Personaliized transcranial electrical stimulation using electroencephalographic features

Abstract

Systems, methods, and computer-readable storage media for transcranial electrical stimulation, and more particularly to systems, methods and software for personalized transcranial electrical stimulation using electroencephalographic features. A system can include an electroencephalography (EEG) system, a feature extraction module, an electrical stimulation target module, and an electrical stimulation system. The EEG system can include sensor electrodes to be attached to a subject's scalp and which provide electrical signals. The electrical signals are processed by the feature extraction module, resulting in corresponding electrical signals for a first period of time and features therefrom. The electrical stimulation target designation module can use those to determine stimulation parameters which result in electrical stimulations provided via the electrodes to the subject's scalp.

Description

Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612PERSONALIIZED TRANSCRANIAL ELECTRICAL STIMULATION USING ELECTROENCEPHALOGRAPHIC FEATURESCROSS-REFERENCE

[0001] This application claims priority to 63 / 723,287, filed November 21, 2024, the contents of which are incorporated herein in their entirety.BACKGROUND1. Technical Field

[0002] The currently claimed embodiments of the present invention relate to transcranial electrical stimulation, and more particularly to systems, methods and software for personalized transcranial electrical stimulation using electroencephalographic features.2. Introduction

[0003] EEG. Electroencephalography (EEG) is a powerful non-invasive technique which allows the study of brain electrophysiology and dynamics. It records neural oscillations i.e. local field potential (LFP) which mainly rise from summed postsynaptic potentials of pyramidal cells in parallel alignment. These oscillations are linked to perceptual, cognitive, motor and emotional processes. EEG has been used clinically to diagnose acute lesions, metabolic encephalopathies, mixed type encephalopathy, craniocerebral trauma, central nervous system (CNS) infections such as encephalitis, Creutzfeldt-Jakob disease, subacute sclerosing panencephalitis and purulent local cerebritis, intracranial tumors, CNS poisoning, epilepsy and seizures.

[0004] While visual observation of the patterns of neural oscillations (e.g. slowing down, spikes) is informative, quantitative analysis of the EEG signal holds a plethora of information which is usually untapped clinically. There are typically 5 different EEG bands which vary by their frequency of oscillation: delta (1-3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-30 Hz) and gamma (30-100 Hz). Each one of these bands can be detected at any location over the scalp, and each band can have a different amplitude (i.e. power). Quantitative analysis of the absolute and relative band powers in tandem with accurate spatial localization of these bands can be instrumental for early and accurate diagnosis. A higher power of low frequency bands, such as delta and theta, relative to the other bands correlates to dementia severity and disease progressionApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 from the earliest stages. Lower alpha power (specifically two subcategories of alpha named a2 and a3) relative to control subjects can signify mild Alzheimer’s Disease (AD) while more theta power can indicate vascular dementia. In addition, the synchronicity between left-frontal and left-parietal and between left-frontal and right-inferior-temporal cortices through coherence analysis indicates healthy retrieval during sentence comprehension. As such, coherence analysis has the potential to explain how different brain locations work in conjunction to orchestrate item distinction, verbal working memory and long-term memory.

[0005] Transcranial Electrical Stimulation (tES). Anodal transcranial Direct Current Stimulation and transcranial Alternating Current Stimulation (tDCS and tACS) are non-invasive electrical stimulation techniques which have shown great promise in treating an array of linguistic and memory impairments. TDCS introduces a low voltage stable current over time to cortical areas of the brain, thus facilitating synaptic transmission, and paired with targeted language intervention has improved both verbal and written naming, as well as spelling, for patients with Primary Progressive Aphasia (PPA). In PPA, tDCS has the potential to normalize the functions of cortical areas and their connections. In addition, tDCS over areas of the Default Mode Network, such as the left dorsolateral prefrontal cortex (DLPFC), enhanced visual recognition memory in patients with AD. Conversely, TACS provides a low voltage current that varies rhythmically above and below zero over time with a specific amplitude and frequency of oscillation. The hypothesis on the mechanism of action of tACS is that it entrains underlying brain oscillations (frequency bands) and leads to synaptic changes via spike-timing dependent plasticity mechanisms. TACS stimulation in the gamma frequency range (~40Hz) on bilateral temporal lobes has shown a decrease of intracerebral p-Tau on patients with AD, suggesting a potential novel therapeutic approach. TACS hasn’t been used for patients with PPA yet.

[0006] Therefore, there remains a need for systems, methods and software for personalized transcranial electrical stimulation using electroencephalographic features.SUMMARY

[0007] Additional features and advantages of the disclosure will be set forth in the description that follows, and in part will be understood from the description, or can be learned by practice of the herein disclosed principles. The features and advantages of the disclosure can be realizedApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 and obtained by means of the instruments and combinations particularly pointed out in the appended claims. These and other features of the disclosure will become more fully apparent from the following description and appended claims, or can be learned by the practice of the principles set forth herein.

[0008] Disclosed are systems, methods, and non-transitory computer-readable storage media which provide a technical solution to the technical problem described. A system for personalized transcranial electrical stimulation using electroencephalographic features configured to perform the concepts disclosed herein can include: an electroencephalography (EEG) system; a feature extraction module configured to communicate with said EEG system; an electrical stimulation target designation module configured to communicate with said feature extraction module; and an electrical stimulation system configured to communicate with said electrical stimulation target designation module, wherein said EEG system comprises a plurality of sensor electrodes to be attached to a subject’s scalp and to provide a corresponding plurality of electrical signals, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for a first period of time and extracts a first set of features therefrom, wherein said electrical stimulation target designation module receives said first set of features from said feature extraction module and determines a first set of stimulation parameters, wherein said electrical stimulation system comprises a plurality of stimulation electrodes to be attached to a subject’s scalp and to provide a corresponding plurality of electrical stimulations therethrough, and wherein said electrical stimulation system is further configured to receive said first set of stimulation parameters from said electrical stimulation target designation module and apply a first electrical stimulation to said subject’s scalp based thereon through said plurality of stimulation electrodes.

[0009] A non-transitory computer-readable storage medium, or computation device, configured as disclosed herein can have instructions stored which, when executed by at least one processor, cause a computer or the at least one processor to perform operations which include: receive, by a feature extraction module implemented on said computer, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receive, by an electrical stimulationApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 target designation module implemented on said computer, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and output said first set of stimulation parameters for use by an electrical stimulation system.

[0010] A computerized method for performing personalized transcranial electrical stimulation using electroencephalographic features disclosed herein can include: receiving, by a feature extraction module implemented on a computer, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receiving, by an electrical stimulation target designation module implemented on said computer, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and outputting said first set of stimulation parameters for use by an electrical stimulation system.

[0011] A non-transitory computer-readable storage medium, or computation device, configured as disclosed herein can have instructions stored which, when executed by at least one processor, cause a computer or the at least one processor to perform operations which include: receive, by a feature extraction module implemented on said computer, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receive, by an electrical stimulation target designation module implemented on said computer, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and output said first set of stimulation parameters for use by an electrical stimulation system.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] FIG. 1 illustrates an example of a schematic illustration of a system for personalized transcranial electrical stimulation;

[0013] FIG. 2 illustrates an example of multiple EEG signals;

[0014] FIG. 3 illustrates an example of EEG feature extraction;

[0015] FIG. 4 illustrates an example of EEG feature extraction to identify functional characteristics; andApplicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612

[0016] FIG. 5 illustrates an example computer system.DETAILED DESCRIPTION

[0017] Some embodiments of the current invention are discussed in detail below. In describing embodiments, specific terminology is employed for the sake of clarity. However, the invention is not intended to be limited to the specific terminology so selected. A person skilled in the relevant art will recognize that other equivalent components can be employed, and other methods developed without departing from the broad concepts of the present invention. All references cited anywhere in this specification are incorporated by reference as if each had been individually incorporated. While specific implementations are described, this is done for illustration purposes only. Other components and configurations may be used without parting from the spirit and scope of the disclosure.

[0018] Closed loop EEG-tES system

[0019] Due to heterogeneity in patients’ neural patterns, as well as disease severity and location of cortical atrophy, it is imperative that EEG and transcranial electrical stimulation (tDCS or tACS) be used in tandem as a closed loop system. Cycles of constant detection of a patient’s level of brain activity via EEG and stimulation with a specific current of set amplitude and frequency for a particular time-period based on the recorded brain activity would provide not only a personalized therapeutic approach but also more control and visibility of the effects of electrical stimulation to the intrinsic brain activity. Furthermore, continuous updating of the electrical stimulation protocol based on the underlying brain activity pattern has the potential to take advantage of machine learning algorithms which can recognize and segregate novel patterns as well as auto-correct, thus personalizing treatment to a greater extent.

[0020] The closed-loop electroencephalographic (EEG) and stimulation system according to an embodiment of the current invention is designed to support quasi-real time personalized stimulation for the individual. Contrary to previous closed-loop approaches, this methodology allows determining the pathological area to stimulate in almost real time, and during each loop the area can be changed based on the EEG patterns observed during each EEG recording session.

[0021] The above can be achieved by an embodiment of the current invention by first recording the EEG through a dedicated EEG recording device. (See the schematic illustration of FIG. 1.)Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612The data are subsequently transmitted to a computer that conducts further analysis. Once the EEG data have been received, they are pre-processed to remove noise and spectral content irrelevant to brain activity. Depending on the individual involved in the session, different methods can be used to select the area of stimulation and its type. For example, in the case of a patient with dementia the EEG signal can be analyzed to identify the electrode with the highest ratio of delta power. Through a series of rules, designed based on prior works, the stimulation parameters can be determined and sent to the stimulation device.

[0022] However, decision making methods according to the general concepts of the current invention are not limited to only rule-based methods. For example, and without limitation, neural networks and other classifiers can be used according to other embodiments.

[0023] Some embodiments of the current invention can include features from the following domains: time, frequency, time-frequency, energy-entropy, statistical, graph theory. The features described in some examples are special cases belonging to one of these domains.

[0024] Examples of such parameters are shown in the figure below:

[0025] a) The area or areas to stimulate can be set by using one or more stimulation montages (combinations of anode and cathode electrodes). These montages can either be pre-calculated on a generic or personalized atlas.

[0026] b) The type of stimulation to administer (e.g. tDCS, tACS) can be selected depending on the availability of equipment and therapy / study protocol.

[0027] c) The stimulation amplitude (in mA) can be set prior to the session or determined depending on the degree of the observed pathology.

[0028] d) The stimulation frequency in the case of tACS can be set to the frequency range of the target EEG rhythm.

[0029] e) The stimulation duration and ramp in seconds.

[0030] f) Finally, an extra parameter is whether or not the protocol calls for actual or sham stimulation.

[0031] After each EEG-stimulation loop, the loop can be repeated a specific number of times (e.g. 5 iterations) or for a set time duration (e.g. 30 minutes)

[0032] Example DevicesApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0033] For the creation of a closed loop EEG-tES system we would require an EEG system, a compatible tES system, hardware for proper connections and hardware / software for real-time processing. We have chosen BRAINVISION as our EEG vendor and SOTERIX as our tES vendor. A suitable EEG is actiCHamp Plus 32 System which includes a cap of 32 electrodes. The minimum number of electrodes that can provide sufficient coverage of the entire scalp depends on the stimulation environment. For example, in a laboratory setting 32 electrodes can be used whereas in a home setting 6 or less electrodes could be available. The proposed system will still function properly albeit with decreased spatial resolution. For the tES system, the SOTERIX Medical MxN-33 HD-tES Custom Waveform Generator which activate up to 32 HD active electrodes (usually the number of active electrodes is less than 10) is suitable and is compatible with the BRAIN VISION system for application of either tDCS or tACS.

[0034] Description of software in steps.

[0035] The following describes some possible software steps according to some embodiments of the current invention. The general concepts of the current invention are not limited to these particular steps.

[0036] Step 1. EEG Pre-processing

[0037] Current implementation:

[0038] The purpose of this step is to remove spectral content irrelevant to brain activity. This is achieved by utilizing using low-order Butterworth filters in the following order:

[0039] a) Notch filter centered on the powerline frequency (depending on region of operation and only if the device is directly connected to the power mains)

[0040] b) Notch filter centered on the first harmonic of the powerline frequency (depending on region of operation and only if the device is directly connected to the power mains)

[0041] c) Notch filter centered on the second harmonic of the powerline frequency (depending on region of operation and only if the device is directly connected to the power mains)

[0042] d) High pass filter to remove low-frequency spectral content with cutoff threshold set at 0.5 Hz, and

[0043] e) Low pass filter to remove high-frequency spectral content with cutoff threshold set at 70 Hz.

[0044] Possible modifications:Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0045] 1) There is no consensus among researchers with regards to the types of filters, filter parameters or frequency thresholds. Such guidelines could be generated by future research.

[0046] 2) While the above combination of filters has been used in various prior applications, e.g. sleep staging and brain age assessment, in the end, the selection of the number of filters and cutoff frequencies should be varied depending on the recording setup and research questions. For example, i) if the recording device is battery powered the first three filters (a, b, and c) can be omitted and ii) if the research question concerns brain rhythms over 70 Hz the low pass filter cutoff threshold (e) should be adjusted accordingly.

[0047] 3) Independent Component Analysis (ICA) is a method that estimates statistically independent signal sources, implemented but not currently used, has been proven useful. However, it requires significant computational time, significant EEG time length and manual ICA component rejection. With the progress of hardware, analytical methods and artificial intelligence, all the above could be addressed in the near future, and hence ICA could be incorporated into the closed-loop system.

[0048] 2. EEG Feature extraction

[0049] Current implementation:

[0050] The feature extraction methods that are currently implemented in the closed-loop system and used to select the stimulation area are:

[0051] a) EEG rhythm power features: Energy ratios of the five main EEG rhythms (delta, theta, alpha, beta and gamma) are calculated along with ratios between rhythm ratios such as delta over gamma.

[0052] b) EEG Functional connectivity features: The Weighted Phase Lag Index (WPLI) derived functional connectivity is calculated, and

[0053] c) EEG derived graph metrics: Betweenness centrality and small world metric are used.

[0054] These can be used to analyze the EEG but also to estimate its quality (presence of noise, electrode movement etc.). Based on the values of these features, the most advantageous for the patient simulation region can be selected.

[0055] Possible modifications:

[0056] 1) There are several methods that enable the extraction of informative EEG features, for example time domain features, spectral domain features, time-frequency features, power-entropyApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 features, functional connectivity features, and graph theory metrics. Each one of the above could be more appropriate depending on the patient group.

[0057] 3. EEG Source estimation

[0058] Current implementation:

[0059] Brain region activation is possible using the standardized brain map Eve283 (or any other available map), and utilizing sLORETA. The steps, implemented to achieve this are as follows:

[0060] a) Record an EEG segment,

[0061] b) Convert the electrode activations into brain bipole activations, achieved usually through matrix multiplication,

[0062] c) Identify all bipoles belonging to the region of interest, commonly provided by the current atlas,

[0063] d) Calculate the average bipole activation, which is henceforth the estimated region activation.

[0064] e) Decide the stimulation area and its respective montage, and

[0065] f) Run the stimulation.

[0066] This methodology provides even more accurate and personalized stimulation sessions but requires the calculation of individual patient data transforms between the mediums i.e. Magnetic Resonance Imaging (MRI) space to EEG coordinates and brain region to montage, which however need not be calculated during the stimulation session but can be calculated beforehand, especially since they require a significant amount of time.

[0067] Possible modifications:

[0068] 1) Brain region activation is also possible using each patient’s MRI to construct highly personalized, and hence more accurate, head models.

[0069] 2) Having the activation of specific brain regions and combined with tools such as ROAST it is possible to derive dynamic and accurate stimulation targeting of deeper brain structures compared to stimulation targets selected through electrode-based analysis.

[0070] 4. Stimulation target area selection

[0071] Current implementation:

[0072] Real time target selection is achieved by taking into account the features calculated on a short duration EEG recording e.g. 1 minute (the duration of the EEG recording is set by theApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 user). The method currently implemented utilizes a set of rule-based decision-making agents that are simple to design, run in very little time and are humanly interpretable. Depending on the disease or syndrome afflicting each patient, the proper method for determining the stimulation area must be selected by the clinician. At present, nine (9) different methods of area selection have been implemented:

[0073] a) Area with maximum delta rhythm

[0074] b) Area with maximum delta rhythm compared to aged-matched control group

[0075] c) Area with minimum gamma rhythm

[0076] d) Area with minimum gamma rhythm compared to aged-matched control group

[0077] e) Area with maximum delta over gamma rhythm

[0078] f) Area with maximum delta over gamma rhythm compared to aged-matched control group

[0079] g) Area with maximum brain rhythm energy difference compared to aged-matched control group

[0080] h) Area with maximum WPLI derived functional connectivity difference compared to aged-matched control group

[0081] i) Area with maximum betweenness centrality difference compared to aged-matched control group.

[0082] In the case of the methods that comparisons are made with the aged-matched control group, the process is not simply finding the maximum difference:

[0083] a) Firstly, the EEG features of the patient are compared to the healthy cohort to find the healthy control that is most similar to the patient.

[0084] b) Then the area of maximum difference is identified in order to provide a higher degree of intervention personalization.

[0085] Possible modifications:

[0086] 1) The artificial intelligence (Al) methods that can be used are unrestricted and can be selected based on which the healthcare practitioner considers most accurate for each patient case. Non-limiting examples of Al methods can include the use of neural networks, machine learning, etc., wherein those Al methods are converted to machine-executable code in the form of one or more Al agents.Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0087] 2) The Al agent, using the current EEG information and prior knowledge attained during training, can select both the most suitable area and montage for stimulation.

[0088] 5. Personalization

[0089] Current implementation:

[0090] The use of the implemented closed-loop system currently requires the clinician to input crucial information into the system by hand. Such parameters are the:

[0091] a) EEG sampling frequency, dependent on recording / device parameters

[0092] b) powerline frequency, dependent on region

[0093] c) anode intensity, crucial for the safety of the equipment and patient

[0094] d) stimulation duration, crucial for the safety of the patient and stimulation effectiveness

[0095] e) number of EEG-tDCS loops, dependent on the disease, syndrome or application

[0096] f) length of EEG recording, dependent on the disease, syndrome or application

[0097] g) method used to select the stimulation region, dependent on the disease, syndrome or application.

[0098] Possible modifications:

[0099] 1) With regards to items a and b, these can be communicated through the individual devices where supported.

[0100] 2) Currently items c, d, e are set by the user, once the efficacy of the system has been determined, these can also be set by an Al agent, within hard limits aiming to protect the patient and involved devices. For example, if a brain region is heavily compromised, i) the anode intensity can be set to a maximum of 4 mA, even if the system estimates that a higher value is required, ii) the maximum stimulation duration could be set to 10 minutes, irrespective of higher system estimates, and iii) the maximum number of EEG-tDCS loops could be set to 10 due to time constraints.

[0101] 3) Since items f and g are directly linked with the pathophysiology of the patient and due to the limits of current knowledge. However, after verifying the validity of the system’s recommendations, efforts could be made to automate their selection.

[0102] Discussion

[0103] The automatic selection of the target area based on the EEG signal enables personalized stimulation in the brain regions that present the highest abnormality based on EEG functionalApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 connectivity. This is not determined a priori, based on general knowledge or previous works, but determined for each patient individually. This is further augmented from the fact that the region may change between each EEG-tDCS loop, rendering its application not individual specific but also time sensitive. This is especially useful if stimulation is provided during a task or set of tasks each of which may indicate abnormality in different brain regions. However, such variations may also be present in non-task (resting-state) stimulation scenarios, since the resting brain state is hardly static. Another way that increases the effectiveness of stimulation on a disease or syndrome level is setting a group of regions of interest that should be stimulated (selected by the clinician), thereby limiting region selection to the ones that are most pertinent for each group of patients.

[0104] 6. Inter-device data transfer

[0105] There are a multitude of ways that data can be transferred between the EEG, tDCS and stimulating computer. In the current implementation, EEG data is transferred from the EEG recording device to the stimulation computer via the Transmission Control Protocol (TCP). Once the EEG data is received, stored, analyzed and a decision has been made on the stimulation area and montage, the parameters are sent to the stimulation device through the Lab Streaming Layer (LSL).

[0106] Possible modifications:

[0107] 1) As new devices are made available, the system can be revised to support multiple data transfer protocols and device combinations.

[0108] In some embodiments all data recorded by the EEG devices as well as prior data used for the decision making can stored in the same system. Therefore, the EEG and stimulation devices can be replaced without any alteration required in terms of data analysis; only the communication method between the devices will be modified.

[0109] Accordingly, FIG. 1 provides a schematic illustration of a system for personalized transcranial electrical stimulation using electroencephalographic features 100 according to an embodiment of the current invention. The system 100 includes an electroencephalography (EEG) system 102; a feature extraction module 104 configured to communicate with the EEG system; an electrical stimulation target designation module 106 configured to communicate with said feature extraction module 104; and an electrical stimulation system 108 configured toApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 communicate with the electrical stimulation target designation module 106. Some embodiments can also include an EEG preprocessing module.

[0110] The EEG system 102 includes a plurality of sensor electrodes (not shown in FIG. 1) to be attached to a subject’s scalp and to provide a corresponding plurality of electrical signals. The feature extraction module 104 receives and processes the plurality of electrical signals from the EEG system 102 for a first period of time and extracts a first set of features therefrom. The electrical stimulation target designation module 106 receives the first set of features from the feature extraction module 104 and determines a first set of stimulation parameters therefrom. The electrical stimulation system 108 includes a plurality of stimulation electrodes (not shown in FIG. 1) to be attached to a subject’s scalp and to provide a corresponding plurality of electrical stimulations therethrough. The stimulation electrodes can be separate from the sensor electrodes in some embodiments, or they may share some or all of the same electrodes for both sensing and stimulation in some embodiments. The electrical stimulation system 108 is further configured to receive the first set of stimulation parameters from the electrical stimulation target designation module 106 and apply a first electrical stimulation to the subject’s scalp based thereon through the plurality of stimulation electrodes. This can be repeated a plurality of times to thereby provide closed loop tDCS and / or tACS. The general concepts of the current invention are not limited to the number of times this is repeated nor to the period of time between measurements and stimulations.[OHl] Some embodiments of the current invention can include an EEG preprocessing module 110 to preprocess the plurality of electrical signals prior to providing the preprocessed signals to the feature extraction module 104. The feature extraction module 104, the electrical stimulation target designation module 106, and the EEG preprocessing module 110 can each be implemented on one or more computers either through software and / or hard wiring, for example.

[0112] FIG. 2 illustrates an example of multiple EEG signals, the EEG signals having distinct EEG rhythms.

[0113] FIG. 3 illustrates an example of EEG feature extraction. The EEG features can be extracted through two distinct pathways. The pathways as illustrated can be executed in parallel or in series, depending on specific configuration.Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0114] As illustrated in the top row, the system can first identify a synchronization likelihood. This is a generalized synchronization metric among two time-series, and it represents the level of synchronization of channel k at time i with reference to all other channels. The generalized synchronization metric can be a weighted undirected graph adjacency matrix, which is symmetric with reference to the diagonal. For example, the system can calculate the generalized synchronization metric via a four-part process by calculating: 1) Time delay embedded series as an input vector; 2) A closeness probability of two embedded vectors; 3) Number of time series with co-operative activity higher than a threshold; and 4) Estimation of synchronization likelihood, resulting in the matrix as illustrated (i.e., the symmetric graph in the top row).

[0115] In a second pathway / process, the system identifies relative wavelet entropy. Here, the values of this matrix represent the extent to which the energy distribution (delta, theta, alpha, beta, and gamma rhythmic activity) from each pair of electrodes are similar to one another. These differences are transformed, using the orthogonal discrete waveform transform into a non- symmetric, weighted, direction graph (i.e., the non-symmetric graph in the lower row), with estimates of the energy frequency bands, relative contribution, etc. illustrated by the level of grayscale.

[0116] The two graphs together can then be used to form nodes / edges in the form of graphs. Such graphs can be multi-dimensional, as illustrated.

[0117] FIG. 4 illustrates an example of EEG feature extraction to identify functional characteristics. In this example, different distributions of the gamma, beta, detla, and whole EEG signals can be used to identify differences between a healthy individual and one which is pathological.

[0118] With reference to FIG. 5, an exemplary system includes a computing device 500 (such as a general-purpose computing device), including a processing unit (CPU or processor) 520 and a system bus 510 that couples various system components including the system memory 530 such as read-only memory (ROM) 540 and random access memory (RAM) 550 to the processor 520. The computing device 500 can include a cache of high-speed memory connected directly with, in close proximity to, or integrated as part of the processor 520. The computing device 500 copies data from the system memory 530 and / or the storage device 560 to the cache for quick access by the processor 520. In this way, the cache provides a performance boost that avoids processorApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612520 delays while waiting for data. These and other modules can control or be configured to control the processor 520 to perform various actions. Other system memory 530 may be available for use as well. The system memory 530 can include multiple different types of memory with different performance characteristics. It can be appreciated that the disclosure may operate on a computing device 500 with more than one processor 520 or on a group or cluster of computing devices networked together to provide greater processing capability. The processor 520 can include any general-purpose processor and a hardware module or software module, such as module 1 562, module 2 564, and module 3 566 stored in storage device 560, configured to control the processor 520 as well as a special-purpose processor where software instructions are incorporated into the actual processor design. The processor 520 may essentially be a completely self-contained computing system, containing multiple cores or processors, a bus, memory controller, cache, etc. A multi-core processor may be symmetric or asymmetric.

[0119] The system bus 510 may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures. A basic input / output (BIOS) stored in memory ROM 540 or the like, may provide the basic routine that helps to transfer information between elements within the computing device 500, such as during start-up. The computing device 500 further includes storage devices 560 such as a hard disk drive, a magnetic disk drive, an optical disk drive, tape drive or the like. The storage device 560 can include software modules 562, 564, 566 for controlling the processor 520. Other hardware or software modules are contemplated. The storage device 560 is connected to the system bus 510 by a drive interface. The drives and the associated computer- readable storage media provide nonvolatile storage of computer-readable instructions, data structures, program modules and other data for the computing device 500. In one aspect, a hardware module that performs a particular function includes the software component stored in a tangible computer-readable storage medium in connection with the necessary hardware components, such as the processor 520, system bus 510, output device 570 (such as a display or speaker), and so forth, to carry out the function. In another aspect, the system can use a processor and computer-readable storage medium to store instructions which, when executed by a processor (e.g., one or more processors), cause the processor to perform a method or other specific actions. The basic components and appropriate variations are contemplated dependingApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 on the type of device, such as whether the computing device 500 is a small, handheld computing device, a desktop computer, or a computer server.

[0120] Although the exemplary embodiment described herein employs the storage device 560 (such as a hard disk), other types of computer-readable media which can store data that are accessible by a computer, such as magnetic cassettes, flash memory cards, digital versatile disks, cartridges, random access memories (RAMs) 550, and read-only memory (ROM) 540, may also be used in the exemplary operating environment. Tangible computer-readable storage media, computer-readable storage devices, or computer-readable memory devices, expressly exclude media such as transitory waves, energy, carrier signals, electromagnetic waves, and signals per se.

[0121] To enable user interaction with the computing device 500, an input device 590 represents any number of input mechanisms, such as a microphone for speech, a touch-sensitive screen for gesture or graphical input, keyboard, mouse, motion input, speech and so forth. An output device 570 can also be one or more of a number of output mechanisms known to those of skill in the art. In some instances, multimodal systems enable a user to provide multiple types of input to communicate with the computing device 500. The communications interface 580 generally governs and manages the user input and system output. There is no restriction on operating on any particular hardware arrangement and therefore the basic features here may easily be substituted for improved hardware or firmware arrangements as they are developed.

[0122] The computing device 500 may be described in the general context of computer systemexecutable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, and so on that perform particular tasks or implement particular abstract data types. In configurations where the computing device 500 is used in a distributed cloud computing environment (such as where the computing device 500 utilizes one or more servers) where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media including memory storage devices.

[0123] The technology discussed herein refers to computer-based systems and actions taken by, and information sent to and from, computer-based systems. One of ordinary skill in the art willApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 recognize that the inherent flexibility of computer-based systems allows for a great variety of possible configurations, combinations, and divisions of tasks and functionality between and among components. For instance, processes discussed herein can be implemented using a single computing device or multiple computing devices working in combination. Databases, memory, instructions, and applications can be implemented on a single system or distributed across multiple systems. Distributed components can operate sequentially or in parallel.

[0124] Neural networks, foundational to modem artificial intelligence, are computational systems designed to process data and generate predictions or classifications by emulating aspects of human brain function. A neural network is a framework of machine learning algorithms that work together to classify inputs based on a previous training process. They power applications like image recognition, natural language processing, and predictive analytics. At their core, neural networks consist of interconnected layers of mathematical units called neurons, organized into an input layer, one or more hidden layers, and an output layer. The input layer receives raw or preprocessed data, such as pixel values or text embeddings, represented as numerical vectors. Hidden layers transform this data into increasingly abstract representations through complex computations, while the output layer produces the result, such as a class probability or a numerical prediction. Each neuron connects to those in the next layer via weighted connections, where weights are numerical values that amplify or diminish the influence of one neuron’s output on another’s input. Additionally, biases — adjustable offsets — enhance the model’s flexibility in fitting data.

[0125] The operation of a neural network begins with a forward pass, where data flows from the input layer through the hidden layers to the output. Each neuron computes a weighted sum of its inputs, adds its bias, and applies a nonlinear activation function, such as a sigmoid, rectified linear unit (ReLU), or hyperbolic tangent (tanh), to produce an output. This process repeats across layers, with each layer extracting more complex features, such as edges in images or semantic patterns in text. The final layer’s output depends on the task: classification tasks yield probabilities (e.g., “90%”), while regression tasks produce continuous values (e.g., a predicted temperature). Crucially, the forward pass does not alter the model’s stored parameters — weights and biases — which represent the network’s learned knowledge. These parameters are stored in digital memory, typically as 32-bit or 16-bit floating-point arrays. Weights form matrices, withApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 rows and columns corresponding to neurons in adjacent layers, while biases are stored as onedimensional arrays. Meta-information, such as layer counts and activation function types, is also stored to define the network’s structure.

[0126] Training a neural network involves adjusting its parameters to minimize prediction errors. During training, a forward pass generates predictions, which are compared to correct outputs using a loss function, such as mean squared error or cross-entropy, to quantify errors. Backpropagation then computes gradients, indicating how much each parameter contributed to the error, by applying the chain rule to propagate errors backward from the output to the input layer. Optimization algorithms, like stochastic gradient descent, adjust weights and biases in directions that reduce the loss. This process iterates over multiple epochs, with parameters gradually converging to values that improve accuracy. Memory usage during training is dynamic: weights and biases are updated incrementally for each data batch, and intermediate results, like neuron activations and gradients, are temporarily stored in buffers to facilitate backpropagation. To ensure progress is saved, parameters are periodically checkpointed to persistent storage, allowing training to resume later. Efficiency techniques, such as reducing parameter precision to 16-bit formats, further optimize memory and computation.

[0127] Once trained, the network enters inference mode, where parameters are fixed, and only forward passes are executed to generate predictions. This mode minimizes memory writes, making it ideal for deployment on resource-constrained devices like mobile phones. Neural networks can reduce memory usage, use unique parameter update mechanisms to enhance training efficiency, use hybrid memory systems combining volatile and non-volatile storage, and / or perform dynamic precision adjustments during training or inference.

[0128] Use of language such as “at least one of X, Y, and Z,” “at least one of X, Y, or Z,” “at least one or more of X, Y, and Z,” “at least one or more of X, Y, or Z,” “at least one or more of X, Y, and / or Z,” or “at least one of X, Y, and / or Z,” are intended to be inclusive of both a single item (e.g., just X, or just Y, or just Z) and multiple items (e.g., {X and Y}, {X and Z}, {Y and Z}, or {X, Y, and Z}). The phrase “at least one of’ and similar phrases are not intended to convey a requirement that each possible item must be present, although each possible item may be present.Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0129] The various embodiments described above are provided by way of illustration only and should not be construed to limit the scope of the disclosure. Various modifications and changes may be made to the principles described herein without following the example embodiments and applications illustrated and described herein, and without departing from the spirit and scope of the disclosure. For example, unless otherwise explicitly indicated, the steps of a process or method may be performed in an order other than the example embodiments discussed above. Likewise, unless otherwise indicated, various components may be omitted, substituted, or arranged in a configuration other than the example embodiments discussed above.

[0130] Further aspects of the present disclosure are provided by the subject matter of the following clauses.

[0131] REFERENCES

[0132] Boggio, P. S., Khoury, L. P., Martins, D. C., Martins, O. E., Macedo, E. C. de, & Fregni, F. (2009). Temporal cortex direct current stimulation enhances performance on a visual recognition memory task in Alzheimer disease. Journal of Neurology, Neurosurgery, and Psychiatry, 80(4), 444-447.

[0133] Cohen, M. X. (2017). Where Does EEG Come From and What Does It Mean? Trends in Neurosciences, 40(4), 208-218. https: / / doi.Org / 10.1016 / j.tins.2017.02.004

[0134] Cotelli, M., Manenti, R., Ferrari, C., Gobbi, E., Macis, A., & Cappa, S. F. (2020). Effectiveness of language training and non-invasive brain stimulation on oral and written naming performance in Primary Progressive Aphasia: A meta-analysis and systematic review. Neuroscience & Biobehavioral Reviews, 108, 498-525. https: / / doi.Org / 10.1016 / j.neubiorev.2019.12.003

[0135] Dhaynaut, M., Cappon, D., Paciorek, R., Macone, J., Connor, A., Guehl, N., ... Santarnecchi, E. (2020). Effects of modulating gamma oscillations via 40Hz transcranial alternating current stimulation (tACS) on Tau PET imaging in mild to moderate Alzheimer’s Disease< / strong>. Journal of Nuclear Medicine, 61(supplement 1), 340.

[0136] Fenner, A. S., Webster, K. T., Ficek, B. N., Frangakis, C. E., & Tsapkini, K. (2019). Written Verb Naming Improves After tDCS Over the Left IFG in Primary Progressive Aphasia. Frontiers in Psychology, 10. https: / / doi.org / 10.3389 / fpsyg.2019.01396Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0137] Ficek, B. N., Wang, Z , Zhao, Y., Webster, K. T., Desmond, J. E., Hillis, A. E., ... Tsapkini, K. (2018). The effect of tDCS on functional connectivity in primary progressive aphasia. NeuroImage: Clinical.

[0138] Gervits, F., Ash, S., Coslett, H. B., Rascovsky, K., Grossman, M., & Hamilton, R.(2016). Transcranial direct current stimulation for the treatment of primary progressive aphasia: An open-label pilot study. Brain and Language, 162, 35-41.

[0139] Herrmann, C. S., Rach, S., Neuling, T., & Striiber, D. (2013). Transcranial alternating current stimulation: A review of the underlying mechanisms and modulation of cognitive processes. Frontiers in Human Neuroscience, 7. https: / / doi.org / 10.3389 / fnhum.2013.00279

[0140] Leite, J., Morales-Quezada, L., Carvalho, S., Thibaut, A., Doruk, D., Chen, C.-F., ... Fregni, F. (2017). Surface EEG-Transcranial Direct Current Stimulation (tDCS) Closed-Loop System. International Journal ofNeural Systems, 27(06), 1750026. https: / / doi.org / 10.1142 / S0129065717500265

[0141] Marson, F., Lasaponara, S., & Cavallo, M. (2021). A Scoping Review of Neuromodulation Techniques in Neurodegenerative Diseases: A Useful Tool for Clinical Practice? Medicina, 57(3), 215. https: / / doi.org / 10.3390 / medicina57030215

[0142] McConathey, E. M., White, N. C., Gervits, F., Ash, S., Coslett, H, Grossman, M., & Hamilton, R. H. (2017). Baseline performance predicts tdcs-mediated improvements in language symptoms in primary progressive aphasia. Frontiers in Human Neuroscience, 11, 347.

[0143] Meyer, L., Grigutsch, M., Schmuck, N., Gaston, P., & Friederici, A. D. (2015). Frontal- posterior theta oscillations reflect memory retrieval during sentence comprehension. Cortex, 71, 205-218. https: / / doi.Org / 10.1016 / j.cortex.2015.06.027

[0144] Mitzdorf, U. (1985). Current source-density method and application in cat cerebral cortex: Investigation of evoked potentials and EEG phenomena. Physiological Reviews, 65(1), 37-100. https: / / doi.org / 10.! 152 / physrev.1985.65.1.37

[0145] Moretti, D. (2004). Individual analysis of EEG frequency and band power in mild Alzheimer’s disease. Clinical Neurophysiology, 115(2), 299-308. https: / / doi.org / 10.1016 / S1388- 2457(03)00345-6

[0146] Niedermeyer, E. (2003). The Clinical Relevance of EEG Interpretation. Clinical Electroencephalography, 34(3), 93-98. https: / / doi.org / 10.1177 / 155005940303400303Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0147] Nissim, N. R., Moberg, P. J., & Hamilton, R. H. (2020). Efficacy of Noninvasive Brain Stimulation (tDCS or TMS) Paired with Language Therapy in the Treatment of Primary Progressive Aphasia: An Exploratory Meta- Analysis. Brain Sciences, 10(9), 597. https: / / doi.org / 10.3390 / brainscil0090597

[0148] Nobili, F., Copello, F., Vitali, P., Prastaro, T., Carozzo, S., Perego, G., & Rodriguez, G.(1999). Timing of Disease Progression by Quantitative EEG in Alzheimer’s Patients. Journal of Clinical Neurophysiology, 16(6). Retrieved from https: / / journals.lww.eom / clinicalneurophys / Fulltext / 1999 / l 1000 / Timing_of_Disease_Progressio n_by_Quantitative_EEG.8. aspx

[0149] Roncero, C., Kniefel, H., Service, E., Thiel, A., Probst, S., & Chertkow, H. (2017). Inferior parietal transcranial direct current stimulation with training improves cognition in anomic Alzheimer’s disease and frontotemporal dementia. Alzheimer’s & Dementia: Translational Research & Clinical Interventions, 3(2), 247-253.

[0150] Schreiter-Gasser, U., Gasser, T., & Ziegler, P. (1994). Quantitative EEG analysis in early onset Alzheimer’s disease: Correlations with severity, clinical characteristics, visual EEG and CCT. Electroencephalography and Clinical Neurophysiology, 90(4), 267-272. https: / / doi .org / 10.1016 / 0013 -4694(94)90144-9

[0151] Siegel, M., Donner, T. H., & Engel, A. K. (2012). Spectral fingerprints of large-scale neuronal interactions. Nature Reviews Neuroscience, 13(2), 121-134. https : / / doi . org / 10.1038 / nrn3137

[0152] Tsapkini, K., Frangakis, C., Gomez, Y., Davis, C., & Hillis, A. E. (2014). Augmentation of spelling therapy with transcranial direct current stimulation in primary progressive aphasia: Preliminary results and challenges. Aphasiology, 28(8-9), 1112-1130. https: / / doi.org / 10.1080 / 02687038.2014.930410

[0153] Chriskos, P., Frantzidis, C. A., Nday, C. M., Gkivogkli, P. T., Bamidis, P. D., & Kourtidou-Papadeli, C. (2021). A review on current trends in automatic sleep staging through bio-signal recordings and future challenges. Sleep medicine reviews, 55, 101377.

[0154] Chriskos, P., Frantzidis, C. A., Gkivogkli, P. T., Bamidis, P. D., & Kourtidou-Papadeli, C. (2019). Automatic sleep staging employing convolutional neural networks and corticalApplicant Ref : C18121_ P18121-02Attorney Docket 002240.615612 connectivity images. IEEE transactions on neural networks and learning systems, 3 / (1), 113- 123.

[0155] Chriskos, P., Frantzidis, C. A., Papanastasiou, E., & Bamidis, P. D. (2021). Applications of Convolutional Neural Networks in neurodegeneration and physiological aging. International Journal of Psychophysiology, 159, 1-10.

[0156] Chriskos, P., Frantzidis, C. A., Gkivogkli, P. T., Papanastasiou, E., Kourtidou-Papadeli, C., & Bamidis, P. D. (2019, July). SmartHypnos: Developing a Toolbox for Polysomnographic Data Visualization and Analysis. In 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 1395-1398). IEEE.

[0157] Chriskos, P., Frantzidis, C. A., Gkivogkli, P. T., Bamidis, P. D., & Kourtidou-Papadeli, C. (2018). Achieving accurate automatic sleep staging on manually pre-processed EEG data through synchronization feature extraction and graph metrics. Frontiers in human neuroscience, 12, 110.

[0158] Tsinalis, O., Matthews, P. M., & Guo, Y. (2016). Automatic sleep stage scoring using time-frequency analysis and stacked sparse autoencoders. Annals of biomedical engineering, 44, 1587-1597.

[0159] Hassan, A. R., & Bhuiyan, M. I. H. (2016). A decision support system for automatic sleep staging from EEG signals using tunable Q-factor wavelet transform and spectral features. Journal of neuroscience methods, 271, 107-118.

[0160] Sousa, T., Cruz, A., Khalighi, S., Pires, G., & Nunes, U. (2015). A two-step automatic sleep stage classification method with dubious range detection. Computers in biology and medicine, 59, 42-53.

[0161] Hardmeier, M., Hatz, F., Bousleiman, H., Schindler, C., Stam, C. J., & Fuhr, P. (2014). Reproducibility of functional connectivity and graph measures based on the phase lag index (PLI) and weighted phase lag index (wPLI) derived from high resolution EEG. PloS one, 9(10), C108648.

[0162] Makarov, V. V., Zhuravlev, M. O., Runnova, A. E., Protasov, P., Maksimenko, V. A., Frolov, N. S., ... & Hramov, A. E. (2018). Betweenness centrality in multiplex brain network during mental task evaluation. Physical Review E, 98(6), 062413.Applicant Ref : C18121_ P18121-02Attorney Docket 002240.615612

[0163] Smit, D. J., Stam, C. J., Posthuma, D., Boomsma, D. I., & De Geus, E. J. (2008).Heritability of “small-world” networks in the brain: A graph theoretical analysis of resting-state EEG functional connectivity. Human brain mapping, 29(12), 1368-1378.

[0164] Pascual-Marqui, R. D. (2002). Standardized low-resolution brain electromagnetic tomography (sLORETA): technical details. Methods Find Exp Clin Pharmacol, 2 - / (Suppl D), 5- 12.

[0165] Huang, Y., Datta, A., Bikson, M., & Parra, L. C. (2019). Realistic volumetric-approach to simulate transcranial electric stimulation — ROAST — a fully automated open-source pipeline.Journal of neural engineering, 16(5), 056006.

[0166] Leung, K. C., Li, V. O., & Yang, D. (2007). An overview of packet reordering in transmission control protocol (TCP): problems, solutions, and challenges. IEEE transactions on parallel and distributed systems, 18(A), 522-535.

[0167] Kothe, C., Shirazi, S. Y., Stenner, T., Medine, D., Boulay, C., Crivich, M. I., ... & Makeig, S. (2024). The Lab Streaming Layer for Synchronized Multimodal Recording. bioRxiv, 2024-02.

Claims

Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612CLAIMSWe claim:

1. A system for personalized transcranial electrical stimulation using electroencephalographic features, comprising: an electroencephalography (EEG) system; a feature extraction module configured to communicate with said EEG system; an electrical stimulation target designation module configured to communicate with said feature extraction module; and an electrical stimulation system configured to communicate with said electrical stimulation target designation module, wherein said EEG system comprises a plurality of sensor electrodes to be attached to a subject’s scalp and to provide a corresponding plurality of electrical signals, wherein said feature extraction module receives and processes said corresponding plurality of electrical signals from said EEG system for a first period of time and extracts a first set of features therefrom, wherein said electrical stimulation target designation module receives said first set of features from said feature extraction module and determines a first set of stimulation parameters, wherein said electrical stimulation system comprises a plurality of stimulation electrodes to be attached to a subject’s scalp and to provide a corresponding plurality of electrical stimulations therethrough, and wherein said electrical stimulation system is further configured to receive said first set of stimulation parameters from said electrical stimulation target designation module and apply a first electrical stimulation to said subject’s scalp based thereon through said plurality of stimulation electrodes.

2. The system according to claim 1, wherein at least some of said plurality of sensor electrodes and said plurality of stimulation electrodes are used for both sensing and stimulation.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.6156123. The system according to claim 1 or 2, wherein said feature extraction module is further configured to process each of said corresponding plurality of electrical signals into a plurality of frequency bands.

4. The system according to claim 3, wherein said plurality of frequency bands are delta (1-3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-30 Hz) and gamma (30-100 Hz) bands.

5. The system according to claim 4, wherein said first set of features comprises at least one of energy ratios of between at least one of said delta, theta, alpha, beta and gamma bands.

6. The system according to claim 4, wherein said first set of features comprises a functional connectivity feature type.

7. The system according to claim 6, wherein said functional connectivity feature type is at least one of a Weighted Phase Lag Index (WPLI) derived functional connectivity between said corresponding plurality of electrical signals from said EEG system, a synchronization likelihood or a relative wavelet entropy.

8. The system according to claim 4, wherein said first set of features comprises a graph metric.

9. The system according to claim 8, wherein said graph metric comprises at least one of a betweenness centrality and small world metric based on said corresponding plurality of electrical signals from said EEG system.

10. The system according to any one of claims 1-3, wherein said electrical stimulation target designation module comprises stimulation rules, andApplicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612 wherein said electrical stimulation target designation module is configured to determine said first set of stimulation parameters based on applying said stimulation rules to said first set of features.

11. The system according to any one of claims 4-9, wherein said electrical stimulation target designation module comprises stimulation rules, and wherein said electrical stimulation target designation module is configured to determine said first set of stimulation parameters based on applying said stimulation rules to said first set of features.

12. The system according to claim 11, wherein said stimulation rules comprise at least one of an area with maximum delta rhythm, an area with maximum delta rhythm compared to an age- matched control group, an area with minimum gamma rhythm, an area with minimum gamma rhythm compared to an age-matched control group, an area with maximum delta over gamma rhythm, an area with maximum delta over gamma rhythm compared to an age-matched control group, an area with maximum brain rhythm energy difference compared to an age-matched control group, an area with maximum WPLI derived functional connectivity difference compared to an age-matched control group, or an area with maximum betweenness centrality difference compared to an age-matched control group.

13. The system according to any one of claims 1-12, wherein said feature extraction module receives and processes said corresponding plurality of electrical signals from said EEG system for a second period of time and extracts a second set of features therefrom, wherein said electrical stimulation target designation module receives said second set of features from said feature extraction module and determines a second set of stimulation parameters, wherein said electrical stimulation system receives said second set of stimulation parameters from said electrical stimulation target designation module and applies a second electrical stimulation to said subject’s scalp based thereon, and wherein said second period of time is subsequent to said first period of time and within a same treatment session.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.61561214. The system according to claim 13, wherein said feature extraction module receives and processes said corresponding plurality of electrical signals from said EEG system for an nth period of time and extracts an nth set of features therefrom, wherein said electrical stimulation target designation module receives said nth set of features from said feature extraction module and determines an nth set of stimulation parameters, wherein said electrical stimulation system receives said nth set of stimulation parameters from said electrical stimulation target designation module and applies an nth electrical stimulation to said subject’s scalp based thereon, and wherein said nth period of time is subsequent to an (n-l)th period of time and within a same treatment session, wherein n is an integer greater than 2.

15. A computation device comprising executable code which when executed causes the computation device to: receive, by a feature extraction module implemented on said computation device, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receive, by an electrical stimulation target designation module implemented on said computation device, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and output said first set of stimulation parameters for use by an electrical stimulation system.

16. The computation device according to claim 15, wherein said feature extraction module is further configured to process each of said plurality of electrical signals into a plurality of frequency bands.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.61561217. The computation device according to claim 16, wherein said plurality of frequency bands are delta (1-3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-30 Hz) and gamma (30-100 Hz) bands.

18. The computation device according to claim 17, wherein said first set of features comprises at least one of energy ratios of between at least one of said delta, theta, alpha, beta and gamma bands.

19. The computation device according to claim 17, wherein said first set of features comprises a functional connectivity feature type.

20. The computation device according to claim 19, wherein said functional connectivity feature type is at least one of a Weighted Phase Lag Index (WPLI) derived functional connectivity between said plurality of electrical signals from said EEG system, a synchronization likelihood or a relative wavelet entropy.

21. The computation device according to claim 17, wherein said first set of features comprises a graph metric.

22. The computation device according to claim 21, wherein said graph metric comprises at least one of a betweenness centrality and small world metric based on said plurality of electrical signals from said EEG system.

23. The computation device according to any one of claims 15-22, wherein said electrical stimulation target designation module comprises stimulation rules, and wherein said electrical stimulation target designation module is configured to determine said first set of stimulation parameters based on applying said stimulation rules to said first set of features.

24. The computation device according to claim 23, wherein said stimulation rules comprise at least one of an area with maximum delta rhythm, an area with maximum delta rhythmApplicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612 compared to an age-matched control group, an area with minimum gamma rhythm, an area with minimum gamma rhythm compared to an age-matched control group, an area with maximum delta over gamma rhythm, an area with maximum delta over gamma rhythm compared to an age- matched control group, an area with maximum brain rhythm energy difference compared to an age-matched control group, an area with maximum WPLI derived functional connectivity difference compared to an age-matched control group, or an area with maximum betweenness centrality difference compared to an age-matched control group.

25. The computation device according to any one of claims 15-24, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for a second period of time and extracts a second set of features therefrom, wherein said electrical stimulation target designation module receives said second set of features from said feature extraction module and determines a second set of stimulation parameters, and wherein said second period of time is subsequent to said first period of time and within a same treatment session.

26. The computation device according to claim 25, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for an nth period of time and extracts an nth set of features therefrom, wherein said electrical stimulation target designation module receives said nth set of features from said feature extraction module and determines an nth set of stimulation parameters, and wherein said nth period of time is subsequent to an (n-l)th period of time and within a same treatment session, wherein n is an integer greater than 2.

27. A computerized method for performing personalized transcranial electrical stimulation using electroencephalographic features, comprising:Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612 receiving, by a feature extraction module implemented on a computer, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receiving, by an electrical stimulation target designation module implemented on said computer, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and outputting said first set of stimulation parameters for use by an electrical stimulation system.

28. The method according to claim 27, wherein said feature extraction module is further configured to process each of said plurality of electrical signals into a plurality of frequency bands.

29. The method according to claim 28, wherein said plurality of frequency bands are delta (1- 3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-30 Hz) and gamma (30-100 Hz) bands.

30. The method according to claim 29, wherein said first set of features comprises at least one of energy ratios of between at least one of said delta, theta, alpha, beta and gamma bands.

31. The method according to claim 29, wherein said first set of features comprises a functional connectivity feature type.

32. The method according to claim 31, wherein said functional connectivity feature type is at least one of a Weighted Phase Lag Index (WPLI) derived functional connectivity between said plurality of electrical signals from said EEG system, a synchronization likelihood or a relative wavelet entropy.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.61561233. The method according to claim 29, wherein said first set of features comprises a graph metric.

34. The method according to claim 33, wherein said graph metric comprises at least one of a betweenness centrality and small world metric based on said plurality of electrical signals from said EEG system.

35. The method according to any one of claims 27-34, wherein said electrical stimulation target designation module comprises stimulation rules, and wherein said electrical stimulation target designation module is configured to determine said first set of stimulation parameters based on applying said stimulation rules to said first set of features.

36. The method according to claim 35, wherein said stimulation rules comprise at least one of an area with maximum delta rhythm, an area with maximum delta rhythm compared to an age-matched control group, an area with minimum gamma rhythm, an area with minimum gamma rhythm compared to an age-matched control group, an area with maximum delta over gamma rhythm, an area with maximum delta over gamma rhythm compared to an age-matched control group, an area with maximum brain rhythm energy difference compared to an age- matched control group, an area with maximum WPLI derived functional connectivity difference compared to an age-matched control group, or an area with maximum betweenness centrality difference compared to an age-matched control group.

37. The method according to any one of claims 27-36, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for a second period of time and extracts a second set of features therefrom, wherein said electrical stimulation target designation module receives said second set of features from said feature extraction module and determines a second set of stimulation parameters, and wherein said second period of time is subsequent to said first period of time and within a same treatment session.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.61561238. The method according to claim 37, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for an nth period of time and extracts an nth set of features therefrom, wherein said electrical stimulation target designation module receives said nth set of features from said feature extraction module and determines an nth set of stimulation parameters, and wherein said nth period of time is subsequent to an (n-l)th period of time and within a same treatment session, wherein n is an integer greater than 2.

39. A computer-readable medium comprising non-transient code which when executed by a computer causes the computer to: receive, by a feature extraction module implemented on said computer, a plurality of electrical signals from a corresponding plurality of sensor electrodes of an electroencephalography (EEG) system, wherein said feature extraction module processes said plurality of electrical signals for a first period of time and extracts a first set of features therefrom; receive, by an electrical stimulation target designation module implemented on said computer, said first set of features from said feature extraction module and determining a first set of stimulation parameters therefrom; and output said first set of stimulation parameters for use by an electrical stimulation system.

40. The computer-readable medium according to claim 39, wherein said feature extraction module is further configured to process each of said plurality of electrical signals into a plurality of frequency bands.

41. The computer-readable medium according to claim 40, wherein said plurality of frequency bands are delta (1-3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-30 Hz) and gamma (30-100 Hz) bands.Applicant Ref.: C18121_ P18121-02Attorney Docket 002240.61561242. The computer-readable medium according to claim 41, wherein said first set of features comprises at least one of energy ratios of between at least one of said delta, theta, alpha, beta and gamma bands.

43. The computer-readable medium according to claim 41, wherein said first set of features comprises a functional connectivity feature type.

44. The computer-readable medium according to claim 43, wherein said functional connectivity feature type is at least one of a Weighted Phase Lag Index (WPLI) derived functional connectivity between said plurality of electrical signals from said EEG system, a synchronization likelihood or a relative wavelet entropy.

45. The computer-readable medium according to claim 41, wherein said first set of features comprises a graph metric.

46. The computer-readable medium according to claim 45, wherein said graph metric comprises at least one of a betweenness centrality and small world metric based on said plurality of electrical signals from said EEG system.

47. The computer-readable medium according to any one of claims 39-46, wherein said electrical stimulation target designation module comprises stimulation rules, and wherein said electrical stimulation target designation module is configured to determine said first set of stimulation parameters based on applying said stimulation rules to said first set of features.

48. The computer-readable medium according to claim 47, wherein said stimulation rules comprise at least one of an area with maximum delta rhythm, an area with maximum delta rhythm compared to an age-matched control group, an area with minimum gamma rhythm, an area with minimum gamma rhythm compared to an age-matched control group, an area with maximum delta over gamma rhythm, an area with maximum delta over gamma rhythmApplicant Ref.: C18121_ P18121-02Attorney Docket 002240.615612 compared to an age-matched control group, an area with maximum brain rhythm energy difference compared to an age-matched control group, an area with maximum WPLI derived functional connectivity difference compared to an age-matched control group, or an area with maximum betweenness centrality difference compared to an age-matched control group.

49. The computer-readable medium according to any one of claims 39-48, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for a second period of time and extracts a second set of features therefrom, wherein said electrical stimulation target designation module receives said second set of features from said feature extraction module and determines a second set of stimulation parameters, and wherein said second period of time is subsequent to said first period of time and within a same treatment session.

50. The computer-readable medium according to claim 49, wherein said feature extraction module receives and processes said plurality of electrical signals from said EEG system for an nth period of time and extracts an nth set of features therefrom, wherein said electrical stimulation target designation module receives said nth set of features from said feature extraction module and determines an nth set of stimulation parameters, and wherein said nth period of time is subsequent to an (n-l)th period of time and within a same treatment session, wherein n is an integer greater than 2.

Images