Systems and methods for processing generative or dynamic images and images to target healthy neural pathways.

Abstract

A system and method for modulating neural pathways to improve functional capacity in a subject is disclosed. The system and method may include the steps of: receiving a first dataset relating to the activity of one or more neural pathways of a subject; receiving a second dataset including at least one of at least one task performed by the subject, at least one symptom, and at least one medical condition of the subject; determining one or more abnormal neural pathways based on the first and second datasets; determining one or more healthy neural pathways associated with the one or more abnormal neural pathways based on one or more criteria from the second dataset; and determining focused or selective stimulation input to healthy neural pathways using a non-invasive stimulation device based on the healthy and abnormal neural pathways.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims priority to U.S. Provisional Application No. 63 / 501,334, filed May 10, 2023, which is hereby incorporated by reference in its entirety.

[0002] The present disclosure generally relates to the fields of neuroscience and neuroengineering, and more specifically to systems and methods for performing neural circuit modulation to restore, remap, or amplify functional capabilities in an individual with disrupted neural pathways.

Background Art

[0003] The human visual system relies on complex neural networks to process visual information and enable object recognition and localization. Disruption of these pathways (e.g., sensory pathways due to retinal or optic nerve diseases) can lead to functional impairments, either directly within the functional circuitry or in parallel, despite the neural and cortical networks remaining intact. Current treatment approaches often involve invasive procedures or rely on pharmaceutical interventions, which limit accessibility and effectiveness. Furthermore, conventional methods may not adequately address the dynamic nature of neuroplasticity, which offers opportunities for functional recovery through targeted stimulation of healthy neural circuits.

[0004] Thus, the present disclosure is directed to non - invasive and innovative approaches that leverage advances in neuroscience and technology to modulate neural circuits and restore functional capabilities in individuals with disrupted or dysfunctional pathways. The description of the background art provided herein is intended to present the context of the present disclosure in a general manner. Unless otherwise indicated herein, the materials described in this section are not admitted to be prior art to the claims of the present application by virtue of inclusion in this section, nor are they recognized as prior art or suggestions of prior art.

Summary of the Invention

[0005] In certain aspects of this disclosure, systems and methods for restoring the functional capacity of an individual with damaged or dysfunctional neural pathways are described. In one embodiment, a computer implementation method is provided for modulating neural pathways to improve the functional capacity of a subject. The computer implementation method may include: receiving a first dataset in a computing device associated with the activity of one or more neural pathways of a subject; receiving a second dataset in a computing device including at least one of at least one task performed by the subject, at least one symptom, and at least one medical condition of the subject; determining one or more abnormal neural pathways based on the first and second datasets using a processor associated with the computing device; determining one or more healthy neural pathways associated with one or more abnormal neural pathways based on one or more criteria using the processor; and determining a stimulus input to the healthy neural pathways using a non-invasive stimulation device based on the healthy and abnormal neural pathways. In another embodiment, a system is provided for modulating neural pathways to improve the functional capacity of a subject. The system may include one or more processors and one or more computer-readable media storing instructions executable by the one or more processors to perform an operation, the operation of which is to: receive a first dataset in a computing device associated with the activity of one or more neural pathways of a subject; receive a second dataset in a computing device including at least one of at least one task performed by the subject, at least one symptom, and at least one medical condition of the subject; determine one or more abnormal neural pathways based on the first and second datasets using a processor associated with the computing device; determine one or more healthy neural pathways associated with one or more abnormal neural pathways based on one or more criteria using a processor; and determine a stimulus input to a healthy neural pathway using a non-invasive or invasive stimulation device based on the healthy and abnormal neural pathways. In yet another embodiment, a non-temporary computer-readable medium for storing computer-executable instructions is provided. Non-temporary computer-readable media store computer-executable instructions, which, when executed by the system, cause the system to perform actions including: receiving a first dataset in a computing device associated with the activity of one or more neural pathways of a subject; receiving a second dataset in a computing device including at least one of at least one task performed by the subject, at least one symptom, and at least one medical condition of the subject; determining one or more abnormal neural pathways based on the first and second datasets using a processor associated with the computing device; determining one or more healthy neural pathways associated with one or more abnormal neural pathways from the second dataset based on one or more criteria using the processor; determining a stimulus input to healthy neural pathways using a non-invasive stimulation device based on healthy and abnormal neural pathways; and determining a stimulus input to healthy neural pathways, or across healthy and diseased pathways that have different effects on healthy pathways, using a non-invasive stimulation device based on healthy and abnormal neural pathways.

[0006] Additional objectives and advantages of the disclosed embodiments are, in part, set forth in the following description, some of which will become apparent from the description, or may be learned through practice of the disclosed embodiments. The objectives and advantages of the disclosed embodiments are realized and achieved by the elements and combinations specifically pointed out in the appended claims.

[0007] Please understand that both the above general description and the following detailed description are merely illustrative and descriptive, and do not limit the embodiments of this disclosure as claimed.

[0008] The accompanying drawings incorporated herein and constituting part of this specification illustrate several embodiments and, together with the description, help to illustrate the principles of this disclosure. [Brief explanation of the drawing]

[0009] [Figure 1A] An exemplary computer system for performing the methods described herein is shown.

[0010] [Figure 1B] This specification provides an exemplary software platform for performing the methods described herein.

[0011] [Figure 2] This disclosure illustrates an exemplary workflow for retraining the central nervous system in general clinical use, according to one or more embodiments of this disclosure.

[0012] [Figure 3] This disclosure provides an exemplary workflow for training a machine learning model that may be used in the application of non-invasive neural stimulation to retrain visual processing pathways in general clinical applications, according to one or more embodiments of this disclosure.

[0013] [Figure 4] This disclosure illustrates an exemplary computing system according to one or more embodiments of this disclosure. [Modes for carrying out the invention]

[0014] The terms used below, even when used in conjunction with the detailed descriptions of specific examples in this disclosure, should be interpreted in the broadest and most reasonable manner. In fact, any term intended to be interpreted restrictively, even if emphasized below, is explicitly and specifically defined in this “Modes for Carrying Out the Invention” section. Both the general description above and the detailed description below are merely illustrative and explanatory and do not limit the claimed features.

[0015] In this disclosure, the term “based on” means “at least partially based on.” Unless otherwise specified by context, the singular forms “a,” “an,” and “the” include plural referents. The term “exemplary” is used to mean “example,” not “ideal.” The terms “comprises,” “comprising,” “includes,” and “including,” or other variations thereof, are intended to cover non-exclusive inclusion; therefore, a process, method, or product that includes a list of elements may not necessarily include only those elements, but may also include other elements not expressly enumerated, or other elements not specific to such process, method, article, or apparatus. Relative terms such as “about,” “approximately,” “substantially,” and “generally” are used to indicate a possible ±10% variation in the stated or understood values. Furthermore, the term “between” used when describing a range of values ​​is intended to include the minimum and maximum values ​​described herein. The use of the term “or” in the claims and specification is used to mean “and / or” unless it is expressly indicated that it refers only to alternatives, or unless the alternatives are mutually exclusive; however, this disclosure supports the definition that refers only to alternatives and “and / or.” As used herein, “another” may mean at least a second or more.

[0016] As used herein, the term “User” generally includes any individual or entity, such as a researcher and / or healthcare professional (e.g., a physician), who may desire information, a solution to a problem, or engage in any other type of interaction with the providers of the systems and methods described herein (e.g., through an application interface present on an electronic device). The terms “Electronic Application” or “Application” may be used interchangeably with other terms such as “Program,” and generally include software configured to interact with, modify, overwrite, supplement, or operate in conjunction with other software.

[0017] Damage to neural pathways, particularly within the visual system, presents significant challenges in restoring function in individuals affected by various conditions. For example, retinal diseases and optic nerve damage can be debilitating conditions, often resulting in severe visual impairment. Even if the complex network of neural connections and cortical processing centers in the brain remains intact, loss of function in these pathways can significantly impact an individual's ability to recognize and interpret visual stimuli. This impairment can not only affect daily activities but also reduce overall quality of life. For instance, in conditions such as macular degeneration or limb loss, the cortical circuits that process visual or tactile information may be intact. However, afferent neural circuits may be dysfunctional or partially absent. For example, a subject may have a scotoma due to central visual field loss (CFL) associated with macular degeneration, which is a partial loss of vision or blind spot in the normally functioning visual cortex. While the subject may have functional vision around the scotoma, the presence of a blind spot in the center of the visual field can still impair object recognition, object manipulation, and other activities of daily living (ADL). Around a scotoma, there may be an immediate surrounding visual area and a larger surrounding visual area.

[0018] Conventional attempts to address these problems have relied primarily on invasive procedures or drug interventions. For example, surgical interventions, while effective in some cases, carry inherent risks and may not be suitable for all subjects. Drug therapy, on the other hand, often alleviates symptoms but may not address the underlying neurological dysfunction. Furthermore, conventional techniques may not fully utilize the dynamic nature of neuroplasticity, which offers opportunities for functional recovery through targeted stimulation of healthy neural circuits. Static stimulation protocols may overlook individual variations in neuronal responsiveness and fail to adapt to changes in neuronal activity over time. Moreover, the lack of individualization and neuronal stratification in conventional approaches may result in suboptimal outcomes for subjects with diverse neurological conditions.

[0019] To address the aforementioned issues, this disclosure aims to propose a novel technique in neural circuit modulation by focusing on non-invasive, personalized stimulation methods tailored to the characteristics and responsiveness of individual subjects. By integrating machine learning algorithms and advanced data analysis techniques, the concepts described herein can optimize stimulation parameters based on real-time subject feedback and monitoring of neural activity. This dynamic approach allows the system to adapt and improve over time to the stimulation pattern, thereby maximizing effectiveness and subject outcomes.

[0020] More specifically, in cases of visual impairment, if a specific neural pathway is dysfunctional or impaired due to a neurological condition such as retinal or optic nerve disease, the brain may rely on alternative pathways to compensate for the loss of function. These alternative pathways are often referred to as “healthy,” “intact,” or “normal” neural circuits, and may also be capable of processing sensory information or generating motor commands. Using neural stimulation techniques, these healthy neural pathways can be activated, essentially rerouting neural signals to bypass the defective pathways. By stimulating brain regions associated with intact sensory processing, motor control, or cognitive function, neural responses contributing to the desired functional outcome can be induced. Through repeated stimulation and training, the brain can undergo a process of functional reorganization or neural plasticity. This adaptive mechanism allows the brain to optimize its neural connections and utilize available neural resources for functional recovery, thereby compensating for the defects of the dysfunctional pathways. Ultimately, the goal of targeting healthy neural pathways is to enable the brain to achieve the same, or at least partial, functional outcomes or tasks that would normally be mediated by the defective pathways. Whether restoring vision, motor coordination, cognitive function, or sensory processing, neurostimulation therapy can leverage the brain's inherent plasticity to promote recovery and functional improvement through available healthy neural networks and circuits. Among visual impairments, in age-related macular degeneration, the loss of retinal photoreceptors leads to a loss of visual input, but the retinal cortical pathways remain intact. Training the preserved cortex can fill the blank. In glaucoma, damage to retinal ganglion cells and the optic nerve results in visual field loss. However, the cortex retains function, and plasticity allows it to map new inputs. In optic neuritis (multiple sclerosis), inflammation of the optic nerve leads to impaired signal transmission. However, the cortex is not damaged and can be compensated for with retraining. In occipital lobe stroke, direct damage to the visual cortex results in blindness. However, the plasticity of preserved cortex allows for functional remapping. Hemispatial neglect and hemispatial neglect result in visual field loss after stroke / trauma. Preserved cortex can be trained to fill in the input. Memory loss results in loss of face recognition.The preserved cortex can be trained to fill in the input. Amblyopia causes visual developmental disorders, but the cortex remains intact. Plasticity may enable suppression of processing. In neurodegenerative disorders, posterior cortical atrophy (a subtype of Alzheimer's disease), occipital / temporal lobe dysfunction impairs visual processing, which can be compensated by suppression. Dementia with Lewy bodies, alpha-synuclein lesions, impair visual processing. The preserved cortex can be retrained. In Parkinson's disease, progressive visual impairment results from dopamine loss. The preserved cortex can be trained to adapt. In Huntington's disease, final visual cortex dysfunction occurs. However, retraining may be beneficial for visual processing while plasticity remains.

[0021] Features of the concepts of the present invention include the use of non-invasive stimulation modalities such as visual, auditory, tactile, electrical, and / or multimodal stimulation delivered by specialized stimulation devices. These stimulations can be accurately calibrated to target specific healthy neural pathways or neural pathways exceeding healthy thresholds identified by comprehensive subject evaluations including brain imaging, electrophysiological recordings, and clinical assessments. Also, machine learning algorithms can play an important role in adjusting stimulation protocols according to individual needs and responses by analyzing large amounts of subject-specific data. Furthermore, the concepts described herein can be integrated with existing treatment modalities and complement pharmacological treatments and surgical interventions. Thus, the systems and methods described herein constitute a promising means for restoring the functional capabilities of individuals with disrupted neural pathways while minimizing the risks and limitations associated with invasive procedures and static treatment protocols.

[0022] In some embodiments, the collective concepts presented herein provide specific and obvious applications in neuroscience and neuroengineering. Furthermore, these concepts represent improvements in computer technology by introducing innovative applications to the common ground of neuroscience and computing. For example, processing and analyzing complex neural data requires advanced computational techniques and algorithms. The concepts described herein can leverage advances in data analysis and processing capabilities to extract meaningful insights from large neural datasets and enable personalized therapeutic approaches tailored to the characteristics of individual subjects. Furthermore, by tailoring stimulation protocols to individual characteristics, including neural activity patterns and responsiveness, the systems and techniques described herein ensure optimal therapeutic outcomes for each subject. Moreover, the concepts described herein address important and specific problems in the field of neurology, namely, the restoration of functional capacity in individuals with damaged neural pathways. Conditions such as retinal diseases and optic nerve damage have a clear debilitating impact on a subject's vision and quality of life. By providing innovative solutions that directly target these challenges, the concepts of the present invention bring practical and meaningful benefits to subjects and clinicians. More specifically, by systematically targeting and stimulating healthy nerve pathways, the effects of abnormal firing pathways can be reduced and / or functional pathways can be enhanced. As a result of prioritizing functional information over information about dysfunction and / or the gate control / blocking effect in the dysfunctional circuit, pain and / or symptoms may be reduced and function restored, depending on the medical condition.

[0023] Here, the subject matter of the present disclosure will be more fully described while referring to the accompanying drawings that form a part of this specification and illustrate certain exemplary embodiments as explanatory figures. Embodiments or aspects described as "exemplary" herein should not be construed, for example, as being more preferable or advantageous than other embodiments or aspects, but are intended to reflect or indicate that the embodiment(s) is / are "exemplary" embodiments. The subject matter can be embodied in various different forms, and thus, it is intended that the subject matter to be protected or claimed is not limited to any of the exemplary embodiments described herein, and the exemplary embodiments are provided merely for the purpose of explanation. Similarly, it is intended that the scope of the claimed or protected subject matter extends quite broadly. For example, the subject matter can be embodied as, among other things, a method, a device, a component, or a system. Accordingly, the various embodiments can take the form of, for example, hardware, software, firmware, or any combination thereof. Therefore, the following detailed description is not intended to be construed in a limiting sense.

[0024] Throughout this specification and the claims, terms can have subtle meanings suggested or implied in the context beyond the explicitly stated meaning. Similarly, the phrases "in one embodiment" or "in some embodiments" or "in one aspect" or "in some aspects" used herein do not necessarily refer to the same embodiment or aspect, and the phrases "in another embodiment" or "in another aspect" used herein do not necessarily refer to different embodiments or aspects. For example, it is intended that the claimed subject matter includes, in whole or in part, combinations of the exemplary embodiments.

[0025] Specific embodiments are described herein, but are not intended to be limiting. Other uses may include migraines, dysfunctional neuropathy, vagus or trigeminal nerve disorders, motor dysfunction resulting in tremors, or other situations in which abnormally firing afferent neurons can negate normal brain feedback or concentrate information in the brain, potentially causing abnormal physiological / autonomous responses such as pain, nausea, and hypotension, discomfort, and / or functional impairment. Furthermore, while specific embodiments described herein may improve dysfunction caused by ocular disorders, other embodiments may utilize the normal afferent function of the eye to treat conditions appearing in other bodily systems and other locations within the body. As described, healthy neural circuits within a given proximity range of an abnormal neural circuit may be targeted. This proximity may be located in processing centers in the subcortex and / or cortex. This proximity may be located at any point in the neural pathway, including within the brain itself. In cases of phantom limb pain, for example, the ocular nerve pathway may be normal and therefore there is no afferent disturbance; however, processing from non-invasive visual stimulation devices may act to interrupt or gate the signal processing of the overlapping system, and abnormal afference associated with the missing limb transmits erroneous pain signals. Thus, normally functioning circuits can alleviate symptoms by overwriting or interrupting abnormal dysfunction signals. This interruption can occur in the nuclear layer, ganglia, and / or cortical layer, and can consequently affect cortical interpretation through signal remapping or prioritization.

[0026] Figure 1A shows an exemplary system capable of performing the method described herein. The exemplary system 100 includes a data acquisition component 10, a database 20, and a device data intelligence component 30, which are operably connected to each other via a network 40. Alternatively or additionally, one or more of the components may be connected locally to another component, for example via a wired connection, without relying on the network connection.

[0027] As disclosed herein, the data acquisition component 10 may include a device or machine capable of measuring electrical activity in the brain. In some embodiments, the data acquisition component 10 may be an electroencephalograph (EEG) machine that includes or is configured to support one or more electrodes, amplifiers, filters, analog-to-digital converters, etc., for performing an EEG test. Other devices, such as a magnetoencephalography (MEG) device, may be used. In some embodiments, the data acquisition component 10 may be a database that receives EEG test data from one or more other sources. In other embodiments, the data acquisition component 10 may be any other brain recording device or modality capable of transmitting information about neural activity. Consumer virtual and augmented reality headsets that integrate the data acquisition component into the headset may be another form of device available.

[0028] The data acquired by the data acquisition component 10 can be transferred to the database 20 via the network 40, or via a direct local or network connection. In some embodiments, the acquired data can be analyzed by the data intelligence component 30 via the network 40 or a local or network connection. Figure 1B shows an exemplary functional module that may be implemented to perform the tasks of the data intelligence component 30.

[0029] Figure 1B shows an exemplary computer system 110 for using the techniques described herein, for example, to retrain a visual processing path. The exemplary system 110 can implement the techniques described herein by implementing a user input / output (I / O) module 120, memory or database 130, data processing module 140, data analysis module 150, classification module 160, network communication module 170, and any other functional modules that may be required to perform a particular task (e.g., an error correction or compensation module, a data compression module, etc.) on one or more computer devices. These modules may correspond to the modules in Figure 1A. For example, database 130 may correspond to database 20, modules 140, 150, 160, and 170 may correspond to data intelligence 30, and the input configuration of module 120 may correspond to data collection 10. As disclosed herein, the user I / O module 120 may further include input submodules such as a keyboard, MEG, EEG, eye-tracking data, etc., and output submodules such as a display (e.g., a printer, television, smartphone, monitor, virtual reality (VR) device, and / or touchpad). In some embodiments, all functions may be performed by a single computer system. In some embodiments, functions may be performed by multiple computer systems. Various modules (e.g., modules for data processing, analysis, classification, communication, etc.) may be one or more processes running in a distributed computing environment. For example, in some embodiments, one or more components of computer system 110 may be network-accessible via a cloud infrastructure. For example, the database 130 used to store data may be stored on one or more remote cloud servers. In this regard, the database may be one or more large storage buckets (e.g., cloud-based storage buckets such as a simple storage service "S3" bucket) from which data can be retrieved on demand.As another example, data processing, analysis, and classification can be performed in a cloud-based environment using services such as cloud-based data processing platforms, serverless computing, and cloud-based machine learning platforms.

[0030] Furthermore, this specification discloses that certain tasks can be performed by implementing one or more functional modules. Specifically, each of the listed modules may consequently include multiple submodules that implement one or more techniques described herein. For example, the data processing module 140 may include submodules for data quality assessment (e.g., submodules for iterative refinement and verification), submodules for normalizing any assigned weights to ensure that those weights contribute proportionally to the overall response, submodules for interpolation or extrapolation, and so on.

[0031] In some embodiments, a user may use the I / O module 120 to manipulate either data available on the local device or data obtainable via a network connection from a remote service device or another user device. For example, the I / O module 120 may enable a user to perform data analysis via a graphical user interface (GUI), for example, via a keyboard, mouse, or touchpad. In some embodiments, a user may manipulate data via voice control. In some embodiments, user authentication may be required before the user is granted access to the requested data. In some embodiments, the user I / O module 120 may be used to manage various functional modules. For example, while an existing data processing session is in progress, a user may request input data via the user I / O module 120. The user may do this by selecting a menu option or by individually typing a command, without interrupting the existing process. In another example, a user may use the user I / O module 120 to provide the computer system 110 with instructions to set various thresholds, configure sample matching settings, and / or how to capture and / or monitor electrical signals in the brain. As disclosed herein, a user may direct and control data processing and analysis via the I / O module 120 using any type of input.

[0032] In some embodiments, the system 110 further includes memory and / or a database 130. In some embodiments, the database 130 includes a local database that can be accessed via a user I / O module 120. In some embodiments, the database 130 includes a remote database that can be accessed by the user I / O module 120 via a network connection. In some embodiments, the database 130 is a local database that stores data acquired from another device (e.g., a user device or a server). In some embodiments, the memory or database 130 may store data acquired in real time from an internet search. In some embodiments, the database 130 may send and receive data to and from one or more other functional modules, which include, but are not limited to, a data acquisition module (not shown), a data processing module 140, a data analysis module 150, a classification module 160, a network communication module 170, etc. In some embodiments, some or all real sample data and / or synthetic sample data may be stored in the database 130.

[0033] In some embodiments, database 130 may be a local database to other functional modules. In some embodiments, database 130 may be a remote database that can be accessed by other functional modules via a wired or wireless network connection (for example, via a network communication module 170). In some embodiments, database 130 may include a local portion and a remote portion.

[0034] In some embodiments, the system 110 includes a data processing module 140. The data processing module 140 may receive real-time data from the I / O module 120 or the database 130. In some embodiments, the data processing module 140 may perform one or more standard data processing algorithms, such as noise reduction, signal enhancement, normalization, interpolation, and / or extrapolation. In some embodiments, the data processing module 140 may be configured to process received and / or collected neural activity data associated with one or more subjects. In various embodiments, the data processing module 140 may further create a training dataset, on which one or more machine learning models (e.g., models for classification, clustering, scoring, etc.) may be trained.

[0035] In some embodiments, the system 110 includes a data analysis module 150. In some embodiments, the data analysis module 150 includes identifying brain activity patterns associated with specific medical conditions, as described in relation to the data processing module 140.

[0036] In some embodiments, the system 110 includes a classification module 160 that can embody a “machine learning model” or “trained classifier.” As used herein, a “machine learning model” or “trained classifier” generally includes instructions, data, and / or a model configured to receive an input and apply one or more weights, biases, classifications, or analyses on the input to produce an output. The output may include, for example, a classification of the input, an analysis based on the input, a design, process, prediction, and / or recommendation associated with the input, or any other suitable type of output. A machine learning model is generally trained using training data, e.g., empirical data and / or samples of input data, fed to the model to establish, adjust, or modify one or more aspects of the model, e.g., weights, biases, criteria for forming classifications or clusters, etc. Aspects of a machine learning model may operate on an input linearly, in parallel, via a network (e.g., a neural network), or via any suitable configuration.

[0037] The execution of a machine learning model(s) may involve the deployment of one or more machine learning techniques, such as k-nearest neighbors, linear regression, logistic regression, random forests, gradient boosting machines (GBMs), deep learning, deep neural networks (e.g., recurrent neural networks (RNNs), convolutional neural networks (CNNs), Transformers), and / or any other suitable machine learning techniques. Supervised training, semi-supervised training, and / or unsupervised training may be employed. For example, supervised learning may involve providing training data and corresponding labels, for example, as ground truth. Unsupervised methods may include clustering or classification. K-means clustering or K-nearest neighbors may also be used, and these may be supervised or unsupervised. Combinations of K-nearest neighbors and unsupervised clustering techniques may also be used. For example, any suitable type of training may be used, such as stochastic, gradient boosting, random seeding, recurrent, epoch-based, or batch-based.

[0038] Furthermore, the techniques described herein may be implemented using multiple machine learning models that can be executed sequentially and / or in parallel. For example, a first machine learning model may detect and / or interpret neurological signals from a patient, and a second machine learning model may generate neurostimulation images to be presented to the patient in order to achieve a desired outcome.

[0039] Neuronal stimulation images can be generated using deep learning models. These deep learning models may have pre-trained weights, or the weights may be learned from training on collected datasets that can combine visual stimuli, neural recordings, and behavioral recordings. Visual stimuli can be generated from deep learning models that simultaneously generate groups of video frames from stimulus parameters in a closed-loop manner, and / or visual stimuli can be generated frame by frame, conditional on changing neural data being recorded in real time. Visual stimuli can also be generated from pre-specified visual features, such as grid patterns or white noise, or from a combination of pre-specified visual features and features generated by the deep learning model.

[0040] In exemplary use cases, a machine learning model is trained to analyze test data from a subject whose specific neural activity related to a medical condition may be unknown, and then identify parts or characteristics of the subject's brain that may be the cause of or result of the medical condition. In some embodiments, one or more parameters may include a score (e.g., a binomial probability score that can be calculated based on logistic regression analysis). As disclosed herein, the binomial probability score may correspond to the likelihood that the subject has a particular medical condition, the likelihood that parts of the subject's brain are active or inactive, the likelihood that a particular stimulus affects a desired part of the brain, etc. For example, a score above a predefined threshold may indicate that a particular stimulus, or sequence or set of stimuli, is effectively stimulating a non-sensory area of ​​the brain.

[0041] As disclosed herein, the network communication module 170 can be used to facilitate communication between a user device, one or more databases, and any other suitable systems or devices via a wired or wireless network connection. Any communication protocol / device may be used, including, but not limited to, modems, Ethernet® connections, network cards (wireless or wired), infrared communication devices, wireless communication devices, and / or chipsets (Bluetooth® devices, 802.11 devices, WiFi devices, WiMAX devices, cellular communication equipment, etc.), near-field communication (NFC), Zigbee® communication, radio frequency (RF) or radio frequency identification (RFID) communication, PLC protocols, 3G / 4G / 5G / LTE-based communication, etc. For example, a user device having a user interface platform for processing / analyzing tumor fraction data may communicate with other user devices with the same platform, ordinary user devices without the same platform (e.g., ordinary smartphones), remote servers, physical devices on a remote IoT local network, wearable devices, user devices responsively connected to the remote server, etc.

[0042] The technologies disclosed herein may be used in combination with the technologies described in U.S. Patent No. 10,736,526 and U.S. Application No. 18 / 044,054, which are each incorporated herein by reference as a whole.

[0043] The functional modules described herein are provided as examples. It will be understood that different functional modules can be combined to create different utilities. It will also be understood that additional functional modules or submodules can be created to implement a particular utility.

[0044] General clinical use Referring here to Figure 2, an exemplary workflow 200 for retraining the nervous system in a general clinical application is provided using the embodiments described herein. The exemplary workflow 200 may be carried out according to some or all of the components described in Figures 1A and 1B.

[0045] In step 205, the system may receive electronic and / or magnetic data related to the activity of one or more neural circuits in the subject. This data may include various types of signals associated with brain activity, including electrical, magnetic, and / or blood flow / oxygenation activity. For example, electrical activity may include signals representing electrical impulses generated by nerve cells in the brain, which can be captured using devices such as EEG. Magnetic activity may include magnetic fields generated by neural activity, which can be measured using techniques and devices such as MEG. Blood flow / oxygenation activity may include changes in blood flow and / or oxygenation in different areas of the brain, which can indicate neural activity. These changes can be captured using techniques such as functional magnetic resonance imaging (fMRI). In some embodiments, the subject may need to undergo certain preparations before data acquisition, depending on the imaging technique being used. For example, in the case of EEG recording, electrodes may be placed on the skin after the skin has been cleansed and prepared. In fMRI examinations, the subject may need to remain still in the MRI machine while data is being acquired. Functional near-infrared spectroscopy (fNIRS), which uses near-infrared light to measure changes in cerebral blood flow by quantifying changes in hemoglobin concentration in the brain based on light intensity measurements, can also be utilized.

[0046] In step 210, the system may receive a second dataset containing data on the symptoms and / or medical conditions of one or more subjects. More specifically, to facilitate this identification, after collection, the raw data may be preprocessed to extract relevant information and remove noise. This preprocessing may include filtering, artifact removal, and / or other signal analysis techniques. The preprocessed data may then be stored for further analysis. In some embodiments, data analysis may include comparing signals across different brain regions, identifying patterns of activity, and correlating activity with specific stimuli or tasks. For example, comparison may include comparing a subject's brain activity to established criteria or reference data to determine deviations indicating healthy and / or abnormal neural pathways. Based on symptoms or signs of a medical condition, one or more abnormal neural pathways may be identified. This may include analyzing data from imaging devices or other diagnostic tools to pinpoint areas of dysfunction. In some cases, abnormal pathways may not be explicitly identified, and interventions may be automatically derived from the subject's brain activity without explicit interpretation of abnormalities or pathways.

[0047] In step 210, the system may receive a second dataset associated with the subject's performance on motor and / or cognitive tasks. This dataset may be acquired during the original data capture session, or at one or more subsequent times, and / or during one or more subsequent data capture sessions. In some embodiments, during the data capture of the second dataset, the subject may be asked to perform one or more motor and / or cognitive tasks, which may help in strengthening or targeting healthy and / or abnormal pathways in step 215. Performance on motor and cognitive tasks serves multiple purposes in relation to non-invasive neural stimulation. For example, performing a task involves specific neural pathways associated with motor or cognitive function. This activation helps to enhance these pathways and increase their sensitivity to subsequent stimulation. Furthermore, by observing the subject's response to the task, a clinician can identify both normal and abnormal neural pathways. Healthy pathways show appropriate responsiveness to the task, while abnormal pathways may show changes or impairments in function. If an abnormal pathway is not explicitly identified, automatically derived interventions may utilize data on task outcomes in addition to neural data.

[0048] In some embodiments, motor tasks may include physical movements or actions performed by the subject. These can range from simple actions such as tapping with fingers or clenching a fist to more complex movements such as walking or reaching for an object. Motor tasks involve motor-related neural circuits and can help assess motor function and coordination. Cognitive tasks, on the other hand, may include mental processes such as attention, memory, language, and executive function. Subjects may be asked to perform tasks such as memory retrieval, problem solving, or concentration. Cognitive tasks involve higher-order cognitive neural circuits and can provide insights into cognitive function and processing. In some embodiments, the selection of tasks performed may be tailored to the subject's symptoms and medical condition. For example, a subject with a visual impairment may be asked to perform tasks related to visual processing, such as object recognition or spatial navigation. In some embodiments, the tasks should involve neural pathways relevant to the subject's condition. For example, motor tasks may target brain regions associated with motor control, while cognitive tasks may target areas involved in visual processing or attention. The tasks should enable subjects to perform safely within a clinical setting and should be appropriate to the subjects' age, health status, and cognitive abilities. During the performance of motor and cognitive tasks, the subjects' behavior and responses may be observed. This may include evaluating cognitive ability metrics such as the quality, accuracy, and speed of movement, or reaction time or precision on cognitive tasks. In some embodiments, different tasks (e.g., walking, talking) may be combined, as a stimulating effect may be possible by engaging different but readily activated parts of the brain to increase resting-state stimulus sensitivity.

[0049] In step 220, the system may utilize a second and / or first dataset to identify one or more neural pathways in the brain that are functioning normally and are suitable targets for neural stimulation. The selection of these pathways may be based on one or more criteria. For example, healthy neural pathways selected for stimulation may be adjacent to identified abnormal pathways or areas of dysfunction. Stimulating healthy pathways adjacent to abnormal areas may enable modulation of neural activity and restoration of function. For example, in subjects with visual impairment due to retinal or optic nerve disease, healthy neural pathways associated with intact visual processing areas may be targeted to compensate for the deficits in the dysfunctional areas. In another embodiment, the selected healthy pathways should be functionally relevant to the subject's symptoms and medical condition. Stimulation pathways that contribute to the desired functional outcome may be important for achieving therapeutic effects. Additionally or alternatively, in another embodiment, the feasibility of stimulating the identified pathways using non-invasive neural stimulation techniques is described. Factors such as accessibility, safety, and potential therapeutic effects may influence the selection of target sites. Additionally or alternatively, in another aspect, the selected pathways may offer therapeutic potential, meaning that stimulation of these pathways is likely to result in meaningful improvement in the subject's symptoms and functional outcomes. Additionally or alternatively, in another example, the safety of neurostimulation interventions may be optimized by selecting only pathways with a favorable safety profile and avoiding areas associated with potential adverse effects or unintended consequences. Additionally or alternatively, in yet another aspect, each subject's individual characteristics, including medical history, neuroanatomical structure, and response to previous treatments, may be considered to modify the selection of healthy pathways to suit the subject's specific needs and circumstances. In some cases, healthy and abnormal pathways may not be clearly identifiable. Automated interventions may infer effective stimulation interventions using data from individuals exhibiting dysfunction and, optionally, data from healthy individuals.

[0050] In some embodiments, one or more methods / techniques may be employed in this data identification step. For example, advanced imaging techniques such as magnetic resonance imaging (MRI), functional MRI (fMRI), positron emission tomography (PET), or diffusion tensor imaging (DTI) can provide detailed insights into brain structure and function. These imaging modalities can be used to visualize neural pathways and identify brain regions associated with healthy function. In other embodiments, electrophysiological methods such as EEG or MEG can measure electrical activity in the brain. By recording neural signals, patterns of activity associated with healthy neural pathways can be identified. In some cases, this step may be bypassed if an automated, purely data-driven approach is used.

[0051] In step 225, neurostimulation techniques for application to identified healthy neural pathways may be determined and / or output to modulate neural activity and promote functional improvement in subjects with neurological disorders. By targeting these pathways, neurological function may be restored or enhanced, symptoms may be alleviated, and overall quality of life may be improved. Neurostimulation may be facilitated by various types of non-invasive neurostimulation. For example, in one embodiment, transcranial magnetic stimulation (TMS) may be used to deliver magnetic pulses to specific areas of the brain to induce currents that modulate neural activity. Traditionally, TMS has been used to stimulate cortical areas associated with motor function, cognition, or mood regulation. In another embodiment, transcranial direct current stimulation (tDCS) may be used to apply low-intensity direct current to the skin to modulate the excitability of neurons. This approach may enhance or suppress neural activity in targeted brain regions and has traditionally been used for cognitive enhancement or mood regulation. In some cases, stimuli may not explicitly target identified healthy pathways, but instead may attempt to converge on patterns of neural activity that are empirically associated with improved functional performance without human interpretation.

[0052] In some embodiments, the frequency of a stimulus can be varied (for example, by or based on a trained machine learning model, as further described herein) to produce different effects on neuronal excitability. For example, high-frequency stimulation may enhance synaptic plasticity and cortical excitability, while low-frequency stimulation may produce an inhibitory effect. In some embodiments, stimuli delivered in a specific time pattern, such as bursts or a series of pulses, may affect the synchronization of neuronal activity and induce changes in the plasticity of neural circuits. In some embodiments, a closed-loop system may dynamically adjust the parameters of a stimulus by integrating real-time feedback from physiological signals or neuronal activity.

[0053] In step 230, if a proximal healthy neural pathway is not stimulated by the first stimulus, or if the stimulus becomes less effective over time, an additional stimulus may be determined and provided to attempt to stimulate a healthy neural pathway or another healthy neural pathway. Stimulation of a healthy neural pathway may be sustained for a predetermined period of time based on subject feedback and / or updates to electronic subject information. For example, electronic subject information may reflect desired retraining of neural pathways in the subject's brain, at which point the stimulation may be terminated.

[0054] In step 235, a biofeedback loop can be incorporated into the neuronal stimulation process to optimize treatment by providing real-time data on neuronal activity, physiological responses, or behavioral changes during stimulation. This feedback mechanism can be used to dynamically adjust stimulation parameters and tailor the therapeutic intervention based on the individual subject's response.

[0055] In some embodiments, biofeedback data may be collected by various monitoring modalities described above, including neuroimaging, electrophysiological recording, physiological sensors, and / or subject-reported outcomes. These data sources capture relevant information regarding neuronal activity, physiological parameters, symptoms, and functional performance. In some embodiments, biofeedback data may be provided as input to a trained machine learning model (further described herein), which may be configured to process the data in substantially real time to extract meaningful insights into the subject's response to neuronal stimulation (e.g., treatment effect, subject progress, etc.). Various algorithms and analytical techniques may be employed to identify patterns, trends, and anomalies in the data to facilitate rapid decision-making (e.g., adjustment of stimulation parameters, treatment protocols, etc.) and adaptive care.

[0056] These techniques can be used, for example, to treat and / or alleviate hallucinations or other symptoms experienced in Charles Bonnet syndrome or dementia by stimulating healthy areas of the brain. In other words, by reactivating suppressed, normal, and functional neural circuits that were otherwise blocked or isolated from normal activity, the resulting detrimental functions are activated.

[0057] Training and development of general models Referring here to Figure 3, the exemplary workflow 300 is provided for training a machine learning model that may be used in the application of non-invasive neural stimulation to retrain visual processing pathways. Embodiments of the exemplary workflow 300 may be carried out according to some or all of the components described in Figures 1A and 1B, as well as some or all of the processes described in exemplary workflow 200.

[0058] In one embodiment, a machine learning algorithm may be employed to analyze electronic data obtained from subjects, including neural activity captured by imaging devices such as EEG or MEG, and performance on behavioral tasks affected by the desired dysfunction. By processing this data, the algorithm can identify patterns and correlations unique to each subject's condition. This personalized approach allows for the customization of neural stimulation protocols to suit individual neural responses and symptoms. For example, the algorithm may determine the optimal type, intensity, and duration of stimulation based on the subject's unique neural profile.

[0059] In step 305, a first set of training data may be collected. In one embodiment, the first set of training data may correspond to inputs provided by a non-invasive stimulation device to multiple subjects receiving neurostimulation therapy. More specifically, each data point in the first set of training data corresponds to a specific stimulus presented to a subject, along with relevant metadata such as the type of stimulus, its duration, intensity, frequency, and other parameters.

[0060] In step 310, a second set of training data may be collected. In one embodiment, the second training dataset may correspond to the response effect of stimuli from step 305 on one or more neurons and / or neural circuits of a subject. This data provides information about how neural activity patterns change in response to various types of stimuli delivered by the stimulating device. Similar to the first set of training data, each data point in the second set is paired with the stimuli input that induced the neural response. However, the focus of the second set is not on the stimuli themselves, but on characterizing the observed changes in neural activity.

[0061] Therefore, the neural responses induced by the stimuli specified in step 305 are recorded using an imaging device such as EEG or MEG, yielding paired data samples consisting of the stimulus input from step 310 and the corresponding neural activity response. Both of these datasets are used in parallel to train a machine learning model to predict and analyze neural responses to non-invasive neural stimulation and optimize treatment strategies tailored to individual subjects.

[0062] In step 315, the collected data may be annotated and labeled to provide supervision for the machine learning algorithm. The supervised learning algorithm relies on the annotated data to identify correlations between input features (e.g., stimuli) and output labels (e.g., neural responses), generalize them from known examples, and enable prediction of undiscovered data. By providing labeled examples during training, the supervised learning algorithm optimizes its performance on a given task by iteratively tuning its internal parameters to minimize prediction errors.

[0063] In some embodiments, step 315 may include a manual or automated process of marking specific segments or features in the collected data with labels or annotations. Depending on the nature of the data, annotation may be performed by a domain expert, a trained annotator, or by an automated algorithm designed to detect specific patterns or events in the data. In some embodiments, annotating the data may correspond to one or more of the following: identifying regions of interest of neural activity signals, labeling different types of stimuli presented during a neural stimulation session, and / or categorizing the data according to specific experimental conditions or clinical outcomes.

[0064] In some embodiments, descriptive labels or categories may be applied to annotated segments of data based on predefined criteria or classification schemes. These labels provide semantic meaning to the data, facilitating subsequent analysis and interpretation. Labeling schemes may vary depending on the purpose of the study and the type of data being annotated. For example, in a neurostimulation study, labels may indicate the presence of abnormal neural activity patterns, the type of stimulus presented (e.g., visual, auditory), or the specific experimental conditions under which the data were collected.

[0065] In step 320, feature extraction may be performed to capture relevant characteristics or patterns from neural activity signals that are useful for predicting or analyzing treatment outcomes. Generally, feature extraction involves converting raw data into a format suitable for input to machine learning algorithms. Features extracted from neural activity data may include various aspects of the signal's characteristics, including temporal, spectral, spatial, and statistical characteristics. Temporal features describe how neural activity changes over time, such as the amplitude, frequency, and latency of event-related potentials (ERPs) or oscillatory rhythms. Spectral features characterize the frequency components of the neural signal, such as power spectral density, coherence, or phase locking measure across different frequency bands. Spatial features capture spatial patterns or distributions of neural activity across different brain regions or electrodes, which may reflect functional connectivity networks or local activation patterns. Statistical features quantify statistical properties of the neural signal, such as mean, variance, or distortion, providing information about the signal's distribution and dynamics.

[0066] Prior to feature extraction, preprocessing steps may be applied to raw neural activity data to enhance signal quality and remove artifacts. This may include filtering to remove noise or unwanted frequency components, artifact removal techniques, baseline correction, and normalization. Preprocessing ensures that extracted features accurately reflect the underlying neural dynamics and minimize the influence of confounding factors or measurement artifacts. Often, neural activity data can be high-dimensional and contain numerous features or channels. Dimensionality reduction techniques may be applied to reduce the complexity of the data and extract its essential characteristics. For example, principal component analysis (PCA), independent component analysis (ICA), or manifold learning methods can be used to identify lower-dimensional representations of the data that preserve the underlying structure and variability. In some embodiments, after feature extraction, feature selection techniques may be employed to identify the most useful subset of features for training a machine learning model. This reduces the complexity of the model, improves its generalization ability, and enhances interpretability. Feature selection methods include univariate statistical tests, wrapper methods (e.g., recursive feature removal), and embedding methods (e.g., L1 regularization), which evaluate the predictive power of individual features or subsets of features.

[0067] In step 325, the selected model may be trained to learn patterns and relationships from training data in order to make predictions or classifications. Under the context of non-invasive neural stimulation, training the model involves using neural activity data and corresponding outcomes to instruct the model to accurately predict therapeutic responses.

[0068] In some cases, depending on the nature of the data, one of several different types of model architectures may be selected. Commonly used models include linear models (e.g., linear regression, logistic regression), decision trees, random forests, support vector machines (SVMs), neural networks (e.g., deep learning models), Gaussian processes, and ensemble methods (e.g., boosting, bagging). In non-invasive neural stimulation research, domain-specific knowledge of neurophysiology, neural circuits, and stimulation protocols can influence the selection of models and functions.

[0069] Before training a selected model, the dataset is typically split into a training set, a validation set, and a test set. The training set is used to instruct the model, the validation set is used to tune hyperparameters and monitor performance during training, and the test set is used to evaluate the final performance of the trained model. Data preprocessing techniques such as normalization, scaling, imputation of missing values, and feature selection may be applied to prepare the dataset for training.

[0070] During training, the model may iteratively process batches of training data, updating its parameters to minimize the loss function. In each training iteration (or epoch), the model computes predictions for the input data, calculates the loss between the predicted output and ground truth, and may perform backpropagation to update the model parameters based on the gradient of the loss function. The training process may continue until a stopping criterion is met, such as reaching a maximum number of epochs or achieving satisfactory performance on a validation set. In some embodiments, during training, the model may iteratively process batches of training data, updating its parameters to minimize the loss function. In each training iteration (or epoch), the model computes predictions for the input data, calculates the loss between the predicted output and ground truth, and may perform backpropagation to update the model parameters based on the gradient of the loss function. The training process may continue until a stopping criterion is met, such as reaching a maximum number of epochs or achieving satisfactory performance on a validation set.

[0071] In step 330, the performance of the trained model can be evaluated. More specifically, cross-validation can be employed to estimate how well the model performs on undiscovered data. By systematically splitting the dataset into training and validation subsets multiple times, cross-validation provides a more robust estimate of the model's performance compared to a single training-test split.

[0072] In a non-restrictive exemplary embodiment of cross-validation, a dataset may be divided into k folds, each containing a portion of the data of approximately equal size. For example, in 5-fold cross-validation, the dataset is divided into five subsets. The model is trained k times, with each iteration using a different fold as the validation set and the remaining folds as the training set. For example, in the first iteration, the first fold is used for validation and the remaining folds are used for training. In the second iteration, the second fold is used for validation, etc. After training the model in each fold, its performance is evaluated against the validation set associated with that fold. Performance metrics such as accuracy, precision, recall, F1 score, or mean squared error are calculated for each iteration. The final performance metric is typically calculated as the mean (or median) of the performance metrics obtained over all k iterations. This provides a more reliable estimate of the model's generalizing ability compared to a single training-test split.

[0073] In step 335, the fully trained model can be deployed for clinical use. More specifically, after successful training and evaluation, the trained model can be deployed in a clinical setting to analyze brain activity, predict neural responses to stimuli, or, based on individual subject data, help diagnose medical conditions and design individualized treatment plans.

[0074] During deployment, electronic and / or magnetic test data may be received, including data associated with the activity of one or more neural circuits of the subject. This data may be received from an imaging device. In some embodiments, the subject's medical status may also be received. In some embodiments, a trained model may identify one or more neural circuits of the subject to target based on the subject's medical status and / or the electronic data associated with the activity of the subject's neural circuits. In some embodiments, the subject may be asked to perform motor tasks and / or cognitive tasks during data capture, which may help facilitate or target healthy and / or abnormal pathways. The specific motor and / or cognitive tasks that the subject is asked to perform may be dynamically identified by the trained model based on the subject's symptoms and / or medical condition. In some embodiments, input may be provided to stimulate one or more identified neural circuits of the subject (e.g., identified healthy neural circuits) using a non-invasive stimulation device. In one embodiment, input may be continuously provided by a non-invasive stimulation device for a predetermined period of time, based on subject feedback and / or feedback from electronic data associated with the activity of one or more determined neural circuits of the subject.

[0075] One technique uses dynamic maps to signal improvements to the algorithm of a machine learning model. In this technique, a first machine learning model is trained using previously recorded visual stimulus and neural data to map the dynamics between visual stimuli and neural dynamics. A second machine learning model searches in real time for visual stimuli that drive the dynamics of the neural data toward a specific target. The dynamic maps from the first machine learning model are used to make the search task of the second machine learning model more efficient.

[0076] Building and deploying trained models for treating subjects with scotoma. The data acquisition and machine learning techniques described above can be purposefully directed towards problems related to the presence of scotomas. A scotoma is an area in the visual field where partial or complete vision is lost. This can be caused by various factors, such as damage to the optic nerve or retina, or certain other neurological conditions. Some or all of the steps described herein can be carried out according to workflows 200 and 300. In some embodiments, data specific to scotomas can be used in model training and deployment.

[0077] In some embodiments, two sets of electronic data may be collected. The first set includes inputs from non-invasive visual stimulation devices provided to multiple subjects. These inputs may include a variety of visual stimuli such as color, gradient, shape, and pattern. The second set includes corresponding effects on one or more neurons, including cortical circuits and / or neurons in the subject's retina. This data includes neural responses induced by the visual stimuli provided in the first set. In some embodiments, the second training dataset may include only healthy neural circuits, only data including unhealthy / dysfunctional neural circuits, or a combination of the two. Training data for healthy neural circuits may be associated with circuits outside the central visual field unaffected by scotoma (e.g., healthy retinal neurons), cortical visual processing circuits, etc. Healthy neural circuits may be associated with circuits affected by scotoma, or may be located within a predetermined proximity range of such circuits. Training data for unhealthy / dysfunctional neural circuits may be associated with retinal neurons affected by scotoma or other visual impairments, or any neural circuits associated with them.

[0078] Both sets of training data can be annotated to provide supervised training for a machine learning system. This annotation helps the system learn the relationship between visual stimuli and neural responses. Specifically, the annotated training data is used to train the machine learning system and potentially employ a variety of machine learning algorithms (e.g., those described above). During clinical use, electronic data and / or neural circuit data associated with inputs provided by the stimulation device can be fed back into the machine learning system for retraining purposes. This makes it possible to fit the model to the responses of individual subjects and improve its effectiveness over time.

[0079] During the deployment of a model trained in a clinical setting, electronic and / or magnetic data associated with the activity of one or more neural circuits of the subject may be received from an imaging device. This data may include data associated with the activity of one or more neural circuits of the subject. In some embodiments, this data may be received from an imaging device. In some embodiments, the subject's medical condition may be received, which may include the condition, location of any blind spots, affected eye(s), severity, etc.

[0080] One or more neural circuits of a subject to be targeted may be determined based on the subject's medical condition and / or electronic data associated with the activity of the subject's neural circuits. In some embodiments, the neural circuits to be targeted may be healthy neural circuits associated with or within a predetermined proximity to abnormal neural circuits. In some embodiments, the subject may be asked to perform motor tasks and / or cognitive tasks during data capture, which may help facilitate or target healthy and / or abnormal pathways. The motor tasks and / or cognitive tasks performed may be determined based on the subject's symptoms and / or medical condition.

[0081] In some embodiments, input may be provided to a subject (e.g., using a non-invasive stimulation device) to visually stimulate one or more neural circuits of the subject. In some embodiments, specific or predetermined neural circuits may be targeted to stimulation based on the subject's stated medical condition. These neural circuits may be limited to healthy neural circuits, such as healthy retinal nerve cells.

[0082] A non-invasive stimulation device may continue to provide input until a predetermined period has passed or a predetermined event has been detected (e.g., specific feedback is received from the subject, specific feedback is received from electronic data associated with the activity of one or more predetermined neural circuits of the subject). In some embodiments, the subject may be instructed to perform a task while input from the non-invasive stimulation device is being shown. For example, the subject may be asked to repeatedly perform a motor or cognitive task, such as lifting an object or thinking about picking up an object. As another example, the subject may be asked to identify an object and / or gaze at the object. In some embodiments, a biofeedback loop may be used to optimize treatment, as described above with respect to workflow 200.

[0083] Construction and deployment of a trained model for the treatment of subjects experiencing phantom limb syndrome. The data acquisition and machine learning techniques described above may be intended to address problems related to phantom limb syndrome. Phantom limb syndrome is a condition in which a subject experiences sensation, whether painful or not, in a limb that is not present. Such a condition is commonly seen in amputees. Some or all of the steps described herein may be carried out according to workflows 200 and 300. In some embodiments, data specific to phantom limb syndrome may be used in model training and deployment.

[0084] In some embodiments, two sets of electronic or magnetic training data may be received. The first training set may include inputs provided to multiple subjects by a non-invasive visual stimulation device. These inputs may include providing subjects with colors, gradients, shapes, pictures, edges, object rotations, patterns, one or more flashes at predetermined time intervals and / or wavelengths. In some embodiments, the first set may also include data indicating the relevant medical condition of the subject, if present. Medical conditions may include limb / finger loss, and / or symptoms such as phantom limb pain, itching, tightness, heat, and / or cold. In some embodiments, subjects may be required to perform motor and / or cognitive tasks during data capture to help them target healthy or abnormal pathways based on their symptoms or medical condition.

[0085] In some embodiments, the second training set may include corresponding effects on nerves and / or neural circuits that may be related to cortical circuits, neural activity in the peripheral nervous system, and human limbs / fingers. The second training set may include data only of healthy neural circuits, data only of unhealthy / dysfunctional neural circuits, or a combination of both. Training data of healthy neural circuits may be associated with a given pathway of an abnormal neural circuit, or otherwise may be within that pathway. The pathway may consist of processing centers that are subcortical and / or cortical. Thus, stimulation of a normal neural pathway, for example, via the eye, may act to stimulate a healthy processing center in the brain that is adjacent to an abnormal processing center in the brain associated with phantom limb dysfunction.

[0086] Both sets of training data can be annotated to facilitate supervised training of the machine learning system. Electronic data associated with stimulus device inputs and neural circuit data can be returned to the machine learning system for retraining purposes.

[0087] During the deployment of a model trained in a clinical setting, electronic and / or magnetic data associated with the activity of one or more neural circuits in the subject may be received from an imaging device. This data may include data associated with the activity of one or more neural circuits in the subject. In some embodiments, the subject's medical condition may include limb / finger loss, and / or symptoms such as pain, itching, tightness, heat, and / or coldness in the phantom limb / finger.

[0088] Based on electronic data associated with the subject's medical condition and neural activity, one or more neural circuits to be targeted may be determined. These targeted circuits may include healthy circuits associated with or near abnormal circuits associated with phantom limb symptoms. In some embodiments, the targeted neural circuits may be healthy circuits associated with or within a predetermined proximity to abnormal neural circuits. In some embodiments, the subject may be asked to perform motor tasks and / or cognitive tasks during data capture, which may help facilitate or target healthy and / or abnormal pathways. The motor and / or cognitive tasks performed may be determined based on the subject's symptoms and / or medical condition.

[0089] In some embodiments, input may be provided to a subject (e.g., using a non-invasive stimulation device) to visually stimulate one or more neural circuits of the subject. In some embodiments, specific or predetermined neural circuits may be targeted to stimulation based on the subject's stated medical condition. This neural circuit may be limited to healthy neural circuits, e.g., healthy cortical circuits associated with afferent nerves of a missing limb, or adjacent or related circuits (e.g., visual stimuli that interrupt thalamic and ganglion-based signal processing and / or cortical interpretation of pain). Limb sensation, ocular sensation, or both may also be stimulated.

[0090] A non-invasive stimulation device may continue to provide input until a predetermined period has passed or a predetermined event has been detected (e.g., specific feedback is received from the subject, specific feedback is received from electronic data associated with the activity of one or more predetermined neural circuits of the subject). In some embodiments, the subject may be instructed to perform a task while input from the non-invasive stimulation device is being shown. For example, the subject may be asked to repeatedly perform a motor or cognitive task, such as lifting an object or thinking about picking up an object. As another example, the subject may be asked to identify an object and / or gaze at the object. In some embodiments, a biofeedback loop may be used to optimize treatment, as described above with respect to workflow 200.

[0091] The functional modules described herein are provided as examples. It will be understood that different functional modules can be combined to create different utilities. It will also be understood that additional functional modules or submodules can be created to implement a particular utility.

[0092] Any process described in this disclosure that is generally understood to be computer implementable may be performed by one or more processors in a computer system such as system environment 110, as described above. A process or process step performed by one or more processors may also be called an operation. One or more processors may be configured to perform such a process by accessing instructions (e.g., software or computer-readable code), and when executed by one or more processors, the instructions cause one or more processors to perform the process. Instructions may be stored in the memory of a computer server. The processors may be a central processing unit (CPU), a graphics processing unit (GPU), or any suitable type of processing unit.

[0093] A computer system such as system environment 110 may include one or more computing devices. If one or more processors of a computer system are implemented as multiple processors, these multiple processors may be contained within a single computing device or distributed among multiple computing devices. If the system environment has multiple computing devices, the memory of the computer system may include the memory of each of the multiple computing devices.

[0094] Figure 4 is a simplified functional block diagram of a computer system 400 that may be configured as a computing device for performing the processes described herein, according to an exemplary embodiment of the present disclosure. Figure 4 is a simplified functional block diagram of a computer that may be configured according to an exemplary embodiment of the present disclosure. In various embodiments, any of the systems herein may be an assembly of hardware including, for example, a data communication interface 420 for packet data communication. The platform may also include a central processing unit ("CPU") 402 in the form of one or more processors for executing program instructions. The platform may include an internal communication bus 408 and storage units 406 (e.g., ROM, HDD, SSD, etc.) that can store data in computer-readable media 422, but the system 400 may receive programming and data (e.g., voice, video, audio, images, or any other data via the electronic network 425) via network communication over an electronic network 425 that may correspond to a network 40. System 400 may also have memory 404 (such as RAM) for storing instructions 424 for performing the techniques presented herein, although the instructions 424 may be stored temporarily or permanently in other modules of System 400 (e.g., a processor 402 and / or a computer-readable medium 422). System 400 may also include input / output ports 412 and / or a display 410 for connecting to input / output devices such as a keyboard, mouse, touchscreen, monitor, and display. Various system functions may be implemented in a distributed manner on multiple similar platforms to distribute the processing load. Alternatively, a server may be implemented by appropriate programming on a single computer hardware platform.

[0095] In this disclosure, the term “based on” means “at least partially based on.” Unless otherwise specified by context, the singular forms “a,” “an,” and “the” include plural referents. The term “exemplary” is used to mean “example,” not “ideal.” The terms “comprises,” “comprising,” “includes,” and “including,” or other variations thereof, are intended to cover non-exclusive inclusion; therefore, a process, method, or product that includes a list of elements may not necessarily include only those elements, but may also include other elements not expressly enumerated, or other elements not specific to such process, method, article, or apparatus. Relative terms such as “about,” “approximately,” “substantially,” and “generally” are used to indicate a possible ±10% variation in the stated or understood values. Furthermore, the term “between” used when describing a range of values ​​is intended to include the minimum and maximum values ​​described herein. The use of the term “or” in the claims and specification is used to mean “and / or” only when the alternatives are mutually exclusive, unless it is explicitly indicated that it refers only to the alternatives; however, this disclosure supports the definition that refers only to the alternatives and “and / or.” As used herein, “another” may mean at least a second or more.

[0096] As used herein, the term “User” generally includes any individual or entity, such as a researcher and / or healthcare professional (e.g., a physician), who may desire information, a solution to a problem, or engage in any other type of interaction with the providers of the systems and methods described herein (e.g., through an application interface present on an electronic device). The terms “Electronic Application” or “Application” may be used interchangeably with other terms such as “Program,” and generally include software configured to interact with, modify, overwrite, supplement, or operate in conjunction with other software.

[0097] The programmatic aspects of technology can typically be considered as “products” or “manufactured goods” in the form of executable code and / or associated data carried on or embodied in a certain type of machine-readable medium. “Storage” type media can include any or all of the tangible memory of a computer, processor, or similar, or associated modules such as various semiconductor memories, tape drives, disk drives, etc., which can provide non-temporary storage for software programming at any time. All or part of the software may sometimes be communicated via the Internet or other various telecommunication networks. Such communication can make it possible, for example, to load software from one computer or processor to another, for example, from a management server or host computer of a mobile communication network to a server's computer platform, and / or from a server to a mobile device. Therefore, other types of media that can hold software elements include light waves, radio waves, and electromagnetic waves, which are used across physical interfaces between local devices via wired and optical terrestrial communication networks and through various air links. Physical elements that carry such waves, such as wired or wireless links, or optical links, etc., can also be considered media that hold software. As used herein, unless limited to non-temporary, tangible “storage” media, terms such as computer or machine “readable media” refer to any medium involved in giving instructions to a processor for execution.

[0098] Furthermore, some embodiments described herein include some features included in other embodiments and do not include other features, but combinations of features of different embodiments are within the scope of the invention and form different embodiments, as will be understood by those skilled in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0099] Therefore, although specific embodiments have been described, those skilled in the art will recognize that further modifications may be made without departing from the spirit of the invention, and it is intended that all such changes and modifications be claimed to fall within the scope of the invention. For example, functions may be added to or removed from the block diagram, and operations may be interchanged between function blocks. Steps may be added to or removed from the described methods within the scope of the invention.

[0100] The subject matter disclosed above should be considered illustrative and not restrictive, and the attached claims are intended to encompass all such modifications, enhancements, and other embodiments, the scope of which is intended to be included in the true spirit and scope of this disclosure. Accordingly, the scope of this disclosure should be determined by the broadest permissible interpretation of the following claims and their equivalents to the maximum extent permitted by law, and should not be limited or restricted by the foregoing detailed description. Although various embodiments of this disclosure are described, it will be apparent to those skilled in the art that many more embodiments are possible within the scope of this disclosure. Accordingly, this disclosure should not be limited except with respect to the attached claims and their equivalents.

Claims

1. A computer implementation method for modulating neural pathways to improve the functional abilities of a subject, A computing device receives a first dataset associated with the activity of one or more neural pathways of the subject, The computing device receives a second dataset which includes at least one of the following: at least one task performed by the subject, at least one symptom, and at least one medical condition of the subject. Using a processor associated with the computing device, determine one or more abnormal neural pathways based on the first dataset and the second dataset. Using the processor, determine from the second dataset one or more healthy neural pathways associated with the one or more abnormal neural pathways based on one or more criteria, Based on the healthy nerve pathway and the abnormal nerve pathway, a non-invasive stimulation device is used to determine the stimulation input to the healthy nerve pathway. The computer implementation method, including the above.

2. The computer implementation method according to claim 1, wherein the first dataset includes at least one signal associated with at least one of electrical activity, magnetic activity, and / or blood flow activity.

3. The computer implementation method according to claim 1, wherein determining one or more abnormal neural pathways includes comparing the first dataset with reference data associated with predetermined normal brain activity.

4. The computer implementation method according to claim 1, wherein the task includes an exercise task and / or a cognitive task.

5. The computer implementation method according to claim 1, wherein the selection of the type of task is determined by a trained machine learning model.

6. The computer implementation method according to claim 1, wherein the one or more criteria include the degree of proximity between one or more normal nerve pathways and one or more abnormal nerve pathways.

7. The computer implementation method according to claim 1, further comprising applying the stimulus input to the subject at a predetermined frequency.

8. The computer implementation method according to claim 7, wherein the application includes applying the stimulus input until a predetermined period of time has elapsed or until a predetermined event is detected.

9. To determine whether the aforementioned stimulus input improved the subject's neural activity and / or functional ability, If it is determined that the stimulus input did not exceed a predetermined threshold and improve the subject's neural activity and / or functional ability, then at least one additional type of stimulus input is applied to one or more healthy neural pathways. The computer implementation method according to claim 1, further comprising:

10. This further includes implementing a biofeedback loop, and such implementation is Receiving biofeedback data from the subject receiving the aforementioned stimulus input, Applying the aforementioned biofeedback data to a trained machine learning model, The trained machine learning model receives an output from the subject corresponding to a second stimulus input and / or a treatment process. The computer implementation method according to claim 1, including the method described in claim 1.

11. A system for modulating neural pathways to improve the functional abilities of a subject, One or more processors, The system comprises one or more computer-readable media storing instructions that can be executed by one or more processors to perform an operation, and the operation is: A computing device associated with the system receives a first dataset associated with the activity of one or more neural pathways of the subject, Based on the first dataset described above, one or more abnormal neural pathways are identified, The computing device receives a second dataset associated with the task performed by the subject, From the second dataset, determine one or more healthy neural pathways associated with one or more abnormal neural pathways based on one or more criteria, Based on the above determination, a non-invasive stimulation device associated with the system is used to apply a stimulation input to one or more healthy nerve pathways. The system including the above.

12. The system according to claim 11, wherein the first dataset includes at least one signal associated with at least one of electrical activity, magnetic activity, and / or blood flow activity.

13. The system according to claim 11, wherein determining one or more abnormal neural pathways includes comparing the first dataset with reference data associated with predetermined normal brain activity.

14. The system according to claim 11, wherein the selection of the type of task is determined by a trained machine learning model.

15. The system according to claim 11, wherein the one or more criteria include the degree of proximity between one or more normal nerve pathways and one or more abnormal nerve pathways.

16. The aforementioned operation further, The system according to claim 11, comprising applying the stimulus input to the subject at a predetermined frequency.

17. The system according to claim 16, wherein the application includes applying the stimulus input until a predetermined period of time has elapsed or until a predetermined event is detected.

18. The aforementioned operation further, To determine whether the aforementioned stimulus input improved the subject's neural activity and / or functional ability, If it is determined that the stimulus input did not exceed a predetermined threshold and improve the subject's neural activity and / or functional ability, then at least one additional type of stimulus input is applied to one or more healthy neural pathways. The system according to claim 11, further comprising:

19. This further includes implementing a biofeedback loop, and such implementation is Receiving biofeedback data from the subject receiving the aforementioned stimulus input, Applying the aforementioned biofeedback data to a trained machine learning model, The trained machine learning model receives an output from the subject corresponding to a second stimulus input and / or a treatment process. The system according to claim 11, including the following:

20. A non-temporary computer-readable medium for storing computer-executable instructions, wherein, when the computer-executable instructions are executed by the system, the system... In a computing device, receiving a first dataset associated with the activity of one or more neural pathways of a subject, Using a processor associated with the computing device, one or more abnormal neural pathways are identified based on the first dataset. The computing device receives a second dataset associated with the task performed by the subject, Using the processor, determine from the second dataset one or more healthy neural pathways associated with the one or more abnormal neural pathways based on one or more criteria, Based on the aforementioned determination, the non-transient computer-readable medium is used to perform an action that includes applying a stimulating input to one or more healthy nerve pathways using a non-invasive stimulating device.

Images