Brain machine interface for rehabilitation of impaired movement

Abstract

A system to rehabilitate stroke or traumatic brain injury survivors includes a brain machine interface that receives brain signals from the brain of a patient. The brain signals represent motor intent of the patient with respect to a hand of the patient that has impaired dexterous hand function due to neurological injury. A limb interface mounts to the hand of the patient. The limb interface detects feedback in the hand that results from the brain signals and helps to control the hand in accordance with the motor intent of the brain signals. The system also includes a processor operatively coupled to the brain machine interface and the limb interface. The processor is configured to synchronize the motor intent with the feedback to increase plasticity between sensory and motor areas of the brain to improve voluntary control of the hand, including dexterous hand functions such as multiple types of grasp.

Description

Atty. Dkt. No. 00100-0414-PCTBRAIN MACHINE INTERFACE FOR REHABILITATION OF IMPAIRED MOVEMENTCROSS-REFERENCE TO RELATED APPLICATION

[0001] The present application claims the priority benefit of U.S. Provisional Patent App. No. 63 / 736,005 filed on December 19, 2024, the entire disclosure of which is incorporated by reference herein.REFERENCE TO GOVERNMENT RIGHTS

[0002] This invention was made with government support under NS094748 awarded by the National Institutes of Health. The government has certain rights in the invention.BACKGROUND

[0003] The restoration of hand and arm function after stroke or traumatic brain injury' (TBI) is one of the most important and challenging unmet needs of neurorehabilitation. Even with the best conventional therapy, up to 60% of the 101 million people with chronic stroke worldwide are still affected by permanent disability due to hand and arm impairment. Less than 10% of stroke patients with moderate to severe upper extremity weakness ultimately regain adequate strength and dexterity in the impaired arm to allow them to perform bimanual tasks fully and independently. Although hemiparetic stroke survivors can use compensatory mechanisms (e.g., the contralateral arm) and assistive devices to perform most basic activities of daily living, survivors are still unable to perform tasks requiring dexterity or bimanual coordination (e.g., dressing, cutting food, handling large objects, or many occupational tasks such as typing). This typically leads such stroke survivors to either be inefficient or reliant on others for assistance in these activities. This, in turn, translates into increased caregiving costs, complications, decreased quality of life for the individual, and increased burden to society.SUMMARY

[0004] An illustrative system to rehabilitate stroke or traumatic brain injury survivors includes a brain machine interface that receives brain signals from the brain of a patient. The brain signals represent motor intent of the patient with respect to a hand of the patient that has impaired dexterous hand function due to neurological injury’. A limb interface mounts to the hand of the patient. The limb interface detects feedback in the hand that results from the brainAtty. Dkt. No. 00100-0414-PCT signals and helps to control the hand in accordance with the motor intent of the brain signals. The system also includes a processor operatively coupled to the brain machine interface and the limb interface. The processor is configured to synchronize the motor intent with the feedback to increase plasticity between sensory and motor areas of the brain to improve voluntary control of the hand, including dexterous hand functions such as multiple types of grasp.

[0005] In an illustrative embodiment, the brain signals represent an intended movement, muscle activation(s), or amount of force to be produced by the hand. In another embodiment, the limb interface comprises a functional electrical stimulation device, an exoskeleton, or other haptic device. In another embodiment, the processor also enables activity that increases plasticity between a motor cortex and spinal motoneurons. In one embodiment, the brain machine interface is an intracranial device. In another embodiment, the brain signals include one or more of electrocorticographic (subdural) signals, intracortical signals, epidural field potentials, intracortical local field potentials or action potentials, and intravascularly recorded signals.

[0006] In one embodiment, the processor generates a motor map of the brain that identifies one or more core regions that represent areas of the brain showing highest modulation with intended hand movement. In another embodiment, the motor map also identifies one or more edge regions that represent areas of the brain showing only partial modulation with the intended hand movement. In another embodiment, the processor controls the BMI within the edge regions to increase modulation of the one or more edge regions synchronously with sensory inputs such that the plasticity increases. In another embodiment, the sensory feedback is controlled by both the BMI and by residual movement in the hand, and wherein a proportion of control may change over time as the patient regains more limb function.

[0007] An illustrative method of rehabilitating people with stroke or traumatic brain injury includes receiving, by a brain machine interface, brain signals from a brain of a patient, wherein the brain signals represent motor intent of the patient with respect to a hand of the patient that has impaired dexterous hand function due to neurological injury. The method also includes controlling, by a limb interface that mounts to the hand of the patient, the hand in accordance with the motor intent of the brain signals, thus providing feedback from the handAtty. Dkt. No. 00100-0414-PCT to the central nervous system. The method also includes synchronizing, by a processor operatively coupled to the brain machine interface and the limb interface, the motor intent with the feedback from the hand to generate signals that increase plasticity between sensory and motor areas of the brain to improve voluntary control of the hand, including dexterous hand functions such as multiple types of grasp.

[0008] In an illustrative embodiment, the brain signals represent an intended movement, muscle activation(s), or amount of force to be produced by the hand. In another embodiment, the limb interface comprises a functional electrical stimulation device, an exoskeleton, or other haptic device. In another embodiment, the processor also enables activity that increases plasticity between a motor cortex and spinal motoneurons. In one embodiment, the brain machine interface is an intracranial device. In another embodiment, the brain signals include one or more of electrocorticographic (subdural) signals, intracortical signals, epidural field potentials, intracortical local field potentials or action potentials, and intravascularly recorded signals.

[0009] The method can also include generating, by the processor, a motor map of the brain that identifies one or more core regions that represent areas of the brain showing highest modulation with intended hand movement. In one embodiment, the motor map also identifies one or more edge regions that represent areas of the brain showing only partial modulation with the intended hand movement. In another embodiment, the method includes controlling, by the processor, the BMI within the edge regions to increase modulation of the one or more edge regions synchronously with sensory inputs such that the plasticity increases. In another embodiment, the sensory feedback is controlled by both the BMI and by residual movement in the hand, and wherein a proportion of control may change over time as the patient regains more limb function.

[0010] Other principal features and advantages of the invention will become apparent to those skilled in the art upon review of the following drawings, the detailed description, and the appended claims.BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Illustrative embodiments of the invention will hereafter be described with reference to the accompanying drawings, wherein like numerals denote like elements.Atty. Dkt. No. 00100-0414-PCT

[0012] Fig. 1 depicts putative plasticity mechanisms from using the invention and shows that sensory feedback driven by functional electrical stimulation (FES), using a brainmachine interface (BMI-FES) strengthens synapses from sensory cortex (SI) to motor cortex (Ml) as well as SI to premotor (PMA, including both premotor and supplementary motor areas) and between Ml and a -motoneurons via antidromic or orthodromic routes (circles) in accordance with an illustrative embodiment.

[0013] Fig. 2A depicts a motor map that is defined by the correlation coefficient (R) between each electrode’s high-gamma (HG) band activity and attempted movement in accordance with an illustrative embodiment.

[0014] Fig. 2B depicts BMI training with the partially modulated edges of the map in accordance with an illustrative embodiment.

[0015] Fig. 2C depicts that R values are recomputed after BMI training, and compared with the original distribution in accordance with an illustrative embodiment.

[0016] Fig. 3A depicts an example motor map; each rectangle represents the change, from before to after training, in R between force and HG activity from one electrode in accordance with an illustrative embodiment.

[0017] Fig. 3B shows that R increased more in the electrodes used in the BMI decoder than in other electrodes in subjects H3 and H4, in accordance with an illustrative embodiment.

[0018] Fig. 4A depicts that a TBI subject used the BMI based on attempted thumb flexion force decoded from HG EEG recorded in a hemicraniectomy area (a portion of the scalp under which the skull has been removed) to control force feedback to the thumb via a custom-built device in accordance with an illustrative embodiment.

[0019] Fig. 4B depicts the distribution of changes in R values with force, from baseline to last day, over all electrodes for BMI and control TBI subjects in accordance with an illustrative embodiment.

[0020] Fig. 4C shows that during HG-band BMI control, the HG peak was more tightly coupled to the force feedback peak (HGlag between peaks: 130 ms) than the mean j.i power nadir was to the force feedback peak during / z-band BMI control in 3 healthy subjects ( / zlag: 400 ms) in accordance with an illustrative embodiment.Atty. Dkt. No. 00100-0414-PCT

[0021] Fig. 5 is a block diagram of a computing system to implement a stroke rehabilitation system in accordance wi th an illustrative embodiment.

[0022] Fig. 6 is a schematic of one embodiment of a BMI, including ECoG electrodes, amplifier and transmitter, BMI processor, and hand exoskeleton and FES system in accordance with an illustrative embodiment.DETAILED DESCRIPTION

[0023] More than 60 million people worldwide have impaired hand function due to stroke. The resulting activity limitations significantly reduce quality of life and lead to increased costs to society due to lost productivity and reliance on caregivers. Despite decades of research, established rehabilitation methods do not restore sufficient hand function. In particular, with currently used rehabilitation therapies, it is difficult to restore multiple types of grasp dexterously, let alone individuated finger movements. Hence, there remains an urgent need for novel techniques that can effectively restore dexterous grasping to improve quality of life after stroke.

[0024] Brain-machine interfaces (BMIs) offer the potential not only to replace function after stroke by controlling prostheses, but also to rehabilitate function by driving brain plasticity. For example, a BMI controlling an exoskeleton or functional electrical stimulation (FES) of a paralyzed arm can synchronize motor intent with sensory feedback and drive plastic changes between sensory and motor areas of the brain. Stroke survivors using noninvasive (mainly EEG-based) BMIs to control haptic feedback or FES have had modest success in improving arm function or reorganizing cortical activity7. However, noninvasive signals lack the spatial, and particularly, temporal resolution that would best enable a BMI to synchronize activity and thus drive plasticity. In contrast, electrocorticography (ECoG) provides high-bandwidth signals, namely the high-gamma band (70- 200 Hertz (Hz)), that have higher spatiotemporal resolution than low-frequency signals. Because plasticity depends upon synchrony between pre- and post-synaptic neurons, it is hypothesized that high- bandwidth. ECoG-based BMIs can create a much higher degree of synchrony between different cortices, or between cortex and spinal motoneurons in the case of FES, and therefore drive plasticity7more effectively between these areas, than can noninvasive methods. It has been shown that ECoG can be used to decode continuous, individuated finger movements and different grasp types that enable BMIs to control prostheses. BMI-based therapy could enableAtty. Dkt. No. 00100-0414-PCT targeted plasticity to areas of cortex with residual function and could eventually be combined with less-specific ways of enhancing excitability, e.g., stimulation or drugs.

[0025] Described herein is a BMI-controlled FES (alternatively, BMI-controlled haptic feedback) system for improving hand and arm function in human survivors of stroke or traumatic brain injury (TBI). The system can be implemented as a safe implantable device for delivering the BM1 therapy. In one embodiment, the system utilizes a high-density ECoG device integrated with an FES system for the forearm to produce dexterous hand movements. In alternative embodiments, different components, such as an epidural device, intracortical device, or intravascular device, may be used.

[0026] The system can be used to help severely impaired stroke survivors. Conventional rehabilitation techniques often do not help patients with severe impairments, largely because they have insufficient remaining function to participate. Despite this need for new treatments, severely impaired patients are often excluded from clinical trials. Since the BMI paradigm uses attempted movements, it will not require patients to have remaining hand function. Hence, this paradigm is specifically designed to help patients with severe hand impairment. It could also help patients with severe arm, in addition to hand, impairment.

[0027] Brain machine interfaces (BMIs) enable subjects to use their brain signals to directly control external devices, such as a robotic arm or exoskeleton. BMIs can provide a powerful tool to drive plastic changes in the brain. For example, a BMI controlling an exoskeleton or functional electrical stimulation (FES) of the paralyzed arm can synchronize motor intent with sensory' feedback and drive plasticity between sensory and motor areas of the brain, as well as between motor cortex and spinal motoneurons. Fig. 1 depicts plasticity mechanisms and shows that BMI-FES-driven sensory feedback may strengthen synapses from SI -Ml and SI -premotor (PMA, including both PM and SMA) as well as between Ml and a- motoneurons via antidromic or orthodromic routes (circles) in accordance with an illustrative embodiment. In Fig. 1. the arrows depict directions of signal flow generated by BMI-FES.

[0028] To date almost all BMI studies for stroke have used noninvasive methods such as electroencephalography (EEG), which have limited ability to acquire high-frequency signals . This reduces the ability of an EEG-based BMI to synchronize motor intent with sensory feedback (and to synchronize upper / lower motor neuron activity), which in turn limitsAtty. Dkt. No. 00100-0414-PCT capacity to drive Hebbian-like (or other) plasticity. Further, noninvasive BMIs have limited movement-related information, and therefore use intention of coarse movements (hand open / close) vs. rest as the control signal for the BMI. The coarse imagery and sensory- feedback may limit the recovery of finer, dexterous movements.

[0029] Compared to EEG, signals recorded in the cerebral cortex (local field potentials, or LFPs) on the surface of the brain (electrocorticography, or ECoG) or surface of the dura mater (epidural field potentials) have much higher information content, spatial resolution, and bandwidth, including the high-gamma band (HG, 70-300 Hz) that also has higher temporal resolution than low-frequency signals in EEG. Also, it has been shown that ECoG or epidural HG can decode detailed grasping, individual finger movements, continuously graded finger force, and electromyography (EMG).

[0030] Although BMI training has shown promise for improving motor function after stroke, the mechanisms by which this occurs are unknown. One possibility is increasing the size of the cortical motor map (i.e., the area of the cerebral cortex that, when activated by stimulation or by volition, results in a given movement), as map enlargement roughly correlates with functional improvement after stroke. Moreover, BMI-controlled haptic feedback training can be used to induce neuroplasticity in multiple locations. It is hypothesized that individuals with poor hand function retain some movement-related modulation of the cortex, with intact (though weak) descending connections. Targeting these areas with the BMI should strengthen their association with the desired movement, thus enabling efficient enhancement of plasticity, since it is easier and faster to strengthen existing connections than build new ones. This enhancement is expected to repurpose these areas to control, and improve function of, the impaired limb. Accordingly, a BMI can be combined with a plurality of nervous system stimulation technologies to both enhance and target plastic changes.

[0031] The use of BMIs enables one to direct and augment the brain’s plasticity in targeted areas. A particular movement is associated with a gradient of brain activity, with an area of highest activity7(modulation) and other areas of w eaker activity (partial modulation), which is believed to be due to having weak corticofugal connections. For convenience, the area of highest modulation can be referred to as the "core" of the movement's motor map and the areas of partial modulation as the "edges," as shown by the shaded areas in Fig. 2, whichAtty. Dkt. No. 00100-0414-PCT is a schematic of targeting plasticity using BMTs. It is noted that these areas are not necessarily spatially contiguous. Fig. 2A depicts a motor map that is defined by the correlation coefficient (R) between each electrode’s HG activity' and attempted movement in accordance with an illustrative embodiment. The map core is defined as the electrodes most correlated with movement, e.g., those with R greater than 95thpercentile (Rcore). The BMI decoder emphasizes those electrodes with partial modulation (the “edges” of the motor map). Fig. 2B depicts BMI training with the partially modulated edges of the map in accordance with an illustrative embodiment. Fig. 2C depicts that R values are recomputed after BMI training, and compared with the original distribution in accordance with an illustrative embodiment. In this example, modulation has increased in the edges, enlarging the map core.

[0032] After a subcortical stroke, connections from the motor map core to the paralyzed muscles are severed or ineffective. It is hypothesized that some edge areas may have weaker, yet preserved, connections. One approach to neurorehabilitation is therefore to expand the motor map by increasing the modulation in the edge areas, using BMI decoders based on activity from these areas. It is expected that edge-targeted BMI use will result in a more substantial increase in modulation, and have a greater therapeutic effect, than would result if the BMI were targeted to just core areas, as is commonplace in BMIs try ing to replace function. By expanding the map, the edge areas are converted into core areas and existing pathways to lower motor neurons are strengthened, thus restoring motor function.

[0033] The inventors showed that this approach is feasible in a study of people with traumatic brain injury who have had hemicraniectomies (which enables HG to be recorded at the scalp with EEG). This was initially done in participants performing BMI control without haptic feedback. The inventors recorded EEG over the hemicraniectomy (hEEG) at 2 kHz to extract HG. These participants used imagined thumb flexion force to control a cursor moving to a force target on a screen. After just one session (20 min.) of BMI training, the correlation (R) of HG activity in the first participant’s (H3’s) electrodes with actual force increased significantly (p=0.03, t-test) in the edge electrodes used in the decoder compared with AR between baseline active pinch recordings a similar time apart (Fig. 3A). Changes in R for those electrodes in the decoder also tended to be greater than those not in the decoder (mean AR 0.02 vs. 0.006, respectively, p=0.09, Fig. 3B, left). The second participant (H4) also had significant changes in R values (p=3xl0‘7) after one session of BMI use, with greater changes in edge decoder electrodes (p=0.002, Fig. 3B, right). Fig. 3 shows how the modulationAtty. Dkt. No. 00100-0414-PCT increased after a single session of BMI training. Specifically, Fig. 3 A depicts squares that represent the change, from before to after training, in R between force and features from one electrode in the hemicraniectomy in accordance with an illustrative embodiment. Fig. 3B shows that R increased more in the electrodes used in the BMI decoder than in other electrodes in subjects H3 and H4, in accordance with an illustrative embodiment.

[0034] The inventors also developed a BMl-controlled haptic force feedback system to the thumb using HG in EEG in an attempt to drive cortical plasticity at 2-4 months postinjury. Specifically, a haptic device was designed to both sense and apply force to the dorsal surface of the thumb, forming an "assisted" pinch. Participants first completed a visual force matching task using actual thumb flexion force. The inventors built Wiener cascade decoders in each session using HG signals from hEEG electrodes in frontal through parietal rows that were also located over the hemicraniectomy. The Wiener cascade decoders were built on the hand-control dataset using 10-fold cross-validation, saving the highest-performing decoder for use in BMI-shared control. During BMI-shared control, visual task presentation was the same as in hand control: randomly selected force targets were presented as colors. Here, control over visual feedback was shared between the participant’s voluntary force, and BMI- controlled force. The shared control paradigm allowed the BMI to be tuned to the individual's level of recovery', with higher proportion of control allotted to BMI output for those with less ability to generate force on their own. With changing ability (e.g. improvement of strength over time), the algorithm could be varied to require a greater degree of hand control, and thus increase the physical challenge to build strength. Importantly, the inventors only input to the BMI HG signals from electrodes that were in the “edges” of the motor map - i.e.. those that correlated moderately with force during hand control (but not in the top 5-10% of electrode-force correlations).

[0035] Two patients used the BMI 1-3 times per week, 1 hour per session for 3-4 total sessions, plus conventional therapy, and 4 controls received conventional physiotherapy only. The inventors measured motor maps and function at the start of each session. Fig. 4 shows the results of BMI use producing HG synchrony. Fig. 4A depicts that a TBI subject used the BMI based on attempted thumb flexion force decoded from high-density. HG EEG in the hemicraniectomy area to control force feedback to the thumb via a custom-built device in accordance with an illustrative embodiment. Motor maps show R between HG and attempted thumb flexion force over 10 min in each session. The top of Fig. 4A shows maps at baseline,Atty. Dkt. No. 00100-0414-PCT days 3 and 4 for subject receiving BMI and occupational therapy (OT). The bottom of Fig. 4A shows maps for a control subject receiving only OT. Fig. 4B depicts the distribution of changes in R values with force, from baseline to last day, over all electrodes for BMI and control TBI subjects in accordance with an illustrative embodiment. BMI subjects had larger increases in R than controls. As a result, the motor map expanded more with the BMI training. Fig. 4C shows that during HG BMI control, HG peak was more tightly coupled to force feedback peak (HGlag between peaks: 130 ms) than mean / z power nadir during / z BMI control in 4 healthy subjects ( / z lag: 400 ms) in accordance with an illustrative embodiment.

[0036] These results from the experiments suggest that HG-based BMI-haptic feedback use can increase map core size, and this map size increase correlates with improvement of hand function (R=0.99 with the Action Research Arm Test). Furthermore, expansion occurred into the motor map edges in these subjects (light areas in Fig 4A). in contrast to the control subjects. Moreover, the results show that BMIs using HG as a control signal significantly improved the timing synchronization between neural modulation onset (BMI input) and haptic feedback, compared to using low-frequency EEG signals for BMI control (Fig. 4C). These results therefore indicate that using implanted electrodes (e.g., ECoG or epidural signals) to control a BMI with HG signals is a powerful method to drive plasticity and improve hand function, and better than low-frequency, standard EEG.

[0037] In addition to using motor maps as discussed above, the system can also be used to quantify plasticity by measuring motor evoked potentials (MEPs) by delivering direct electrical stimulation (DES) as described for motor maps, every' 2 months, but here using single pulses only. The MEP amplitudes can be measured using EMGs. Changes in the distribution of MEP response within 24 ms (i.e.. corticospinal tract-mediated) would indicate synaptic strength changes between Ml / premotor area (PM) and a-motoneurons due to BMI therapy. The system can also be used to examine the change in spatial extent of MEPs per stimulus site by summing MEP amplitudes over all EMG electrodes. The system can also be used to examine changes in cortical excitability using MEP threshold and MEP recruitment curve slopes.

[0038] Additionally, deep neural network (DNN) algorithms (starting with combined recurrent and convolutional) can be used to decode intended movement from ECoG. HGAtty. Dkt. No. 00100-0414-PCT power or a plurality of other frequency bands recorded during attempted movements can be used as input features to the DNN. The DNN can be updated as new data become available.

[0039] The decoder will update in 10-ms time bins in one embodiment, and the loss function will be designed to tune the decoder weights based on MEP latency such that it optimizes synchrony among cortices and a- motoneurons on fine timescales, which cannot be achieved with EEG BMIs. HG is a marker of local network synchronous firing; thus, if two networks synchronously firing are made to be in closer synchrony, this should provide conditions more conducive to plasticity.

[0040] The BMI decoding will progress according to stages, from hand open / close to grasp types, wrist extension / flexion, and thumb / index flex / ext. At each stage, decoders will be built from HG recorded while subjects attempt to perform instructed hand movements (displayed on a screen). The system will initially include all electrodes with reasonably high modulation in the decoder, to facilitate high-quality control and maintain patient engagement early on. The inventors will then gradually remove electrodes in the core of the motor map from the decoder, eventually using only those electrodes in the edges of the motor map (defined as those in frontal cortices with R values with movement in the 50th-95th %ile). This will guide the participant to learn to increase the modulation in these edge areas with each movement. By doing so, the system aims to increase the connectivity between motor map edges and other areas of motor / premotor / somatosensory cortices or a-motoneurons. After these adjustments, in the first 1-2 weeks of each stage, the decoder will be fixed to allow the participant to adapt to the BMI.

[0041] An FES control paradigm will use EMG electrodes, calibrated at the beginning of each session, to make the desired movements. Bony landmarks and henna markers can be used, as well as a 3D scanner, to ensure consistency across sessions. Since the EMGs will be in a similar position in each session, subsequent sessions require minimal fine-tuning to find these electrode combinations. The BMI decoder will translate HG to activate the FES pattern to produce the intended movement. To evoke graded movements, the BMI decoder output will be scaled via a transfer function to modulate the FES amplitude. The transfer function gain will remain constant over time to maximize a- MN recruitment and thereby optimize synchrony with cortical inputs.Atty. Dkt. No. 00100-0414-PCT

[0042] In an illustrative embodiment, any of the operations described herein can be performed by a computing device that includes a processor, memory, user interface, transceiver (e.g., a transmitter and a receiver), etc. For example, the operations can be stored as computer-readable instructions in the memory, and the processor can execute the computer-readable instructions to perform the operations described herein. For example, Fig. 5 is a block diagram of a computing system 500 to implement a neurorehabilitation system in accordance with an illustrative embodiment.

[0043] The computing system 500 is in communication with a network 535, a brain machine interface 540, and a functional electrical stimulation device 545. Alternatively, the functional electrical stimulation device 545 can be replaced by an exoskeleton or other haptic device. The computing system 500 can communicate directly with the brain machine interface 540 and the functional electrical stimulation device 545, or indirectly through the network 535. In one embodiment, the computing system 500 may be incorporated into the brain machine interface 540 and / or the functional electrical stimulation device 545. The computing system 500 includes a processor 505, an operating system 510, a memory 515, an input / output (I / O) system 520, a network interface 525, and a neurorehabilitation application 530. In alternative embodiments, the computing system 500 may include fewer, additional, and / or different components.

[0044] The components of the computing system 500 communicate with one another via one or more buses or any other interconnect system. The computing system 500 can be any type of networked computing device. For example, the computing system 500 can be a smartphone, a tablet, a laptop computer, a dedicated device specific to the neurorehabilitation application, etc.

[0045] The processor 505 can be in electrical communication with and used to control any of the system components described herein. The processor 505 can be any type of computer processor known in the art, and can include a plurality of processors and / or a plurality of processing cores. The processor 505 can include a controller, a microcontroller, an audio processor, a graphics processing unit, a hardware accelerator, a digital signal processor, etc. Additionally, the processor 505 may be implemented as a complex instruction set computer processor, a reduced instruction set computer processor, an x86 instruction setAtty. Dkt. No. 00100-0414-PCT computer processor, etc. The processor 505 is used to run the operating system 510, which can be any type of operating system.

[0046] The operating system 510 is stored in the memory 515, which is also used to store programs, user data, network and communications data, peripheral component data, the neurorehabilitation application 530, and other operating instructions. The memory 515 can be one or more memory systems that include various types of computer memory such as flash memory, random access memory (RAM), dynamic (RAM), static (RAM), a universal serial bus (USB) drive, an optical disk drive, a tape drive, an internal storage device, a non-volatile storage device, a hard disk drive (HDD), a volatile storage device, etc.

[0047] The I / O system 520 is the framework which enables users and peripheral devices to interact with the computing system 500. The I / O system 520 can include one or more displays (e.g., light-emitting diode display, liquid crystal display, touch screen display, etc.), a speaker, a microphone, etc. that allow the user to interact with and control the computing system 500. The I / O system 520 also includes circuitry and a bus structure to interface with peripheral computing devices such as power sources, USB devices, data acquisition cards, peripheral component interconnect express (PCIe) devices, serial advanced technology attachment (SATA) devices, high definition multimedia interface (HD MI) devices, proprietary connection devices, etc.

[0048] The network interface 525 includes transceiver circuitry (e g., a transmitter and a receiver) that allows the computing system to transmit and receive data to / from other devices such as the brain machine interface 540, the functional electrical stimulation device 545, other remote computing systems, servers, websites, etc. The data received from the brain machine interface 540 and the functional electrical stimulation device 545 can include captured brain signals, sensor data, etc. The network interface 525 enables communication through the network 535, which can be one or more communication networks. The network 535 can include a cable network, a fiber network, a cellular network, a wi-fi network, a landline telephone network, a microwave network, a satellite network, etc. The network interface 525 also includes circuitry to allow device-to-device communication such as Bluetooth® communication.

[0049] The neurorehabilitation application 530 can include software and algorithms in the form of computer-readable instructions which, upon execution by the processor 505,Atty. Dkt. No. 00100-0414-PCT performs any of the various operations described herein such as receiving signals, processing captured brain signals, generating motor maps, determining intended movements and force, receiving sensor data, analyzing the sensor data, generating signals to increase plasticity as discussed herein, etc. The neurorehabilitation application 530 can utilize the processor 505 and / or the memory 515 as discussed above. In an alternative implementation, the neurorehabilitation application 530 can be remote or independent from the computing system 500, but in communication therewith.

[0050] Fig. 6 shows an illustrative embodiment of a BMI, including ECoG electrodes (such as those made by Adtech, PMT, or other companies) that record the brain signals, amplifier that digitizes and conditions the signals, transmitter that sends the digitized signals to the receiver, either via cables or wirelessly. The receiver and BMI processor can be implanted under the skin or can be external to the body. The BMI processor decodes the signals and transmits them (either wired or wirelessly) to the receiver and processor for the output device. The output device can be a hand exoskeleton or FES system, as shown in Fig. 6. The FES system stimulates the electrodes, which can be external to the body, such as the Battelle Neuralife, or implanted, such as the Freehand system, which in turn stimulate the hand or arm muscles.

[0051] The word "illustrative" is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as "illustrative" is not necessarily to be construed as preferred or advantageous over other aspects or designs. Further, for the purposes of this disclosure and unless otherwise specified, "a" or "an" means "one or more.”

[0052] The foregoing description of illustrative embodiments of the invention has been presented for purposes of illustration and of description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. The embodiments were chosen and described in order to explain the principles of the invention and as practical applications of the invention to enable one skilled in the art to utilize the invention in various embodiments and with various modifications as suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.

Claims

Atty. Dkt. No. 00100-0414-PCTWHAT IS CLAIMED IS:

1. A system to rehabilitate people with stroke or traumatic brain injury, the system comprising: a brain machine interface that receives brain signals from a brain of a patient, wherein the brain signals represent motor intent of the patient with respect to a hand of the patient that has impaired dexterous hand function due to neurological injury; a limb interface that mounts to the hand of the patient, wherein the limb interface helps to control the hand in accordance with the motor intent of the brain signals, thus providing feedback from the hand to the central nervous system; and a processor operati vely coupled to the brain machine interface and the limb interface, wherein the processor is configured to synchronize the motor intent with the feedback from the hand to generate signals that increase plasticity between sensor}’ and motor areas of the brain to improve voluntary control of the hand, including dexterous hand functions such as multiple types of grasp.

2. The system of claim 1, wherein the brain signals represent an intended movement, muscle activation(s), or amount of force to be produced by the hand.

3. The system of claim 1, wherein the limb interface comprises a functional electrical stimulation device, an exoskeleton, or other haptic device.

4. The system of claim 1, wherein the processor also enables activity that increases plasticity between a motor cortex and spinal motoneurons.

5. The system of claim 1, wherein the brain machine interface is an intracranial device.

6. The system of claim 1, wherein the brain signals include one or more of electrocorticographic (subdural) signals, intracortical signals, epidural field potentials, intracortical local field potentials or action potentials, and intravascularly recorded signals.Atty. Dkt. No. 00100-0414-PCT7. The system of claim 1, wherein the processor generates a motor map of the brain that identifies one or more core regions that represent areas of the brain showing highest modulation with intended hand movement.

8. The system of claim 7, wherein the motor map also identifies one or more edge regions that represent areas of the brain showing only partial modulation with the intended hand movement.

9. The system of claim 8, wherein the processor controls the BMI within the edge regions to increase modulation of the one or more edge regions synchronously with sensory inputs such that the plasticity increases.

10. The system of claim 9, wherein the sensory feedback is controlled by both the BMI and by residual movement in the hand, and wherein a proportion of control may change over time as the patient regains more limb function.1 1. A method of rehabilitating people with stroke or traumatic brain injury, the method comprising: receiving, by a brain machine interface, brain signals from a brain of a patient, wherein the brain signals represent motor intent of the patient with respect to a hand of the patient that has impaired dexterous hand function due to neurological injury; controlling, by a limb interface that mounts to the hand of the patient, the hand in accordance with the motor intent of the brain signals, thus providing feedback from the hand to the central nervous system; and synchronizing, by a processor operatively coupled to the brain machine interface and the limb interface, the motor intent with the feedback from the hand to generate signals that increase plasticity between sensory and motor areas of the brain to improve voluntary control of the hand, including dexterous hand functions such as multiple ty pes of grasp.

12. The method of claim 11, wherein the brain signals represent an intended movement, muscle activation(s), or amount of force to be produced by the hand.Atty. Dkt. No. 00100-0414-PCT13. The method of claim 11, wherein the limb interface comprises a functional electrical stimulation device, an exoskeleton, or other haptic device.

14. The method of claim 11. wherein the processor also enables activity that increases plasticity between a motor cortex and spinal motoneurons.

15. The method of claim 11, wherein the brain machine interface is an intracranial device.

16. The method of claim 1 1, wherein the brain signals include one or more of electrocorticographic (subdural) signals, intracortical signals, epidural field potentials, intracortical local field potentials or action potentials, and intravascularly recorded signals.

17. The method of claim 11 , further comprising generating, by the processor, a motor map of the brain that identifies one or more core regions that represent areas of the brain showing highest modulation with intended hand movement.

18. The method of claim 17, wherein the motor map also identifies one or more edge regions that represent areas of the brain showing only partial modulation with the intended hand movement.

19. The method of claim 18, further comprising controlling, by the processor, the BMI within the edge regions to increase modulation of the one or more edge regions synchronously with sensory inputs such that the plasticity increases.

20. The method of claim 19, wherein the sensory feedback is controlled by both the BMI and by residual movement in the hand, and wherein a proportion of control may change over time as the patient regains more limb function.

Images