Systems and methods of generative neurostimulation and neuromodulation for treatment of disease

Abstract

Systems and methods are disclosed for modulating neural pathways to improve functional ability in a subject. The systems and methods may contain steps, including: providing for presentation, on a monitor associated with a computing device, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, at the computing device and using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; initiating, at the computing device, a priming stimulation process; and presenting, at the computing device, a visual perception learning (VPL) task to the patient after the priming stimulation process.

Description

SYSTEMS AND METHODS OF GENERATIVE NEUROSTIMULATION AND NEUROMODULATION FOR TREATMENT OF DISEASECROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application No. 63 / 596,507, filed November 6, 2023, which is incorporated by reference herein in its entirety.TECHNICAL FIELD

[0002] The present disclosure relates generally to neurostimulatory therapies for treating visual impairments and, more specifically, to systems and methods for optimizing neural network function in the visual cortex to improve visual performance through the use of generative video technology.BACKGROUND

[0003] Age-related macular degeneration (AMD) is a leading cause of blindness, currently affecting millions of people worldwide. While there have been significant advancements in the treatment of the wet form of AMD, such as anti- vascular endothelial growth factor (anti-VEGF) injections, treatment options for dry AMD are still limited. Dry AMD affects a larger proportion of AMD patients (e.g., 90- 90%) and is characterized by progressive neural dysfunction, which impairs visual function even in the absence of significant retinal damage. Conventional therapies focus on the damaged eye and its underlying pathology, leaving the disrupted neural networks in the visual cortex unaddressed. Therefore, a need exists for a therapeutic approach that can treat the neuronal disruption resulting from AMD and improve visual functions such as reading, contrast sensitivity, and peripheral vision.

[0004] The present disclosure is accordingly directed to techniques for treating AMD and other visual impairments through a neurostimulatory therapy that utilizes dynamic, generative video content. The background description provided herein is for the purpose of generally presenting context of the disclosure. Unless otherwise indicated herein, the materials described in this section are not prior art to the claims in this application and are not admitted to be prior art, or suggestions of the prior art, by inclusion in this section.SUMMARY OF THE DISCLOSURE

[0005] According to certain aspects of the disclosure, systems and methods are described for restoring functional ability in individuals afflicted with visual impairments.

[0006] In one aspect, a computer-implemented method for improving visual function in a patient is provided. The computer-implemented method may include: providing for presentation, on a monitor associated with a computing device, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, at the computing device and using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; initiating, at the computing device, a priming stimulation process to improve the visual function in the patient.

[0007] In another aspect, a system for improving visual function in a patient is provided. The system may include: one or more processors; and one or more computer readable media storing instructions that are executable by the one or more processors to perform operations comprising: providing for presentation, on a computing device associated with the system, one or more visual stimuli to thepatient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; initiating a priming stimulation process to improve the visual function in the patient.

[0008] In yet another aspect, a non-transitory computer-readable medium storing computer-executable instructions is provided. The non-transitory computer- readable medium stores computer-executable instructions which, when executed by a system, may cause the system to perform operations including: providing for presentation, on a monitor associated with a computing device, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, at the computing device and using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; initiating, at the computing device, a priming stimulation process to improve the visual function in the patient.

[0009] Additional objects and advantages of the disclosed embodiments will be set forth in part in the description that follows, and in part will be apparent from the description, or may be learned by practice of the disclosed embodiments. The objects and advantages of the disclosed embodiments will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims.

[0010] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosed embodiments, as claimed.BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate several embodiments and together with the description, serve to explain the principles of the disclosure.

[0012] FIG. 1A depicts an exemplary computer system for executing the methods described herein.

[0013] FIG. 1 B depicts an exemplary software platform for executing the methods described herein.

[0014] FIG. 2 depicts an exemplary workflow for utilizing the embodiments described herein to develop a platform that may improve visual function in a patient, according to one or more embodiments of the present disclosure.

[0015] FIG. 3 depicts a graph showing neurofunctional stratification based on simulated scotomas of varying sizes, according to one or more embodiments of the present disclosure.

[0016] FIG. 4 depicts a diagram of an exemplary visual perceptual learning (VPL) task, according to one or more embodiments of the present disclosure.

[0017] FIG. 5 depicts a chart that illustrates how various textures affect the brain’s neural response, according to one or more embodiments of the present disclosure.

[0018] FIG. 6 depicts an example computing system, according to one or more embodiments of the present disclosure.

[0019] FIG. 7A depicts a diagram of a simplified periodic contrast modulation, according to one or more embodiments of the present disclosure.

[0020] FIG. 7B depicts a graph that represents a typical example of a SNR- corrected EEG SSVEP response spectrum at the Oz electrode, according to one or more embodiments of the present disclosure.

[0021] FIG. 7C depicts a diagram that illustrates a projection of the multidimensional space onto the most significant dimensions, according to one or more embodiments of the present disclosure.

[0022] FIG. 8A depicts a diagram that shows the superiority of Al-generated stimulation over checkerboard flicker with different apertures in healthy controls, according to one or more embodiments of the present disclosure.

[0023] FIG. 8B depicts a diagram that shows the superiority of Al-generated stimulation over checkerboard flicker with different apertures in AMD patients, according to one or more embodiments of the present disclosure.

[0024] FIG. 9A depicts a graph showing the improved contrast sensitivity after multiple days of training with perceptual task and visual stimulation, according to one or more embodiments of the present disclosure.

[0025] FIG. 9B depicts a graph showing that the strength of the entrainments is strongly correlated with the contrast sensitivity as measured post session, according to one or more embodiments of the present disclosure.DETAILED DESCRIPTION OF EMBODIMENTS

[0026] The terminology used below may be interpreted in its broadest reasonable manner, even though it is being used in conjunction with a detailed description of certain specific examples of the present disclosure. Indeed, certain terms may even be emphasized below; however, any terminology intended to be interpreted in any restricted manner will be overtly and specifically defined as such in this Detailed Description section. Both the foregoing general description and thefollowing detailed description are exemplary and explanatory only and are not restrictive of the features, as claimed.

[0027] In this disclosure, the term “based on” means “based at least in part on.” The singular forms “a,” “an,” and “the” include plural referents unless the context dictates otherwise. The term “exemplary” is used in the sense of “example” rather than “ideal.” The terms “comprises,” “comprising,” “includes,” “including,” or other variations thereof, are intended to cover a non-exclusive inclusion such that a process, method, or product that comprises a list of elements does not necessarily include only those elements, but may include other elements not expressly listed or inherent to such a process, method, article, or apparatus. Relative terms, such as “about,” “approximately,” “substantially,” and “generally,” are used to indicate a possible variation of ±10% of a stated or understood value. In addition, the term “between” used in describing ranges 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 explicitly indicated to refer to alternatives only or the alternatives are mutually exclusive, although the disclosure supports a definition that refers to only alternatives and “and / or.” As used herein “another” may mean at least a second or more.

[0028] As used herein, the term “user” generally encompasses any person or entity, such as a researcher and / or a care provider (e.g., a doctor, etc.), that may desire information, resolution of an issue, or engage in any other type of interaction with a provider of the systems and methods described herein (e.g., via an application interface resident on their electronic device, etc.). The term “electronic application” or “application” may be used interchangeably with other terms like “program,” or thelike, and generally encompasses software that is configured to interact with, modify, override, supplement, or operate in conjunction with other software.

[0029] Age-related macular degeneration (AMD) is a leading cause of vision loss and blindness. The macula, a small area in the center of the retina responsible for sharp, central vision, deteriorates in individuals with AMD, leading to progressive vision loss. Currently, AMD affects hundreds of millions of people worldwide, a number that is expected to rise significantly in the coming decades due to several factors, including an aging population.

[0030] AMD is categorized into two forms: wet (exudative) and dry (atrophic). Wet AMD is caused by abnormal blood vessel growth that damages the macula. Current therapies for wet AMD, such as anti-vascular endothelial growth factor (anti- VEGF) injections, have made significant progress in slowing disease progression. However, these treatments are invasive, costly, and inaccessible for many, thereby leaving millions untreated or under-treated. Furthermore, they primarily target the underlying ocular pathology without fully addressing the functional loss experienced by patients. Additionally, wet AMD encompasses only approximately 10-15% of AMD cases, with the vast majority of AMD patients suffering from the dry form, for which treatment options are far more limited. Dry AMD involves the gradual degeneration of retinal cells, which leads to scotomas (i.e., blind spots) in central vision and is also associated with disrupted neuronal networks in the visual cortex, which further impairs visual function. Unlike wet AMD, there is no effective treatment that can stop or reverse the progression of this form of the disease. More particularly, conventional approaches focus on slowing the degeneration of retinal cells and limiting the growth of scotomas. However, these approaches fail to address the patient’s visualfunctions directly and critical aspects of daily life, such as reading, face recognition, and mobility, are still impaired even with available treatments.

[0031] The limitations of these conventional treatments highlight the need for a different approach, one that focuses on retraining the brain rather than just treating the eye. More particularly, much like how a camera with a cracked lens can either be repaired mechanically or have its software adjusted to compensate for the flaw, it is possible to “reprogram” the brain to adapt to and compensate for vision loss. One conventional therapeutic retraining strategy, learning with visual perceptual learning (VPL), involves repeated behavioral exercise, often focused on tasks like reading, which require intensive visual processing. By practicing these exercises over extended periods (e.g., hours per day for weeks or months), the brain can gradually reorganize its cortical connections, improving visual skills and quality of life. Although VPL enables the brain to adapt and compensate for areas of vision loss by enhancing peripheral vision and higher-order processing, it is demanding and timeintensive, requiring a great deal of effort from patients who may already be struggling with vision impairment. Additionally, results are slow, with improvements only seen after long periods of dedicated practice, making it difficult for many patients to sustain the effort required. To make VPL more efficient, another conventional retraining strategy combines VPL with transcranial electrical stimulation (tES), which uses alternating currents or magnetic stimulation to “prime” the brain. This priming is achieved by synchronizing the brain’s cortex with visual flickers or other stimulation, often measured by an increase in EEG alpha waves. By preparing the brain for learning, this method may accelerate the effects of VPL, making the training process shorter and potentially more effective. However, this approach also has limitations because it focuses on a single neural target (e.g., such as boosting alpha waves),which may therefore not address the broader range of neural dysfunctions in AMD. Furthermore, the need for specialized equipment, such as EEG sensors and electrical stimulation devices, make this approach costly and less accessible for most patients. Additionally, clinical settings and / or specialized personnel may also be required, which limits the possibility of employing this approach for widespread, at- home use.

[0032] To address the foregoing issues, the present disclosure contemplates a novel neurostimulatory approach for treating AMD by focusing on the brain’s neural networks rather than solely targeting the damaged retina. This approach leverages generative video technology to stimulate and reprogram the brain’s disrupted visual networks. The video stimuli may be tailored to specific neural and visual perceptual learning VPL objectives, and they may be continuously optimized in real time using feedback from biosensors, such as electroencephalogram (EEG) and eye movement tracking. This feedback allows the system to adjust the video frames pixel by pixel, ensuring that the neural stimuli are as effective as possible. Over time, this neuro- adaptive learning helps patients retrain their brains to improve visual processing, even as the physical damage to the retina persists. This generative video approach shifts the focus from treating the eye to training the brain and addresses the underlying functional loss in AMD by reorganizing cortical connections to enhance visual function. By engaging the brain’s natural plasticity, it helps patients utilize their remaining peripheral vision more effectively, compensating for central vision loss. In an aspect, the videos may be designed to “prime” the brain for learning, accelerating the process of visual perceptual learning, which traditionally requires extensive and time-consuming behavioral training.

[0033] The concept summarized above, and further elaborated upon herein, overcomes several issues faced by conventional techniques. For instance, it provides a non-invasive, accessible therapy that may be delivered at home via common devices like tablets, smart phones, and / or other mobile or personal devices. This reduces the cost and logistical barriers associated with current AMD treatments, such as frequent intraretinal injections. Additionally, by directly stimulating the brain’s neural networks, the novel concepts bridge the gap between retinal structure and functional vision improvement. It offers the possibility of significant, lasting improvements in visual function without the need for pharmaceutical intervention. Furthermore, the systems and processes described herein offer a solution for the majority of AMD patients who have dry AMD, a group for whom current treatments offer limited scope.

[0034] In an aspect, the collective concepts presented in this disclosure offer concrete and tangible applications in neuroscience and neuroengineering. These concepts represent improvements in computer technology by introducing innovative applications at the intersection of neuroscience and computing. For instance, the novel concepts utilize Al-driven video generation as a tool for neurostimulation, a novel application of generative models in healthcare. By continuously optimizing video stimuli in real time based on neural feedback (e.g., EEG and eye-tracking), the system effectively turns video content into a therapeutic medium, something not previously possible with standard video technologies. Additionally, the integration of closed-loop feedback systems in this processes also enhances the adaptive nature of the therapy, allowing it to adjust to individual patient’s neural responses. This requires the processing and analysis of large amounts of data in real time, pushing the boundaries of current computer technologies in terms of both computationalpower and algorithmic efficiency. The dynamic optimization of video content to achieve specific neural objectives represents a breakthrough in how Al and machine learning can be applied to personalized medical treatments.

[0035] The novel processes described herein also advance the technical field of neurostimulation and cognitive therapy by making these treatments more accessible and non-invasive. Existing neurostimulation techniques often rely on specialized, expensive equipment and are confined to clinical settings. The processes discussed here, by contrast, may be employed on everyday devices such as, for example, tablets, smartphones, virtual reality headsets, other types of mobile devices, and the like. This enables patients to access cutting-edge treatments at home with minimal equipment. Additionally, the system’s ability to optimize neurostimulatory video without the need for invasive sensors further enhances accessibility, making the technology scalable and practice for widespread use. Furthermore, this approach to neuro-adaptive learning addresses the limitations of conventional therapies by focusing on brain plasticity. By using Al-generated video to stimulate the brain, the system optimize the learning environment for visual perceptual tasks, improving both the speed and effectiveness of treatment. This represents a technical improvement over traditional vision therapy, which is typically time-consuming and resource-intensive, as it enables faster and more generalized improvements in visual function. The application of machine learning to continuously refine the therapy based on patient-specific neural and behavioral data creates an individualized treatment experience, enhancing the precision and efficacy of the therapy.

[0036] The concepts described in this neurostimulatory approach cannot practically be performed in the human mind due to the complexity, scale, and real-time nature of the computational tasks involved. For instance, the use of biofeedback, such as EEG signals and eye-tracking data, allows the system to continuously monitor neural activity and adjust the generative video stimuli accordingly. These adjustments are made in real time based on complex data sets that may include neural signals, video frame data, and patient-specific responses. The optimization of stimuli requires continuous recalibration of as many as thousands of parameters or more within milliseconds, a task that requires immense computational power and speed. Human cognition is not equipped to process and adjust this type of data instantaneously or at such a high level of precision.

[0037] The subject matter of the present disclosure will now be described more fully hereinafter with reference to the accompanying drawings, which form a part hereof, and which show, by way of illustration, specific exemplary embodiments. An embodiment or implementation described herein as “exemplary” is not to be construed as preferred or advantageous, for example, over other embodiments or implementations; rather, it is intended to reflect or indicate that the embodiment(s) is / are “example” embodiment(s). Subject matter may be embodied in a variety of different forms and, therefore, covered or claimed subject matter is intended to be construed as not being limited to any exemplary embodiments set forth herein; exemplary embodiments are provided merely to be illustrative. Likewise, a reasonably broad scope for claimed or covered subject matter is intended. Among other things, for example, subject matter may be embodied as methods, devices, components, or systems. Accordingly, embodiments may, for example, take the form of hardware, software, firmware, or any combination thereof. The following detailed description is, therefore, not intended to be taken in a limiting sense.

[0038] Throughout the specification and claims, terms may have nuanced meanings suggested or implied in context beyond an explicitly stated meaning. Likewise, the phrase “in one embodiment” or “in some embodiments,” or “in one aspect” or “in some aspects” as used herein does not necessarily refer to the same embodiment or aspect, and the phrase “in another embodiment” or “in another aspect” as used herein does not necessarily refer to a different embodiment or aspect. It is intended, for example, that claimed subject matter include combinations of exemplary embodiments in whole or in part.

[0039] FIG. 1 A depicts an exemplary system by which the methods described herein may be executed. Exemplary system 100 includes a data collection component 10, a database 20, and device data intelligence component 30, operably connected to each other via network 40. Alternatively, or additionally, one or more of the components may be connected with another component locally without reliance on network connection; e.g., through a wired connection.

[0040] As disclosed herein, data collection component 10 may include a device or machine with which electrical activity in the brain may be measured. In some embodiments, data collection component 10 may be an electroencephalograph (EEG) machine that contains, or is configured to support, one or more electrodes, amplifiers, filters, analog to digital converters, etc., by which to conduct an EEG test. Other devices such as a magnetoencephalography (MEG) machine may be used. In some aspects, data collection component 10 may be a database that receives EEG test data from one or more other sources. In other aspects, data collection component 10 may be any other brain recording device or modality that may convey information about neural activity. Consumer-grade virtualand augmented reality headsets that integrate data collection components into the headsets may be another form of device that can be utilized.

[0041] Data acquired by the data collection component 10 may be transferred to database 20 via network 40 or a direct, local or network connection. In some embodiments, the collected data may be analyzed by data intelligence component 30, via network 40 or a local or network connection. FIG. 1 B depicts exemplary functional modules that may be implemented to perform tasks of data intelligence component 30.

[0042] FIG. 1 B depicts an exemplary computer system 110 for using techniques discussed herein, for example for improving visual function, particularly for patients suffering from AMD. Exemplary system 110 may practice the techniques discussed herein by implementing, on one or more computer devices, user input and 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 needed for carrying out a particular task (e.g., an error correction or compensation module, a data compression module, etc.). These modules may correspond to the modules of FIG. 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 aspect of module 120 may correspond to data collection 10. As disclosed herein, user I / O module 120 may further include an input sub-module, such as a keyboard, MEG, EEG, eye tracking data, and an output sub-module, such as a display (e.g., a printer, a television, a smartphone, a monitor, a virtual reality (VR) device, and / or a touchpad). In some embodiments, all functionalities may be performed by one computer system. In some embodiments, the functionalities are performed by more than one computersystem. The various modules (e.g., for data processing, analysis, classification, communication, etc.) may be one or more processes executing in a distributed computing environment. For instance, in some embodiments, one or more components of the computer system 110 may be network accessible via cloud infrastructure. For example, the database 130 used to store data may be stored in 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 simple storage service “S3” buckets, etc.) from which data may be retrieved on demand. As another example, data processing, analysis, and classification may be performed in cloudbased environments using services like cloud-based data processing platforms, serverless computing, cloud-based machine learning platforms, and the like.

[0043] Also disclosed herein, a particular task may be performed by implementing one or more functional modules. In particular, each of the enumerated modules itself may, in turn, include multiple sub-modules implementing one or more techniques discussed herein. For example, data processing module 140 may include a sub-module for data quality evaluation (e.g., for performing iterative refinement and validation), a sub-module for normalizing any assigned weights to ensure that the weights contribute proportionally to the overall response, a sub-module for performing interpolation or extrapolation, and the like.

[0044] In some embodiments, a user may use I / O module 120 to manipulate data that is available either on a local device or can be obtained via a network connection from a remote service device or another user device. For example, I / O module 120 may allow a user, e.g., via a keyboard, a mouse, or a touchpad, to perform data analysis via a graphical user interface (GUI). In some embodiments, a user may manipulate data via voice control. In some embodiments, userauthentication may be required before a user is granted access to the data being requested. In some embodiments, user I / O module 120 may be used to manage various functional modules. For example, a user may request via user I / O module 120 input data while an existing data processing session is in process. A user may do so by selecting a menu option or type in a command discretely without interrupting the existing process. In another example, a user may utilize user I / O module 120 to set various thresholds, configure sample matching settings, and / or provide other instructions to computer system 110 that dictate how electrical signals in the brain are captured and / or monitored. As disclosed herein, a user may use any type of input to direct and control data processing and analysis via I / O module 120.

[0045] In some embodiments, system 110 further comprises a memory and / or database 130. In some embodiments, database 130 comprises a local database that may be accessed via user I / O module 120. In some embodiments, database 130 comprises a remote database that may be accessed by user I / O module 120 via network connection. In some embodiments, database 130 is a local database that stores data retrieved from another device (e.g., a user device or a server). In some embodiments, memory or database 130 may store data retrieved in real-time from internet searches. In some embodiments, database 130 may send data to and receive data from one or more of the other functional modules, including, but not limited to, a data collection module (not shown), data processing module 140, data analysis module 150, classification module 160, network communication module 170, and etc. In some embodiments, some or all real-sample data and / or synthetic sample data may be stored on database 130.

[0046] In some embodiments, database 130 may be a database local to the other functional modules. In some embodiments, database 130 may be a remotedatabase that may be accessed by the other functional modules via wired or wireless network connection (e.g., via network communication module 170). In some embodiments, database 130 may include a local portion and a remote portion.

[0047] In some embodiments, system 110 comprises a data processing module 140. Data processing module 140 may receive the real-time data, from I / O module 120 or database 130. In some embodiments, data processing module 140 may perform standard data processing algorithms, such as one or more of noise reduction, signal enhancement, normalization, interpolation and / or extrapolation, etc. In some embodiments, 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, data processing module 140 may additionally create a training data set, on which one or more machine-learning models (e.g., for classification, clustering, scoring, etc.) may be trained.

[0048] In some embodiments, system 110 comprises a data analysis module 150. In some embodiments, data analysis module 150 includes identifying brain activity patterns associated with particular medical conditions, as described in connection with data processing module 140.

[0049] In some embodiments, system 110 comprises a classification module 160, which may embody a “machine-learning model” or “trained classifier.” As used herein, a “machine-learning model” or “trained classifier” generally encompasses instructions, data, and / or a model configured to receive input, and apply one or more of a weight, bias, classification, and / or analysis on the input to generate 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 generallytrained using training data, e.g., experiential data and / or samples of input data, which are fed into the model in order to establish, tune, or modify one or more aspects of the model, e.g., the weights, biases, criteria for forming classifications or clusters, or the like. 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.

[0050] The execution of the machine-learning model(s) may include deployment of one or more machine-learning techniques, such as k-nearest neighbors, linear regression, logistic regression, random forest, gradient boosted machine (GBM), deep learning, a deep neural network (e.g., recurrent neural network (RNN), convolutional neural network (CNN), Transformers) and / or any other suitable machine-learning technique. Supervised, semi-supervised, and / or unsupervised training may be employed. For example, supervised learning may include providing training data and labels corresponding to the training data, e.g., as ground truth. Unsupervised approaches may include clustering, classification or the like. K-means clustering or K-Nearest Neighbors may also be used, which may be supervised or unsupervised. Combinations of K-Nearest Neighbors and an unsupervised cluster technique may also be used. Any suitable type of training may be used, e.g., stochastic, gradient boosted, random seeded, recursive, epoch or batch-based, etc.

[0051] Techniques discussed herein may also be implemented using multiple machine learning models, which may be executed in series and / or in parallel. For example, a first machine learning model may detect and / or interpret a neurological signal from the patient, and a second machine learning model maygenerate neurostimulatory imagery to present to the patient to achieve the desired result.

[0052] Neurostimulatory imagery may be generated using deep learning models. The deep learning models may have pre-trained weights or the weights may be learned from training on collected datasets which may combine visual stimulation, neural recordings, and behavioral recordings. The visual stimuli may be generated from a deep learning model that generates a group of video frames simultaneously from the stimulation parameters in a closed-loop fashion, and / or they may be generated frame-by-frame, conditioned on the changing neural data being recorded in real-time. Visual stimulation may also be generated from pre-specified visual features, e.g., gratings or white noise, or from combinations of pre-specified visual features and features generated from a deep learning model.

[0053] In an exemplary use case, a machine-learning model may be trained to analyze test data from a test subject whose specific neural activity with respect to a medical condition may be unknown and then subsequently identifying portions or characteristics of the test subject’s brain that may be responsible for or may be resultant of the medical condition. In some embodiments, the one or more parameters may include a score (e.g., a binomial probability score that may be calculated based on logistic regression analysis). As disclosed herein, the binomial probability score may correspond to the likelihood of a subject having a certain medical condition, the likelihood of a portion of the subject’s brain being active or inactive, the likelihood of a particular stimuli affecting a desired portion of the brain, etc. For example, a score of over a predefined threshold may indicate that a specific stimulus or sequence or set of stimuli has effectively stimulated a non-sensory region of the brain.

[0054] As disclosed herein, network communication module 170 may be used to facilitate communications between a user device, one or more databases, and any other suitable system or device through a wired or wireless network connection. Any communication protocol / device may be used, including, without limitation, a modem, an Ethernet connection, a network card (wireless or wired), an infrared communication device, a wireless communication device, and / or a chipset (such as a Bluetooth™ device, an 802.11 device, a WiFi device, a WiMax device, cellular communication facilities, etc.), a near-field communication (NFC), a Zigbee communication, a radio frequency (RF) or radio-frequency identification (RFID) communication, a PLC protocol, a 3G / 4G / 5G / LTE based communication, and / or the like. For example, a user device having a user interface platform for processing / analyzing tumor fraction data may communicate with another user device with the same platform, a regular user device without the same platform (e.g., a regular smartphone), a remote server, a physical device of a remote loT local network, a wearable device, a user device communicably connected to a remote server, and etc.

[0055] The functional modules described herein are provided by way of example. It will be understood that different functional modules may be combined to create different utilities. It will also be understood that additional functional modules or sub-modules may be created to implement a certain utility.

[0056] Referring now to FIG. 2, an exemplary workflow 200 is provided for utilizing the embodiments described herein to develop a platform that may improve visual function in a patient. Aspects of the exemplary workflow 200 may be performed in accordance with some or all components described in FIGS. 1A and 1 B.

[0057] At step 205, visual stimuli aimed at stimulating various parts of the user’s visual field may be generated. In an aspect, generative artificial intelligence (Al) (e.g., a form of Al that creates new content from learned patterns) may be utilized to produce dynamic and complex video stimuli tailored to each user’s neural and visual needs. For instance, these stimuli may consist of textures, shapes, and movements that are periodically or continuously optimized to target specific areas of the visual field, including foveal (central vision), parafoveal (near-central vision), and peripheral regions. By generating video content in real-time or substantially real-time, the system may produce stimuli that are highly customized for each patient’s unique condition, such as those with macular degeneration, where vision loss occurs primarily in the central part of the visual field. In an aspect, the Al-driven generator may utilize a small set of input parameters, such as contrast levels, phase dynamics, and spatial frequencies, to create a vast array of visual patterns that stimulate different neural circuits in the brain.

[0058] At step 210, a baseline of neural activity may be established in response to visual stimuli. More particularly, during initial sessions, patients may be exposed to various visual patterns or tasks designed to engage the visual cortex, such as lateral masking tasks or reading exercises. In one aspect, as the patients perform these tasks, a 64-channel EEG array (or a similar sensor configuration) records the electrical signals in the brain, particularly focusing on regions involved in visual processing. The system may capture neural data over short time windows (e.g., in the range of 500 milliseconds) immediately before or during the presentation of the visual stimuli. This data may provide a snapshot of how the brain reacts to different types of visual input.

[0059] In an aspect, once the data is collected, techniques such as dimensionality reduction and / or z-scoring may be applied to process the EEG signals. Dimensionality reduction may help to simplify the complex, high-dimensional neural data may focusing on the most relevant features, thereby making it easier to interpret and model. Z-scoring may normalize the data, allowing the system to compare different neural states on a consistent scale. The result of this analysis may be the development of a neural target / vector (e.g., a mathematical representation of the target neural state that encapsulates the optimal patterns of brain activity associated with the desired outcome, whether that is improved visual perception or another functional goal). This neural target may serve as the benchmark that the system strives to achieve during therapy sessions, as further described herein. In an aspect, the development of this neural target is what makes the neurostimulation therapy highly personalized and adaptive. Each user’s brain responds differently to visual stimuli, and the system may tailor the neural target to reflect the individual’s unique neural patterns. This personalized neural map allows for more precise stimulation, which is crucial for achieving meaningful improvements in visual performance, especially in conditions like AMD where vision loss is often complex and varied.

[0060] At step 215, patients may undergo a priming optimization process to prepare the brain for the upcoming VPL tasks by enhancing its readiness for learning and improving neural plasticity. By delivering targeted neurostimulation through priming, the brain’s visual cortex is “primed” to be more receptive and responsive during various task-based trials, as further described herein. To facilitate this process, in an aspect, patients may be placed in a controlled environment where they sit in a fixed position at a specified distance (e.g., 60 centimeters, etc.) from amonitor or screen and are exposed to priming stimulation. The priming stimuli may be generated based on specific stimulus parameters, such as contrast, spatial frequency, or motion, which may be selected to engage the brain’s visual processing areas effectively.

[0061] During this process, EEG data may be recorded continuously, e.g., using 64-channel EEG arrays or a smaller number of sensors depending on the setup. In an aspect, EEG sensors may be placed on the patient’s scalp to measure brain activity, particularly of the visual cortex. The system may monitor neural signals in real-time and may calculate a neural objective. In this context, the neural objective may refer to a short-term neural goal (e.g., a more immediate target that the system wants to achieve to enhance brain readiness before or during a VPL task). For example, the system may aim to increase the power of certain brainwave frequencies, such as alpha waves in the visual cortex, as this is often associated with a state of readiness for learning. Calculation of the neural objective may include metrics such as the power of certain brainwave frequencies, the phase difference between multiple EEG channels, and / or the patient’s behavioral performance (e.g., such as their accuracy and response time in completing the VPL tasks).

[0062] In an aspect, priming stimulation may last for a predetermined time period (e.g., 30 seconds, etc.) either before, during, or in some cases, both before and during the VPL tasks. In an aspect, the stimuli utilized for priming may be generated through procedural methods or by using advanced models such as a variational autoencoder (VAE), which is a type of neural network-based generative model. In the VAE, the stimulus parameters may be encoded as a vector in a latent space, which feeds into the VAE’s decoder to generate either single images or sequences of frames in a video. A dynamics module may be used to generatesequences in the latent space, e.g., using an autoregressive model, where the coefficients of the model form part of the stimulus parameters. To refine the stimuli, a stochastic optimizer may be employed. This optimization algorithm may dynamically adjust the stimulus parameters based on the feedback from the patient’s neural and behavioral responses. For example, if the EEG data indicates that the power in the alpha frequency band is not reaching the target level, or if the patient’s task performance is suboptimal, the optimizer may adjust the stimuli parameters to better engage the visual cortex and improve performance.

[0063] In some aspects, the priming optimization may be extended into laboratory testing, where participants undergo multiple days of VPL training in combination with priming stimulation. The priming videos may be optimized for specific VPL tasks, which may include exercises like lateral masking, crowded letter recognition, single letter recognition, etc. In these tasks, the visual stimuli may be presented in the peripheral vision to target the areas most affected by vision loss in conditions like AMD. After several days of training, the system may collect data from groups of participants, which may be used to identify the most effective stimulus configurations for different populations. These configurations may then be fine-tuned based on participant characteristics (e.g., visual acuity, task performance), which may be gathered through surveys or vision tests, and deployed for future therapy sessions.

[0064] At step 220, once the neural target is developed (e.g., based on baseline neural data gathered from earlier visual stimulation sessions) and the patients have been exposed to priming stimulation, the system may use this target to evaluate the patient’s brain activity during ongoing therapy. Specifically, the generated visual stimuli may be dynamically adjusted in an optimization process inresponse to real-time feedback from the user’s neural activity. Accordingly, in an aspect, the generative Al process may not be static, but rather, may be part of a closed-loop system that responds to real-time feedback from the user’s brain. In this regard, the initially generated visual stimuli may be presented to the patient and as the patient views the stimuli, one or more EEG sensors record the brain’s electrical activity, particularly in regions responsible for visual processing. The system may then evaluate how closely the recorded neural responses align with the desired or target neural state, which may be associated with improved visual perceptual learning, enhanced visual acuity, or other therapeutic goals. In an aspect, this evaluation may be facilitated by analyzing specific neural metrics, such as the power of certain brainwave frequencies (e.g., alpha waves) or the phase coherence between different brain regions.

[0065] If the neural feedback indicates that the stimuli are not eliciting the desired response, e.g., if the brain activity is below the target level, the system may dynamically adapt the stimuli in real-time. This process may involve altering parameters such as contrast, brightness, spatial frequency, or movement patterns in the generated video to better engage the brain’s visual cortex. In an aspect, the optimization process may leverage advanced algorithms, including Bayesian Optimization or Genetic Algorithms, to efficiently explore different stimulus configurations and determine which are most effective at driving the desired neural outcomes. Accordingly, the closed-loop optimization process ensures that the visual stimuli are continually refined based on each user’s unique and evolving neural responses, thereby creating a personalized and adaptive therapy experience. This process enhances the efficiency and effectiveness of the treatment by ensuring thatthe brain is consistently stimulated in the most beneficial way, maximizing the potential for neuroplastic changes and functional improvements.

[0066] In an aspect, task-based trials may be implemented in which specific visual tasks are presented to patients in order to collect real-time neural feedback and optimize the stimuli accordingly. These trials may involve showing patients various visual stimuli, such as patterns, textures, or shapes, while they perform tasks designed to engage different aspects of visual perception. For example, in an exemplary task, patients may be asked to detect or differentiate between visual elements, such as identifying which interval contains a specific visual stimulus, like a Gabor patch flanked by other distractor elements. The contrast of the visual stimulus may be dynamically adjusted through a staircase method to ensure the task remains challenging and tests the patient’s visual processing capability. During these trials, the patient’s brain activity may be monitored using an EEG, which records neural responses from the visual cortex as they complete each task. This neural data may be leveraged in the closed-loop optimization process described above. In an aspect, the task-based trials may be structured with a large number of trials (e.g., 120 trials, etc.) to ensure that enough data is collected to make meaningful adjustments to the stimuli. The stimuli may be presented in short intervals (e.g., 33 milliseconds, etc.), with breaks between them to reset the patient’s visual processing. The results of each task trial may provide a feedback loop that guides the optimization algorithms to refine the stimuli parameters to help the brain achieve the desired neural target.

[0067] As an example of the foregoing, two trials may be instituted that each consist of two 33-millisecods intervals, one containing a Gabor patch (the target) with two Gabor flankers and the second showing just the flankers interrupted by a blank screen (e.g., 500 milliseconds). Patients may be asked to judge which intervalcontains the target. The target’s contrast may be varied according to a 3: 1 staircase(-0.1 log unit for 3 consecutive correct responses, -0.1 log for 1 wrong response).

[0068] In some aspects, the visual stimuli may be tested and refined across different participant groups to calibrate the system and ensure it is effectively driving the brain toward the desired neural target / state. This process may involve initially testing the visual stimuli on a first group of healthy participants, who serve as a baseline for understanding how the brain should respond to the visual tasks when there is no underlying visual impairment. During this phase, the system collects EEG data from participants while they engage in the task-based trials, as described above. The neural activity is recorded and analyzed to identify patterns of brain activity that correspond to the successful performance on the visual tasks. These patterns may form the neural target, which represents the optimal brain state the systems aims to achieve in all participants.

[0069] In an aspect, once the neural target is established with healthy participants, the system may apply neural optimization to another group of participants. This group may include either additional healthy individuals or those with visual impairments, such as mild to moderate AMD. The goal of this phase is to apply the previously established neural target and adjust the visual stimuli in realtime to optimize each participant’s brain activity, driving it closer to the target state. The system monitors EEG data from these participants during the trials, comparing their neural responses to the ideal target and making dynamic adjustments to the stimuli, such as altering the contrast, spatial frequency, or timing, to improve alignment with the target neural state.

[0070] At step 225, the performance of the optimization process may be evaluated and the effectiveness of the stimuli in driving the brain toward a desiredneural state may be determined. More particularly, in this step, the system analyzes the neural and behavioral data collected during the task-based trials and experimental group phases to assess how well the stimuli are helping participants achieve the predefined neural target. To conduct this analysis, the system may process the EEG data recorded during the trials, comparing the participants’ neural responses to the target state. Statistical methods (e.g., Euclidean distance calculations, etc.) may be utilized to quantify how close the brain activity is to the target. This distance provides a metric of how effective the visual stimuli were at engaging the visual cortex and moving the brain toward the desired neural outcome. In an aspect, behavioral data, such as task performance and accuracy, may also be analyzed to correlate improvements in visual function with the corresponding neural changes.

[0071] Once the data is collected, the system may evaluate the performance of the different optimization algorithms that were used to adjust the stimuli in real time. The goal is to rank these algorithms based on their ability to minimize the distance between the current and target neural states within a set timeframe, often no more than an hour of trial time. The ranking process helps determine which algorithms are most effective at dynamically adapting the visual stimuli to each participant’s unique neural responses. In aspect, factors such as speed, accuracy, and the ability handle noisy neural data may be considered when ranking the algorithms. By identifying the most effective optimization techniques, the system may prioritize these approaches in subsequent trials or therapy sessions, leading to more efficient and targeted neurostimulation.

[0072] In an aspect, a neurofunctional data registry may be created that captures and maps the neural responses of participants during visual tasks, whichmay be used for understanding and treating conditions like AMD. The registry may be built by collecting neurophysiologic data (e.g., such as EEG or magnetoencephalography, MEG) while participants, particularly healthy subjects, watch carefully curated visual stimuli designed to simulate visual impairments like AMD.

[0073] In an aspect, the creation of this data registry may begin by showing healthy participants videos that simulate the effects of scotomas, which are blind spots in the visual field commonly experienced by individuals with AMD. The scotomas may be simulated on a monitor, varying in size (e.g., 0, 5, or 10 degrees of the visual field) using eye-tracking technology to ensure accuracy in where these simulated blind spots appear. By manipulating the size and location of these scotomas, the system may model how visual processing changes as the visual field is reduced, thereby helping researchers understand how the brain compensates for vision loss. The neurofunctional biomarkers (e.g., neural signatures that correlate with different scotoma sizes and visual processing changes) may be extracted from the EEG or MEG data. These biomarkers helps develop a data map that links disease characteristics with functional brain responses and help illustrate how both the disease and brain function can be described through neurofunctional data.

[0074] Referring now to FIG. 3, graph 300 depicts neurofunctional stratification based on simulated scotomas of varying sizes, using a scatter plot of principal components (PC1 plotted on the x-axis against PC2 plotted on the y-axis) to differentiate between groups of subjects exposed to different levels of visual impairment. The stratification may be based on neurophysiologic data recorded while participants viewed visual stimuli with simulated scotomas of different sizes (e.g.,0.0, 5.0, and 10.0 degrees in diameter). Graph 300 shows three clusters, 305, 310,and 315. Data points encompassed by cluster 305 represent participants with no simulated scotoma, meaning they had normal visual processing during the experiment. Data points encompassed by cluster 310 represent subjects experiencing a mid-range scotoma size in the simulated field, mimicking a partial visual impairment. Data points encompassed by cluster 315 correspond to participants experiencing the largest simulated scotoma, representing a severe visual impairment. Cluster 305 is tightly packed, indicating that participants with normal vision process the visual stimuli in a fundamentally different way compared to those with visual field impairments. Clusters 310 and 315 show more spread and overlap, reflecting the more complex neural adaptations required for visual processing with larger blind spots. Graph 300 supports the idea that these biomarkers derived while freely viewing naturalistic videos can differentiate stages of progressive visual impairment as seen in AMD.

[0075] In an aspect, Narrow Alpha Band Entrainment (NABE) may additionally be utilized to enhance participants’ performance in peripheral visual tasks. NABE leverages specific frequencies of brainwave stimulation to synchronize brain activity in a way that improves cognitive and visual performance. In this case, participants may be exposed to complex visual textures that are designed to entrain alpha brainwaves, thereby enhancing their ability to perform VPL tasks. In an aspect, NABE may be further optimized with priming videos that are designed to boost the Steady State Visual Evoked Potentials (SSVEP), which is a type of brain response that reflects the brain’s reaction to repetitive visual stimulation.

[0076] One exemplary VPL task may be a crowded trigram task, in which patients may be tasked with identifying a target letter in a crowded visual field where the letters are presented at a fixed distance in the peripheral vision. The size of thefront remains constant, but the crowding gap (i.e., the space between the letters) varies, making the task more or less difficult depending on the trial. For example, diagram 400 in FIG. 4 illustrates exemplary crowded trigrams placed in the participant’s peripheral vision, with a specific focus on the critical spacing, for example, predetermined relative orientation, between the letters. For example, the crowded trigram may be presented at 8 degree of visual periphery. The task may require participants to correctly identify the middle letter of the trigram, which helps to assess how the brain processes visual information in crowded peripheral vision - a common issue for patients with visual impairments like AMD. The number of symbols and their relative orientation may vary. In an aspect, these random letter sets are non-learnable in the sense that participants cannot rely on previous knowledge or patterns to improve performance. In an aspect, each trigram is different, and the font size is constant, which standardizes the visual load and isolates the effects of spacing and peripheral vision challenges on the task performance.

[0077] Participants in the experimental group may be exposed to the SSVEP-optimizing priming video for a predetermined period (e g., 20 seconds) before each block of trials. This priming video may be designed to enhance the brain’s readiness for the visual task by entraining neural activity in a way that maximizes visual processing. Data from the experiment may be analyzed in two ways: via a texture search task or VPL task. In the former, the SSVEP response, measured as the signal-to-noise ratio (SNR) for each visual texture, is z-scored within participants and then averaged to find which textures elicit the strongest neural responses. This helps determine which types of visual stimuli are most effective at engaging the brain. In the VPL task, the accuracy of target identification in the crowded trigram task may be analyzed using t-tests to compare improvementsin performance between trial blocks. Statistically significant improvements in accuracy are monitored for after exposure to the priming videos.

[0078] Experimental results show that certain visual textures, when paired with contrast flicker, produce significantly stronger SSVEP responses (e.g., measured as 1.8 z-scores higher than the media texture), indicating that specific stimuli are more effective at driving neural activity. In the VPL task, the participants in the experimental group (e.g., those exposed to the priming video) showed an average accuracy increase of 5% or more in identifying the target letter in the crowded visual field. This improvement is not seen in the comparison group, which was not exposed to the priming video, suggesting that the SSVEP-maximizing priming video plays an important role in enhancing visual performance.

[0079] Referring now to FIG. 5, chart 500 is provided of the SSVEP SNR, z- scored across participants for different textures in a texture search task. Chart 500 illustrates how various textures affect the brain’s neural response, as measured by the SSVEP SNR on the y-axis. Examination of chart 500 reveals a clear upward trend, with certain textures producing significantly stronger SSVEP responses (as indicated by z-scored SNR values exceeding 1.5). These textures are likely to be effective in driving neural plasticity and improving visual function when used in neurostimulation or VPL training. In an aspect, the results align with the disclosure’s claim that SSVEP-maximizing stimulation can improve visual performance, particularly in tasks requiring peripheral vision processing or managing crowded visual scenes.

[0080] In an aspect, the practical implementation of the system described herein for users may involve several key components to ensure it is accessible, effective, and adaptable to a wide range of visual impairments, particularly forindividuals with AMD. In an aspect, an initial survey or vision test may be administered to each user, which would classify them into a specific group based on their visual abilities and needs. This classification process may help tailor the therapy by selecting the most appropriate priming stimulation, task durations, and VPL parameters for that group, based on prior lab-tested configurations. The system may be designed to be flexible, meaning that while advanced features like EEG monitoring may be included fortracking neural responses and adjusting tasks dynamically, these features wouldn’t always be required. This flexibility makes the system accessible to users without sophisticated equipment, ensuring broader adoption.

[0081] In an aspect, the system may be delivered on common consumer devices like tablets, smartphones, or virtual reality headsets, making it feasible for home use. In one exemplary implementation, users may download an application where the visual stimulation and VPL tasks are delivered through either pregenerated stimulation videos (e.g., stored locally on the device, etc.) or onlinegenerated videos, which may be streamed from the cloud if the device has limited computation capacity. This cloud-based approach may minimize the memory and processing requirements for the user’s device, allowing for real-time video generation and streaming without burdening the hardware. The app may guide the user through each session, including visual tasks and priming exercises, ensuring that the therapy is both user-friendly and effective.

[0082] In an aspect, during the sessions, users may perform tasks designed to stimulate and train their visual system, as described herein (e.g., lateral masking, fixation exercises, and reading tasks). In an aspect, eye tracking technology may be employed to monitor the user’s preferred retinal locus (PRL), ensuring that thestimuli are targeting the appropriate areas of the visual field to maximize the training’s effectiveness. The tasks may be adjusted based on the user’s performance real-time feedback from the system, especially if EEG is used to track brain activity, enabling a closed-loop optimization where the system continuously refines the stimuli based on neural responses.

[0083] For users with more severe visual impairments, or those looking to accelerate their progress, the system may incorporate generative Al to create dynamic video stimuli. Although the generative Al may be high-dimensional, these stimuli may be driven by low-dimensional control parameters, which may adjust the spatiotemporal dynamics of the video to specifically target the user’s neural deficits, such as stimulating healthy neurons in the peripheral vision to compensate for central vision loss. For example, a particular video may be characterized by a highdimensional vector (e.g., tens of thousands of dimensions or more), but a control signal controls the movement of this vector over time, which may have fewer degrees of freedom (e.g., 10 - 100). In an aspect, the system’s machine learning algorithms may help tailor these videos to the user’s evolving needs, adjusting the visual tasks and priming sessions as they progress through the therapy.

[0084] In terms of usability, the system may incorporate features like voice instructions, high-contrast interfaces, and automatic triggers that start a session when the user engages with the device, ensuring that even individuals with severe vision loss may navigate and use the system independently. Reminders and prompts may be built into the application to encourage consistent use, enhancing the likelihood of long-term improvements.

[0085] Studies have been conducted to demonstrate how neuro-adaptive priming videos may accelerate VPL and improve other visual functions in participantswith mild to moderate dry AMD. These studies show that by optimizing video-based stimulation, VPL may be enhanced in terms of both speed and effectiveness, potentially lead to improvements in visual tasks such as contrast sensitivity, visual acuity, and reading ability. The rationale for this study is based on existing evidence that VPL, when combined with oculomotor training or electrical stimulation, may help improve visual functions in patients with AMD and other visual impairments. Accordingly, similar benefits may be achieved using optimized visual stimulation (e.g., neuro-adaptive priming videos), which are designed to engage specific neural pathways and facilitate learning. This stimulation helps prepare the brain to adapt more efficiently during visual tasks, potentially accelerating the process of VPL and enabling broader improvements in visual function.

[0086] In an aspect, the study may be designed to involve two groups of participants with dry AMD: one experimental group and one control group. The experimental group may receive optimized neuro-adaptive priming videos, while the control group receives a random (sham) video, which serves as a placebo to compare the effect of the optimized video. In an aspect, the participants may undergo VPL training across three distinct tasks: lateral masking, fixation training, and reading. Lateral masking may involve detecting visual stimuli surrounded by nearby distracting stimuli (flankers) and is used to improve contrast sensitivity. Fixation training may involve participants focusing on a flickering dot pattern, which becomes a uniform gray when they maintain stable fixation. This exercise helps improve eye movement control. Reading tasks may involve reading words in the smallest print determined by the Freiburg Acuity and Contrast Test (FrACT), with two blocks of 100 words presented to test reading speed and accuracy.

[0087] To objectively assess the effectiveness of the priming and VPL training, four key metrics may be collected at the beginning of the first session: contrast threshold (e.g., measured through lateral masking at a specified spatial frequency (e.g., 6 cycle per degree), which tests the participants’ ability to discern differences in contrast), visual acuity (e.g., assessed using the Landolt C and Sloan letter tests from the FrACT, which measure how well participants can see fine details), crowding (e.g., measured using Sloan letters with flankers (i.e., nearby distracting stimuli) on each side, which tests the participants’ ability to recognize letters in a crowded visual field, a task often impaired in AMD, reading speed (e.g., assessed using a computerized version of the Minnesota low vision reading (MNREAD) app, which measures how quickly and accurately participants can read common English words presented in small print.

[0088] In an aspect, to facilitate the study, the participants may take part in three separate training sessions on consecutive days, where they undergo the three VPL tasks (e.g., lateral masking, fixation training, and reading). In each session, eye tracking may be used to monitor each participant’s preferred retinal locus (PRL), the area they use for vision compensating for the central vision loss caused by AMD. Additionally, the stimulus presentation in each session may be triggered once the participant’s gaze aligns with the PRL, ensuring that the training targets the appropriate part of their visual field. Furthermore, each session may be capped at a predetermined time limit, for example 80 minutes, with variations depending on how quickly the participant aligns their gaze and completes the tasks.

[0089] In an aspect, after the third training session, participants may return after a predetermined time period, for example at least one week, for a post-test assessment to gauge the effectiveness of the training. EEG may be recorded duringthis assessment to capture any cortical changes associated with visual improvement.The same four metrics (e.g., contrast threshold, crowding, visual acuity, and reading speed) may be reassessed, and the data from the pre- and post-test assessment are compared. In an aspect, this data analysis may involve calculating the change between pre-test and post-test values for each metric. The results may then be compared between the experimental and control groups using a mixed-model Analysis of Variance (ANOVA), with a significance level of alpha < 0.05. This analysis may enable researchers to determine whether the neuro-adaptive priming videos significantly improved visual functions compared to the sham videos.

[0090] In an aspect, if the optimized stimulation doesn’t sufficiently accelerate learning or generalize improvements, alternative methods may be employed. For instance, Steady-State Visual Evoked Potential (SSVEP), which is a periodic visual stimulation technique, may be used as an alternative priming method based on prior research showing that it can enhance VPL. In an aspect, if the gazecontingent presentation of the stimuli prolongs the duration of the trials, the reading task may be removed to keep the sessions with the time limits.

[0091] In an aspect, the techniques described herein may go beyond improving visual function alone and may extend into enhancing facial recognition and visual memory - areas that are often affected by neurodegenerative diseases. Memory loss, particularly for faces, may lead to a loss of ability to recognize familiar individuals, a condition that may be distressing both for patients and their caregivers. The methods described here offer a solution by leveraging advanced machine learning models, particularly generative models, to retrain the brain in recognizing faces. In an aspect, a machine learning model may be trained on a dataset of facial images to generate three-dimension representations, either in still image or videoform, creating lifelike avatars. In some aspects, these avatars may be designed to speak using algorithm-generated or user-provided speech, adding a dynamic and personalized dimension to the therapy.

[0092] In this method, the patient may be exposed to the generated faces avatars while receiving neurostimulatory images or codes. These images or codes may be strategically designed to stimulate areas of the brain associated with facial recognition and memory, such as the visual cortex or other memory-related regions. By stimulating these areas while the patient interacts with the face avatars, the system effectively retrains the brain to retain or improve its ability to recognize faces. For patients with neurodegenerative conditions that affect memory, this technique could potentially slow or reverse the decline in facial recognition ability. In an aspect, users may also upload images of certain “important” individuals (e.g., family members or caregivers), and the system could generate three-dimensional interactive avatars based on these photos. This may allow patients to practice recognizing specific faces in combination with the neurostimulatory techniques, reinforcing their neural networks related to facial recognition.

[0093] Additionally or alternatively to the foregoing, the techniques described herein may be expanded to improve other sensory networks, such as the sensorimotor network (SMN). By targeting the SMN with priming neurostimulatory images, it is possible to activate and reorganize neural pathways, which may lead to better sensory integration and motor control. This may help mitigate issues like balance problems, often caused by desensitization to visual or sensory cues, which in turn could help prevent falls - an important benefit for older individuals or those with neurodegenerative conditions.

[0094] In an aspect, these techniques may also have potential for performance enhancement. More particularly, by applying neurostimulatory codes to specific brain regions, such as the SMN or visual cortex, cognitive and sensory abilities related to reaction time, peripheral vision, or night vision may be improved. This may have applications for professionals requiring high levels of visual or sensory performance, such as video gamers, pilots, military personnel, and others. By enhancing brain functions related to these tasks, individuals may experience sharper response times, better accuracy, and improved visual processing.

[0095] Moreover, the techniques discussed in this disclosure may have applications in stem cell therapy. More particularly, when new neurons are generated through stem cell therapy, these neurostimulatory techniques may help guide and integrate the new cells into the brain’s existing sensory networks. In an aspect, machine learning algorithms, trained to present specific neurostimulatory images, may condition these newly generated stem cells to take on specific sensory or cognitive functions, such as enhancing vision, hearing, or touch. For example, visual neurostimulatory images may be used to guide stem cells in the auditory cortex, helping them participate in hearing function, or in the sensory cortex, improving tactile perception.

[0096] The functional modules described herein are provided by way of example. It will be understood that different functional modules may be combined to create different utilities. It will also be understood that additional functional modules or sub-modules may be created to implement a certain utility.

[0097] In general, any process discussed in this disclosure that is understood to be computer-implementable may be performed by one or more processors of a computer system, such as system environment 110, as describedabove. A process or process step performed by one or more processors may also be referred to as an operation. The one or more processors may be configured to perform such processes by having access to instructions (e.g., software or computer-readable code) that, when executed by the one or more processors, cause the one or more processors to perform the processes. The instructions may be stored in a memory of the computer server. A processor may be a central processing unit (CPU), a graphics processing unit (GPU), or any suitable types of processing unit.

[0098] A computer system, such as system environment 110, may include one or more computing devices. If the one or more processors of the computer system are implemented as a plurality of processors, the plurality of processors may be included in a single computing device or distributed among a plurality of computing devices. If a system environment comprises a plurality of computing devices, the memory of the computer system may include the respective memory of each computing device of the plurality of computing devices.

[0099] FIG. 6 is a simplified functional block diagram of a computer system 600 that may be configured as a computing device for executing the processes described herein, according to exemplary embodiments of the present disclosure. FIG. 6 is a simplified functional block diagram of a computer that may be configured according to exemplary embodiments 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 620 for packet data communication. The platform also may include a central processing unit (“CPU”) 602, in the form of one or more processors, for executing program instructions. The platform may include an internal communication bus 608, and a storage unit 606 (such as ROM,HDD, SDD, etc.) that may store data on a computer readable medium 622, although the system 600 may receive programming and data via network communications via electronic network 625, which may correspond to network 625 (e.g., voice, video, audio, images, or any other data over the electronic network 625). The system 600 may also have a memory 604 (such as RAM) storing instructions 624 for executing techniques presented herein, although the instructions 624 may be stored temporarily or permanently within other modules of system 600 (e.g., processor 602 and / or computer readable medium 622). The system 600 also may include input and output ports 612 and / or a display 610 to connect with input and output devices such as keyboards, mice, touchscreens, monitors, displays, etc. The various system functions may be implemented in a distributed fashion on a number of similar platforms, to distribute the processing load. Alternatively, the systems may be implemented by appropriate programming of one computer hardware platform.Additional Information

[0100] For stimulus preparation, complex neuromodulatory stimuli may be generated by an Al-powered platform, which may be configured to further optimize them to boost narrow-band alpha EEG response. For example, 14 healthy participants were stimulated with visual periodic entrainment stimuli (see Fig 1 a) for 20 or 2 seconds in closed-loop, using neural feedback to iterate and optimize the video stimulation. The stimulus evoking the strongest EEG SSVEP was selected for use in the optimized video. FIG. 7 A presents a diagram 705 of a simplified periodic contrast modulation. FIG. 7B presents a graph 710 that represents an example of SNR-corrected EEG SSVEP response spectrum at the Oz electrode. The vertical line at 10 Hz may correspond to the stimulation frequency and the 1stharmonicresponse. FIG. 70 presents a diagram 715 that illustrates a projection of the multidimensional space onto the most significant dimensions.

[0101] As part of an experimental design, the effect of optimized versus control stimulation was prospectively tested in 7 patients (6 female, 1 male; mean age = 76) with advanced dry AMD and 8 healthy participants (6 female, 2 male; mean age = 60). Both groups were presented with (a) the generative video optimized for SSVEP response, and (b) unoptimized periodic stimuli (checkerboard flicker) as a control. The stimuli were presented in foveal (6 degrees of visual angle), wholescreen and extrafoveal (whole-screen minus foveal) stimulation modes. Stimulation was delivered at 10 Hz corresponding to the alpha frequency band. Participants from both groups were instructed to fixate at the center of the display. Every combination of visual stimuli and aperture was presented 8 times and each trial lasted 10 seconds. The participants’ neural response was monitored using an active high density EEG system.

[0102] The results of the foregoing experiment show that EEG amplitudes at the stimulation frequency in healthy participants were twice as high as those of AMD patients (p<0.001). The Al-generated optimized video increased the SSVEP response in both groups (3x in controls, 4x in patients, p<0.001 ). In both groups, the strongest response to the optimized video was evoked by whole-screen stimulation, followed by extrafoveal and foveal stimulation. The control stimuli entrained narrowband alpha EEG responses in both AMD and healthy populations. However, the optimized Al-generated stimulation was much more effective, and, significantly, extrafoveal stimulation played an important role, making this approach particularly suitable for patients with compromised central vision. Diagram 800 in FIG. 8A and diagram 805 in FIG. 8B collectively show the superiority of Al-generated stimulationover checkerboard flicker with different apertures in healthy controls (FIG. 8A) and AMD patients (FIG. 8B).

[0103] To further explore, the potential of generative video for accelerating perceptual learning and improving functional outcome in AMD, 6 patients (5 female, 1 male, mean age = 74) with dry bilateral intermediate-stage AMD participated in 3 training sessions based on visual perceptual learning. Prior to the first training session and following the last session, their Best Corrected Visual Acuity (BC A) and contrast sensitivity were measured using FrACT7 and their reading ability was measured using MNREAD8 tests. As part of training, patients were given a Lateral Masking task9 in which they were required to identify 3 vs 2 Gabor patches. The task was repeated over 80 trials. Each trial was preceded by 10 seconds of visual stimulation with the optimized generative video. Task performance and visual acuity were evaluated following the last training session (post session). Perceptual performance in the masking task improved, pointing to an improvement in contrast sensitivity. Visual acuity measures before and after training showed improvement in acuity in 4 out of 5 patients - the 6th patient was ill on that day. The BCVA improvements seen after 3 days of training in 4 out of 6 participants may be seen in Table 1 below. Patient PGH004 was ill on the day of training.Table 1

[0104] Graphs 900 in FIG. 9A and graph 905 in FIG. 9B collectively represent the efficacy of entrained visual perceptual learning. Graph 900 in FIG. 9A presents the improved contrast sensitivity after 3 days of training with perceptual task and visual stimulation. Graph 905 in FIG. 9B illustrates that the strength of the entrainments is strongly correlated with the contrast sensitivity as measured post session.

[0105] The research above demonstrates the advantages of Al-generated video optimized for peripheral visual entrainment in maximizing the amplitude of the narrow- band alpha EEG response. The results demonstrate that visual entrainment is at least partially preserved in the presence of retinal damage such as scotomas. The preliminary results thus point to the important potential of non-invasive visual stimulation delivered via Al-generated video to facilitate perceptual training and enhance visual function in AMD patients.

[0106] In this disclosure, the term “based on” means “based at least in part on.” The singular forms “a,” “an,” and “the” include plural referents unless the context dictates otherwise. The term “exemplary” is used in the sense of “example” rather than “ideal.” The terms “comprises,” “comprising,” “includes,” “including,” or other variations thereof, are intended to cover a non-exclusive inclusion such that a process, method, or product that comprises a list of elements does not necessarily include only those elements, but may include other elements not expressly listed or inherent to such a process, method, article, or apparatus. Relative terms, such as “about,” “approximately,” “substantially,” and “generally,” are used to indicate a possible variation of ±10% of a stated or understood value. In addition, the term “between” used in describing ranges of values is intended to include the minimum and maximum values described herein. The use of the term “or” in the claims andspecification is used to mean “and / or” unless explicitly indicated to refer to alternatives only if the alternatives are mutually exclusive, although the disclosure supports a definition that refers to only alternatives and “and / or.” As used herein “another” may mean at least a second or more.

[0107] As used herein, the term “user” generally encompasses any person or entity, such as a researcher and / or a care provider (e.g., a doctor, etc.), that may desire information, resolution of an issue, or engage in any other type of interaction with a provider of the systems and methods described herein (e.g., via an application interface resident on their electronic device, etc.). The term “electronic application” or “application” may be used interchangeably with other terms like “program,” or the like, and generally encompasses software that is configured to interact with, modify, override, supplement, or operate in conjunction with other software.

[0108] Program aspects of the technology may be thought of as “products” or “articles of manufacture” typically in the form of executable code and / or associated data that is carried on or embodied in a type of machine-readable medium. “Storage” type media include any or all of the tangible memory of the computers, processors or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives and the like, which may provide non-transitory storage at any time for the software programming. All or portions of the software may at times be communicated through the Internet or various other telecommunication networks. Such communications, for example, may enable loading of the software from one computer or processor into another, for example, from a management server or host computer of the mobile communication network into the computer platform of a server and / or from a server to the mobile device. Thus, another type of media that may bear the software elements includes optical,electrical and electromagnetic waves, such as used across physical interfaces between local devices, through wired and optical landline networks and over various air-links. The physical elements that carry such waves, such as wired or wireless links, optical links, or the like, also may be considered as media bearing the software. As used herein, unless restricted to non-transitory, tangible “storage” media, terms such as computer or machine “readable medium” refer to any medium that participates in providing instructions to a processor for execution.

[0109] Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would 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.

[0110] Thus, while certain embodiments have been described, those skilled in the art will recognize that other and further modifications may be made thereto without departing from the spirit of the invention, and it is intended to claim all such changes and modifications as falling within the scope of the invention. For example, functionality may be added or deleted from the block diagrams and operations may be interchanged among functional blocks. Steps may be added or deleted to methods described within the scope of the present invention.

[0111] The above disclosed subject matter is to be considered illustrative, and not restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other implementations, which fall within the true spirit and scope of the present disclosure. Thus, to the maximum extent allowed by law, the scope of the present disclosure is to be determined by the broadestpermissible interpretation of the following claims and their equivalents, and shall not be restricted or limited by the foregoing detailed description. While various implementations of the disclosure have been described, it will be apparent to those of ordinary skill in the art that many more implementations are possible within the scope of the disclosure. Accordingly, the disclosure is not to be restricted except in light of the attached claims and their equivalents.

Claims

WHAT IS CLAIMED IS:1 . A computer-implemented method for improving visual function in a patient, the computer-implemented method comprising: providing for presentation, on a monitor associated with a computing device, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, at the computing device and using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; and initiating, at the computing device, a priming stimulation process to improve the visual function in the patient.

2. The method of claim 1 , wherein the neurophysiological data comprises electroencephalography (EEG) data or magnetoencephalography (MEG) data.

3. The method of claim 2, wherein the EEG data is captured from at least one EEG sensor positioned to detect visual cortex activity of the patient.

4. The method of claim 1 , further comprising presenting, at the computing device, a visual perception learning (VPL) task to the patient in associated with the priming stimulation process.

5. The method of claim 4, wherein the initiating the priming stimulation process comprises presenting a priming stimulation video for a predetermined duration before the VPL task.

6. The method of claim 5, wherein the priming stimulation video is generated using a deep generative model trained to optimize visual textures that drive neural activity of the patient toward a target state.

7. The method of claim 4, wherein the VPL task is a crowded trigram task and wherein the presenting the VPL task comprises: presenting a set of letters to the patient that are separated by predetermined relative spacing and positioned at a predetermined distance from the patient in a visual periphery of the patient; and prompting the patient to identify a target letter in the set of letters.

8. The method of claim 7, further comprising varying the predetermined relative spacing between the set of letters between presentation iterations.

9. The method of claim 4, further comprising: detecting, using one or more sensors associated with the device, task performance data of the patient during the VPL task; and adjusting the one or more visual stimuli based on the task performance data to improve visual function over time.

10. The method of claim 9, further comprising:storing the neurophysiological data and the task performance data in a data registry; and optimizing, using the data registry, future visual stimuli.11 . A system for improving visual function in a patient, the system comprising: one or more processors; and one or more computer readable media storing instructions that are executable by the one or more processors to perform operations comprising: providing for presentation, on a computing device associated with the system, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; and initiating a priming stimulation process to improve the visual function in the patient.

12. The system of claim 1 1 , wherein the neurophysiological data comprises electroencephalography (EEG) data or magnetoencephalography (MEG) data.

13. The system of claim 1 1 , wherein the computing device is one of: a smartphone, tablet, or virtual reality headset.

14. The system of claim 1 1 , further comprising presenting a visual perception learning (VPL) task to the patient in association with the priming stimulation process.

15. The system of claim 14, wherein the initiating the priming stimulation process comprises presenting a priming stimulation video for a predetermined duration before the VPL task.

16. The system of claim 15, wherein the priming stimulation video is generated using a deep generative model trained to optimize visual textures that drive neural activity of the patient toward a target state.

17. The system of claim 14, wherein the VPL task is a crowded trigram task and wherein the presenting the VPL task comprises: presenting a set of letters to the patient that are separated by predetermined relative spacing and positioned at a predetermined distance from the patient in a visual periphery of the patient; and prompting the patient to identify a target letter in the set of letters.

18. The system of claim 14, further comprising: detecting task performance data of the patient during the VPL task; and adjusting the one or more visual stimuli based on the task performance data to improve visual function over time.

19. The system of claim 18, further comprising:storing the neurophysiological data and the task performance data in a data registry; and optimizing, using the data registry, future visual stimuli.

20. A non-transitory computer-readable medium storing computer-executable instructions which, when executed by a system, cause the system to perform operations comprising: providing for presentation, on a monitor associated with a computing device, one or more visual stimuli to the patient; receiving, using at least one sensor associated with the computing device, neurophysiological data from the patient during provision of the one or more visual stimuli; determining, at the computing device and using the neurophysiological data, a baseline of neural activity for the patient to determine a neural response to the one or more visual stimuli; and initiating, at the computing device, a priming stimulation process to improve the visual function in the patient.

Images