to transmit telemetry information from and to the programming wand. Once the system is programmed, it can operate continuously at theprogrammed settings until they are reprogrammed by means of the external computer and the programming wand.

[0142] The logic and control section of the IPG controls an output circuit or section which generates the programmed signal levels appropriate for the condition being treated. The output section and its programmed output signal are coupled (e.g., directly, capacitively, or inductively) to an electrical connector on the housing of the generator and to a lead assembly connected to the stimulating electrodes. Thus, the programmed output signal of the IPG can be applied to the electrode set implanted on the subject’s vagus nerve to modulate vagal activity in the desired manner.

[0143] The housing in which the IPG is encased is hermetically sealed and composed of a material such as titanium, ceramics, or glass which are biologically compatible with the fluids and tissues of the subject's body. In particular aspects, the material is glass.

[0144] The IPG can be programmed using a personal computer employing appropriate software. The software permits non-invasive communication with the generator after the latter is implanted, which is useful for both activation and monitoring functions. Programming capabilities should include the ability to modify the adjustable parameters of the IPG and its output signal, to test device diagnostics, and to store and retrieve telemetered data. The implantable stimulator may be programmed with a tablet, such as using bluetooth.

[0145] Diagnostics testing should be implemented to verify proper operation of the device. The nerve electrodes are capable of indefinite use absent indication of a problem with them observed on such testing.

[0146] A source of power supplies a pulse of electric charge to the electrodes, such that the electrodes produce an electric current and / or an electric field within the subject. The electrical stimulator is configured to induce a peak pulse voltage sufficient to produce an electric field in the vicinity of a nerve such as a vagus nerve, to cause the nerve to depolarize and reach a threshold for action potential propagation. By way of example, the threshold electric field for stimulation of the nerve may be about 8 V / m at 1000 Hz. For example, the device may produce an electric field within the subject of about 10 to 600 V / m (preferably less than 100 V / m) and an electrical field gradient of greater than 2 V / m / mm. Electric fields that are produced at the vagus nerve are generally sufficient to excite all myelinated A and B fibers, but not necessarily the unmyelinated C fibers. However, by using a reduced amplitude of stimulation, excitation of A- delta and B fibers may also be avoided.

[0147] Current passing through an electrode may be about 0 to 40 mA, with voltage across the electrodes of about 0 to 30 volts. The current is passed through the electrodes in bursts of pulses. There may be 1 to 30 pulses per burst, such as 5, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 25pulses per burst, particularly 15 or 16 pulses per burst. Each pulse within a burst has a duration of about 20 to 1000 microseconds, such as 30, 40, 50, 60, 70, 80, 90, 100, 125, 150, 175, or 200 microseconds, preferably 100 microseconds. A burst followed by a silent inter-burst interval repeats at 1 to 5000 bursts per second (bps, similar to Hz), preferably at 15-50 bps, and even more preferably at 30 bps. The preferred shape of each pulse is a full biphasic current controlled charge balanced wave. In certain embodiments, the electrical signal is applied one to 1000 times during a therapy session, such as 10, 25, 50, 75, 100, 500 or 1000 times during a VNS treatment. The vagus nerve stimulation treatment according may be conducted for 0.1 seconds to five seconds, preferably about 0.5 seconds (each defined as a single dose).

[0148] The electric pulse train of the VNS may have a current amplitude of 0.1 to 2.0 milliamps (mA), such as between 0.4 to 1.0 mA, or between 0.7 to 0.9 mA, such as at around 0.8 mA. The electric pulse train may also have a duration of 30 to 5000 milliseconds (ms), such as 125 to 2000 ms, 400 to 600 ms, or 500 ms. For example, the electric pulse train with a duration of 500 ms typically consists of 15 pulses at 30 Hz. An increase in pulse train duration would be associated with an increase in the number of pulses or a decrease in frequency. Conversely, a decrease in pulse train duration would be associated with a decrease in the number of pulses or an increase in frequency.

[0149] In some embodiments, the VNS may be applied continuously for a given period of time. The term “continuously stimulate” as defined herein means stimulation that follows a certain On / Off pattern continuously 24 hours / day. For example, existing implantable vagal nerve stimulators “continuously stimulate” the vagus nerve with a pattern of 30 seconds ON / 5 minutes OFF for 24 hours / day and seven days / week. However, the treatment may then be modified on an individualized basis, depending on the response of each particular subject.

[0150] The VNS can be administered 1 day to 6 months, up to years after injury. For example, the individual can be treated immediately after injury, or within 1, 2, 3, 4, 5, 6 days of injury, or within 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 13, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50 days or more of injury, or within 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20 or more years after injury.

[0151] The preferred stimulator shapes an elongated electric field of effect that can be oriented parallel to a long nerve, such as a vagus nerve. By selecting a suitable waveform to stimulate the nerve, along with suitable parameters such as current, voltage, pulse width, pulses per burst, or inter-burst interval, the stimulator produces a correspondingly selective physiological response in an individual subject. Such a suitable waveform and parameters are simultaneously selected to avoid substantially stimulating nerves and tissue other than the target nerve, particularly avoiding the stimulation of nerves in the skin that produce pain.

[0152] The methods for verifying and monitoring stimulation of the vagus nerve rely on the stimulated vagus nerve causing some physiological response that can be measured, such as some change in the subject’s voice (by virtue of stimulation of a recurrent laryngeal nerve, which is a branch of the vagus nerve), autonomic nervous system, evoked potential, chemistry of the blood, or blood flow within the brain are described, for example, in U.S. Patent No. 9,254,383.

[0153] In some embodiments, one or more triggering methods may be used to determine selection criteria, preprocess collected data, and activate triggering of therapeutic stimulation of the subject. FIGs. 2A-2C show examples of a graphical description of a dynamic algorithm, a static threshold algorithm, and a periodic algorithm respectfully. FIGS. 3A-3C show examples of triggering during rehabilitative exercises using the respective triggering methods according to some embodiments.

[0154] The triggering methods were validated using data from 16 stroke subjects using motion controllers to perform rehabilitative exercises. Numerous streams of sensor data were reduced into a single signal, then a trigger method was used to determine whether or not to deliver stimulation as the rehabilitative exercise progressed.

[0155] Data are reported as mean + / - standard error of the mean (SEM) or median with interquartile range (IQR). Where appropriate, standard parametric statistical tests (unpaired t- tests) were used to make comparisons. Statistical tests for each comparison are noted in the text. Unpaired two-tailed t-tests were used to determine differences in the triggering quality of the triggering methods. Unpaired two-tailed t-tests were used to determine differences in the stimulation pairings of each algorithm. The threshold for statistical significance was set at p < 0.05. Error bars in figures represent SEM.

[0156] In an aspect, the dynamic algorithm automatically individualizes the triggering criteria without the need to fine tune parameters for each person or session. An adaptive minimum activity threshold fluctuates according to recent sensor input to achieve selectivity on the largest or best sensor inputs while maintaining the optimum rate of stimulations. The dynamic threshold ensures the dynamic algorithm can be applied to various exercises across a range of impairment levels without losing its main advantages. These characteristics may increase the number of ideal VNS pairings during therapy.

[0157] Turning to FIG. 2A, in some embodiments, a method 210 for performing a dynamic algorithm may include a step 212 of measuring a current movement, a step 214 of adjusting a threshold, a step 216 of determining when the current movement is above the threshold, a step 218 of triggering a stimulation therapy when the current movement is above the threshold (yes), and a step 217 when the current movement is not above the threshold (no). During step 217, the method returns to step 212. In an example, step 214 of adjusting a threshold may include settinga minimum activity threshold based on a recent or contemporaneous current movement or sensor input. Tn an example, the dynamic algorithm delivers stimulation triggers when the sensor input signal exceeded a dynamically-adjusted minimum activity threshold, which varied during the exercise session based on recent sensor input. In an example, the minimum activity threshold may be set to the 90th percentile of recent sensor input.

[0158] In some aspects, the dynamic algorithm is configured to continually adapt over time to adjust for intermittent periods of rest and person-to-person variability, which can flexibly be applied to signals from different types of controllers while maintaining similar triggering characteristics. The dynamic algorithm reliably selected the best sensor inputs with desired timing intervals.

[0159] Turning to FIG. 3A, results of a dynamic algorithm applied to activity data over time is shown according to an example. In some embodiments, a dynamic algorithm may pair VNS with clear sensor inputs and produces minimum triggering rate variability. First, the performance of a dynamic algorithm was explored. Since variability of impairment levels between stroke, SCI, or TBI subjects is a consideration, the dynamic algorithm was designed to actively adjust the stimulation threshold to account for differences in performance across subjects and exercises. Multiple minimum activity thresholds were created based on recent sensor input for analysis (Percentiles: 65%, 70%, 75%, 80%, 85%, 90%, 95%). Analysis of the dynamic algorithm showed VNS was routinely paired with sensor inputs above the set minimum percentile of recent movement. When the dynamic algorithm was set to pair VNS with sensor inputs above the 90th percentile of recent movements, the dynamic algorithm triggered at 5.48 (IQR: 2.34) stimulations per minute and selected movements of 91.83% (IQR: 1.33) during VNS (FIGs. 7C and 7D).

[0160] All triggers were paired with sensor inputs above the 90th percentile of recent sensor inputs (FIG. 7A). Moreover, all sessions resulted in at least one stimulation trigger, indicating that this approach is robust across varying types of signals and levels of participant performance. As expected, there was a clear trade-off between sensor input selectivity and triggering rate. The dynamic algorithm selected higher sensor input magnitudes compared to the static threshold algorithm across a wide range of triggering rates (FIG. 8).

[0161] In an aspect, the static threshold algorithm is capable of selecting large movements while maintaining an optimum stimulation rate. The capability to select clear movements makes a static threshold appealing, but the realistic application of such an algorithm is hindered by the large variability of impairment levels observed in subjects with neurological injuries. Additionally, motor performance can fluctuate day-to-day or with improvement over the course of rehabilitation, which complicates selection of an appropriate static threshold. If the threshold is set too high, a subject may not perform any movements of a magnitude great enough to triggerVNS. If the threshold is set too low, sensor input of virtually any magnitude will trigger stimulation, which limits selection of the largest or best movements.

[0162] Turning to FIG. 2B, in some embodiments, a method 220 for performing a static threshold algorithm may include a step 212 of measuring a current movement, a step 216 of determining when the current movement is above the threshold, a step 218 of triggering a stimulation therapy when the current movement is above the threshold (yes), and a step 217 when the current movement is not above the threshold (no). In an aspect, the static threshold algorithm is configured to deliver stimulation triggers when the sensor input signal exceeds a fixed minimum activity threshold, which may be set to a level which produces a median triggering rate similar to the dynamic algorithm.

[0163] Turning to FIG. 3B, results of a static threshold algorithm applied to activity data over time is shown according to an example. A static threshold algorithm pairs VNS with clear sensor inputs, but produces substantial triggering rate variability. Second, performance of a static threshold algorithm was evaluated. The static threshold reduces the complexity of the signal processing, but it could reduce the flexibility to accommodate differences in performance across different subjects and exercises. The performance of the static threshold algorithm was explored at multiple fold increases over the noise floor (sensor input minimum multipliers: 2, 4, 6, 8, 10, 12, 14). FIGS. 7C and 7D show a sensor input minimum multiplier of lOx produces a triggering rate of 5.39 (IQR: 6.25) stimulations per minute and selectivity of 78.0% (IQR: 42.3). 13.7% of sessions resulted in no stimulation triggers during the entirety of the session.

[0164] Individual analyses of these sessions revealed that participants were generating sensor input during the session, but the measured sensor inputs were not large enough to surpass the static threshold. 71.3% of sessions resulted in total sensor input- VNS pairings below 90% selectivity. Individual analyses of these sessions revealed an abundance of activity above the minimum activity threshold that was set.

[0165] When the triggering threshold was increased, selectivity of the static threshold algorithm increased and the triggering rate decreased (FIG. 8). At high thresholds, selection emphasized performance-based triggering, so stimulations only occurred on the clearest sensor inputs. However, this consequently produced less frequent triggers. Similarly, when the triggering threshold was decreased, the selectivity of the static threshold algorithm decreased and the triggering rate increased.

[0166] In an aspect, the periodic algorithm employs simple timing for the sake of prioritizing the rate of VNS over the quality of its pairings. The periodic algorithm may be beneficial for ease of implementation and simplifies validation testing. Turning to FIG. 2C, in some embodiments, a method 220 for performing a periodic algorithm may include a step 232 of counting down a timerand a step 218 of triggering a stimulation therapy when the countdown is complete. In an example, the countdown timer can be set at a regular interval, irrespective of the movement signal. Turning to FIG. 3C, results of a periodic algorithm applied to activity data over time is shown according to an example. A periodic algorithm was found to provide consistent interstimulation intervals but poor selectivity. Performance was evaluated at multiple inter-stimulus intervals (6, 6.67, 7.5, 10, 12, 15 and 20 seconds). By design of the periodic algorithm, each timing parameter resulted in a consistent stimulation interval at the set value. (See FIG. 7C).

[0167] Because the periodic algorithm does not consider sensor input when determining triggering, the periodic algorithm often produced trigger events when no sensor input was occurring and only rarely produced trigger events during large sensor inputs. As a result, the periodic algorithm consistently stimulated on sensor inputs that were below the 50th percentile of recent movements. The median movement selectivity was 0 (IQR: 9.67) when the inter- stimulus interval was set to 12 seconds. (See FIG. 7D). The periodic algorithm frequently triggered stimulation during periods of rest, which decreased the percentile of the selected sensor inputs. When the inter-stimulus interval was shortened or lengthened, the selection characteristics of the periodic algorithm do not improve (FIG. 8). Thus, the periodic algorithm produced reliable trigger intervals, but demonstrated poor selectivity for large sensor inputs.

[0168] In an aspect, all proposed algorithms may be used to create an appropriate triggering interval in combination or isolation. In some implementations, one or more of the triggering methods may be applied to one or more sensors tracking a portion of a subject’s activity. In an example, a first algorithm may be used for monitoring a first portion of a subject’s activity and a second algorithm may be used for monitoring a second portion of a subject’s activity. In some implementations, one or more of the algorithms may maintain a desired triggering interval while providing 5-12 triggers per minute. In an example, a first algorithm may be used across exercise types and subjects. (FIG. 7D). In an example, a second algorithm may be used to prevent stimulation events from occurring in quick succession by setting a minimum inter-stimulus interval. This second or separate algorithm may ensure stimulations are separated by at least the minimum inter-stimulus interval (e.g., 5 seconds). In an example, the minimum activity threshold in the static threshold algorithm was set to 10 times the movement sensor input. The resulting mean inter- stimulus interval was 19.07 + / - 1.09 seconds for sessions where triggering occurred. When the minimum percentile of recent sensor input was set to 90% in the dynamic algorithm, the average inter-stimulus interval was 12.83 + / - 0.29 seconds. In an example, each algorithm may be configured to produce a trigger around a 12 second interval.

[0169] Vagus nerve stimulation may enhance rehabilitation following neurological injury. Manually triggered VNS paired with upper limb rehabilitation may reduce long-term deficits following stroke.

[0170] In some implementations, VNS may be delivered by a therapist who manually pushes a button during movements they want to reinforce. This method may enhance motor function and improvements persist over time. This approach uses a stimulation paradigm congruent to the periodic algorithm described. The absence of continued improvements in function may reflect the lack of consistent stimulation during the best or largest magnitude movements. Thus, the dynamic algorithm described here represents a means to deliver unsupervised closed-loop stimulation during rehabilitation, which would facilitate longer and more intensive VNS therapy regimens.

[0171] A feasibility study of RePlay was done containing both traditional repetition-based exercises and dynamic game-controlling movements, each measured by a sensor array housed in a selected controller. All participants performed the same exercises and played each game at the same difficulty level. If subjects were not able to grasp and hold the device, the device was affixed to a stabilizing base. All procedures were approved by the Institutional Review Board at the University of Texas at Dallas (UTD IRB: 19-119). 16 participants ages 22 to 76 with a history of upper limb motor impairment due to stroke or TBI (mean time since neurological injury was 8 + 2 years) and 6 healthy participants ages 23 to 49 were recruited. During an initial office visit, participants used the FitMi controller to perform rehabilitative exercises while playing games on an Android tablet.

[0172] Collection of quantitative upper limb activity data from stroke and TBI subjects. The study sought to design a real-time algorithm to identify a largest magnitude sensor input during a variety of different rehabilitative exercises and activities of daily living and on average produce five stimulation pairings per minute, based on effective paradigms from preclinical studies. A previous set of rehabilitative activity data or dataset collected from 30 stroke or TBI subjects with impairments in upper limb motor function was utilized. The dataset included captured activity from 3 sensing devices, 7 games, and 22 different upper-limb rehabilitation exercises. Data was collected from 670 exercise sessions of one minute or longer. Individual sessions containing processed data had unique characteristics relative to each participant, game, exercise and controller type. FitMi puck movement was measured as rotation angle or force. Touch screen movement was measured as swipe velocity, and typing movement was measured as the time interval between correct key presses.

[0173] To capture activity data during rehabilitative exercises, the RePlay application streamed and processed incoming data from the controllers and saved the data to local storage for offline analysis. Custom Python routines were developed for simulations.

[0174] Processing of rotation and force signals created movement vectors that described acceleration over time. To begin, unprocessed movement signals were smoothed with a moving average filter (Discrete, linear convolution, 36 samples from activity data, kernel size: 5). At each point in time, a buffer or gradient of the 21 most recent values of the smoothed signal was obtained. The sum of all values in the gradient was calculated, resulting in a single value. Each consecutive movement signal value was processed in the same manner. This signal processing resulted in a representative movement sensor signal that could be used as input to an algorithm for stimulation.

[0175] Processing of typing and touchscreen data created signals that described the speed of touch over time. Incoming typing data was analyzed to determine the rate of correct key presses. If the interval between the last two key presses was above a selected percentile of the last 5 key press intervals, a stimulation trigger was requested. Similarly, the touchscreen data was processed by considering the swipe velocity of the current movement. If the current swipe velocity was above a selected percentile of a buffer of swipe velocity samples (e.g., last 300), a stimulation trigger was requested. In an example, the swipe velocity may be measured by a touch screen in pixel over time.

[0176] Movement minimums were selected for each exercise to separate movement from noise. The sensed movement minimum for each exercise was determined by measuring the mean maximum value of the signal when a control subject held the controller with the arm at rest. Any activity that did not exceed the sensed movement minimum was excluded.

[0177] The dynamic algorithm maximizes sensor input magnitude across all exercises and capabilities. Triggering stimulation to coincide with the largest sensor inputs during rehabilitation is necessary for VNS-dependent benefits. Selectivity of the triggering methods was compared across the entire data set to determine which algorithm balanced consistent timing with triggering on the largest sensor inputs. Quality of the sensor inputs selected by the triggering methods are represented as the percent of maximum sensor inputs during each exercise.

[0178] Overall, the dynamic algorithm resulted in the greatest percent of maximum sensor input. This indicates that the dynamic algorithm provides the most reliable selection of the largest sensor input. (See FIGs. 7A, 8). The dynamic algorithm selects large sensor inputs better than supervised manual triggering and the dynamic algorithm produces triggering at the end of motor execution. (See FIG. 7B).

[0179] Thus, provided herein is the design of a triggering method capable of pairing VNS with clear sensor inputs during upper limb rehabilitation following neurological injury.

[0180] Sensor data was recorded from stroke or TBI subjects and healthy participants during a variety of different rehabilitative exercises. The sensor data was used to develop a triggering method and compare it to alternative and clinically employed algorithms to compare which strategy exhibits the best triggering criteria. After testing a range of parameters within each of the triggering methods, a set of parameters within the dynamic algorithm were identified selecting the largest sensor inputs while maintaining a consistent median triggering rate.

[0181] A study was performed to demonstrate effectiveness of triggering methods to facilitate feedback-controlled neurostimulation during rehabilitation, aimed at increasing the dose and quality of VNS pairings. Stimulation timing and trial selection affect a magnitude of VNS- dependent enhancement of post-stroke recovery. A recent preclinical study illustrated reliance of VNS effects on trial selection, finding that pairing VNS with the strongest forelimb movements during rehabilitative training significantly enhanced recovery of forelimb strength, whereas pairing the weakest movements failed to promote recovery. Moreover, several studies confirm that a matched stimulation intensity that is not paired with activity or movement fails to enhance recovery. These provide the rationale for developing a triggering method that can trigger stimulation concurrent with the largest sensor inputs generated by muscle activity. Several additional lines of evidence demonstrate the importance of stimulation timing on VNS-dependent effects. Faster rates of stimulation (i.e., shorter inter-stimulation intervals) are associated with smaller VNS-dependent effects in a preclinical model. Moreover, clinical evidence shows large amounts of periodic VNS during rehabilitation does not enhance, and may in fact limit, recovery. Together, these findings reinforce the importance of incorporating inter-stimulation timing into an algorithm for unsupervised stimulation.

[0182] Aspects of the systems and methods provided herein, such as the computer or control system, RePlay App, may be embodied in programming. Various aspects of the technology may be thought of as "products" or "articles of manufacture" typically in the form of machine (or processor) executable code and / or associated data that is carried on or embodied in a type of machine-readable medium. Machine-executable code may be stored on an electronic storage unit, such as memory (e.g., read-only memory, random-access memory, flash memory) or a hard disk. "Storage" type media may include any or all of the tangible memory of the computers, processors or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives and the like, which may provide non-transitory storage at any time for the software programming. All or portions of the software may at times be communicated through the Internet or various other telecommunication networks. Such communications, for example, may enableloading of the software from one computer or processor into another, for example, from a management server or host computer into the computer platform of an application server. Thus, another type of media that may bear the software elements includes optical, electrical, and electromagnetic waves, such as used across physical interfaces between local devices, through wired and optical landline networks and over various air-links. The physical elements that carry such waves, such as wired or wireless links, optical links, or the like, also may be considered as media bearing the software. As used herein, unless restricted to non-transitory, tangible "storage" media, terms such as computer or machine "readable medium" refer to any medium that participates in providing instructions to a processor for execution.

[0183] Hence, a machine-readable medium, such as computer-executable code, may take many forms, including but not limited to, a tangible storage medium, a carrier wave medium or physical transmission medium. Non-volatile storage media include, for example, optical or magnetic disks, such as any of the storage devices in any computer(s) or the like, such as may be used to implement the databases, etc. shown in the drawings. Volatile storage media include dynamic memory, such as main memory of such a computer platform. Tangible transmission media include coaxial cables, copper wire and fiber optics, including the wires that comprise a bus within a computer system. Carrier-wave transmission media may take the form of electric or electromagnetic signals, or acoustic or light waves such as those generated during radio frequency (RF) and infrared (IR) data communications. Common forms of computer-readable media therefore include for example: a floppy disk, a flexible disk, hard disk, magnetic tape, any other magnetic medium, a CD-ROM, DVD or DVD-ROM, any other optical medium, punch cards paper tape, any other physical storage medium with patterns of holes, a RAM, a ROM, a PROM and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave transporting data or instructions, cables or links transporting such a carrier wave, or any other medium from which a computer may read programming code and / or data. Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.

[0184] The present disclosure provides computer systems that are programmed to implement methods of the disclosure. The computer system may be an electronic device of a user or a computer system that is remotely located with respect to the electronic device. The electronic device may be a mobile electronic device. The computer system may include or be in communication with an electronic display that comprises a user interface (UI) for providing, for example, selection of electro-stimulation parameters. Examples of UT’s include, without limitation, a graphical user interface (GUI) and web-based user interface. Methods and systemsof the present disclosure may be implemented by way of one or more algorithms. An algorithm may be implemented by way of software upon execution by a central processing unit.

[0185] While preferred embodiments of the present disclosure have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. It is not intended that the present disclosure be limited by the specific examples provided within the specification. While many embodiments of the present disclosure have been described with reference to the aforementioned specification, the descriptions and illustrations of the embodiments herein are not meant to be construed in a limiting sense. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the scope of the present disclosure. Furthermore, it shall be understood that all aspects of the present disclosure are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the present disclosure described herein may be employed in practicing the embodiments of the present disclosure. It is therefore contemplated that the present disclosure shall also cover any such alternatives, modifications, variations, or equivalents. It is intended that the following claims define the scope of the invention and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

CLAIMSWHAT TS CLAIMED IS:

1. A method of treating a subject having or at risk of having a dysfunction, the method comprising: establishing a set of task outcomes for at least one task, the set of task outcomes including one or more patterns of progressive activity, a progression threshold, and one or more patterns of antagonist activity; monitoring a subject’s activity performing the at least one task, including monitoring the one or more patterns of progressive activity; comparing the subject’s monitored activity to the set of task outcomes; and triggering a stimulation of a nerve when performance of the at least one task, comprises the monitored one or more patterns of progressive activity that meets the progression threshold and does not involve the one or more patterns of antagonist activity.

2. The method of claim 1 , wherein the task comprises a performance of a rehabilitative exercise.

3. The method of claim 1 , wherein the task comprises a performance of a game.

4. The method of claim 1 , wherein the task comprises a motor response, a sensory response, a physiological response, or a cognitive response.

5. The method of claim 1, wherein the task comprises an activity of daily living.

6. The method of claim 1 , wherein the nerve is a cranial nerve.

7. The method of claim 1, wherein the nerve is the IO111cranial nerve.

8. The method of claim 1, wherein the stimulation occurs at a frequency once every five seconds or longer.

9. The method of claim 1 , wherein the stimulation occurs after at least a refractory period of a subject’s movement.

10. The method of claim 1, wherein the stimulation does not occur at or near an onset of the subject’s activity.

11. The method of claim 1 , wherein the task outcome includes one or more of a task completion beyond the progression threshold, an activity, a sequence of activities, a pattern of activities, an activity repetition, a sequence repetition, a pattern repetition, a peak force, a peak range of motion, a peak acceleration, a peak velocity, a peak angular velocity, a peak angular acceleration, or electromyogram activity,12. The method of claim 1, wherein the antagonist activity is one or more of a compensatory movement, a non-sequential movement, a muscle tremor, a muscle spasticity, or a failure to activate a target muscle or a muscle group, or activation of muscles opposing the task outcomes, the at least one task, or one or more patterns of progressive activity.

13. The method of claim 1, wherein the progression threshold is at least at the 50thpercentile, 75thpercentile, 80thpercentile, 85thpercentile, 90thpercentile, 95thpercentile, 98thpercentile or 99thpercentile of the set of task outcomes.

14. The method of claim 1, wherein the set of task outcomes includes 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 75, 100, 500, or 1000 tasks.

15. The method of claim 1, wherein the monitoring is done using one or more sensors in contact with the subject.

16. The method of claim 1, wherein the monitoring is performed using a controller.

17. The method of claim 1, wherein the monitoring is performed using a camera.

18. The method of claim 1 , wherein the monitoring is performed using one or more EMG sensors.

19. The method of claim 1, wherein the monitoring is performed using one or more EEG sensors.

20. The method of claim 1, wherein the dysfunction is a motor disability, a sensory disability, a dysfunction of the ankle, knee, hip, thumb, finger / fingers, wrist, elbow, shoulder or trunk, amotor sensory dysfunction, a result of a stroke, multiple sclerosis, surgical procedures, traumatic brain injury, spinal cord injury, PTSD, aphasia, a result of a peripheral nerve or limb injury, a walking dysfunction, or ankle dorsiflexion.

21. The method of claim 1, wherein the stimulation is applied for a pre-determined time to invoke the subject’s Hebbian process.

22. The method of claim 1, further comprising assessing a stop condition and instructing the subject to stop the subject’s activity when the stop condition is met.

23. The method of claim 1, further comprising adjusting a duration of stimulation.

24. The method of claim 1, further comprising adjusting the progression threshold for stimulation.

25. The method of claim 1, wherein the progression threshold includes a threshold duration.

26. The method of claim 25, further comprising adjusting the threshold duration.

27. The method of claim 1 , further comprising updating the set of task outcomes based on the monitoring of the subject’s activity performing the at least one task.

28. The method of claim 1, further comprising adjusting an intensity of the stimulation within a repetition of the performance of the at least one task.

29. The method of claim 1, further comprising adjusting an intensity of the stimulation based on activation of the subject’s recurrent laryngeal nerve.

30. The method of claim 1, further comprising adjusting the progression threshold such that the subject receives a predetermined number of nerve stimulations per unit time.

31. The method of claim 30, wherein the predetermined number of nerve stimulations per unit time is about 200 simulations per day.

32. The method of claim 1 , further comprising adjusting the progression threshold based on a buffer of task outcomes.

33. The method of claim 1, further comprising measuring at least one characteristic of the subject’s response to the at least one task and adjusting the progression threshold based on the at least one characteristic of the subject’s response to the at least one task.

34. The method of claim 33, wherein the measured at least one characteristic is an interstimulation interval in response to a speech task.

35. The method of claim 34, wherein the speech task is a sequence production task for conversational speech.

36. The method of claim 1, further comprising preprocessing activity data.

37. The method of claim 36, wherein preprocessing activity data comprises averaging or smoothing the subject’s activity performing the at least one task.

38. The method of claim 37, wherein smoothing the subject’s activity performing the at least one task comprises removing outliers or noise.

39. The method of claim 37, wherein smoothing the subject’s activity performing the at least one task comprises applying a moving average to a buffer of the monitoring of the subject’s activity performing the at least one task.

40. The method of claim 39, wherein the buffer includes a predetermined time of about 300 ms.

41. The method of claim 36, wherein preprocessing activity data comprises extracting one or more principal components from the subject’s activity.

42. The method of claim 41 , wherein the one or more principal components from the subject’s activity includes attributes of displacement, velocity, acceleration, a Euclidean distance from the subject’s activity to the at least one task, one or more of changes in direction, intermittent differences in range, and intermittent differences in pauses.

43. The method of claim 41 , wherein the one or more principal components from the subject’s activity includes anatomical features of the subject such as their body composition,muscles, joints, tendons, flexibility, range of motion among other subject-specific factors such as missing digits or severity of condition.

44. The method of claim 1, wherein the progression threshold is established from the set of task outcomes.

45. The method of claim 1, wherein the stimulation is limited to no more than once every 5 seconds.

46. The method of claim 1 , wherein the stimulation is not performed when a compensatory activity is detected.

47. The method of claim 1, further comprising determining when a continuous improvement of the subject’s activity performing the at least one task occurs, and triggering a stimulation based on the continuous improvement.

48. The method of claim 47, wherein the continuous improvement of the subject’s activity performing the at least one task is based on a quantitative measure of the subject’s function or task outcome.

49. A method of treating a subject having or at risk of having a dysfunction, the method comprising: selecting a task including at least one associated activity, at least one task outcome, one or more patterns of antagonist activity, and a progression threshold; capturing a subject performing the task; determining when performance of the task meets the progression threshold; and stimulating the subject’s nerve when the performance of the task meets the progression threshold and does not involve the one or more patterns of antagonist activity.

50. A system for treating a subject having or at risk of having a dysfunction, the system comprising: a computer processing device comprising a processor and a non-volatile storage medium with instructions for the processor to perform the method of any one of claims 1 to 49.

Images