阅读说明：本技术 用于脑活动解析的系统与方法 (System and method for brain activity resolution ) 是由 内森·英特拉托 于 2016-02-16 设计创作，主要内容包括：本发明提供了一种计算机实施的方法,包括：获得表示特定个体的脑活动的电信号数据；基于与特定脑状态相关联的预定预测因子处理电信号数据,预测因子选自包含多个预定预测因子的预测因子库,其中与特定脑状态相关联的预定预测因子包括：由预定的母小波产生的预定的代表性小波包原子集,小波包原子的预定排序以及预定的归一化因子集,其中该处理包括：利用预定的代表性小波包原子集将电信号数据解构成多个预定的解构小波包原子,其中电信号数据的时间窗口投射到预定的代表性小波包原子集上,其中每个预定的代表性小波包原子对应于特定的预定脑活动特征。(The invention provides a computer-implemented method, comprising: obtaining electrical signal data representative of brain activity of a particular individual; processing the electrical signal data based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein the predetermined predictors associated with the particular brain state include: a predetermined representative set of wavelet packet atoms generated from a predetermined mother wavelet, a predetermined ordering of wavelet packet atoms, and a predetermined set of normalization factors, wherein the process comprises: deconstructing the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined representative wavelet packet original subset, wherein a time window of the electrical signal data is projected onto the predetermined representative wavelet packet original subset, wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity characteristic.)

1. A computer-implemented method, comprising:

a. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

b. processing in real-time the electrical signal data representative of brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictor associated with a particular brain state comprises:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original set of wavelet packets generated from the predetermined mother wavelet,

a predetermined ordering of wavelet packet atoms, and

a predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing, by the specially programmed processor, the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms using the predetermined representative wavelet packet atomic set,

wherein the time window of electrical signal data is projected onto the predetermined representative original set of wavelet packets,

wherein the projection is via convolution or inner product, an

Wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing the plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering said stored plurality of predetermined deconstructed wavelet packet atoms within said computer data object based on utilizing a predetermined order by said specially programmed processor;

obtaining a statistical measure of activity for each of said reordered plurality of predetermined deconstructed wavelet packet atoms; and

v. normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

c. outputting a visual indication of at least one personalized mental state of the particular individual, a visual indication of at least one personalized neurological condition of the particular individual, or both based on the processing,

wherein an individual predetermined predictor associated with a particular brain state of the plurality of predetermined predictors is generated by:

i. obtaining the predetermined representative wavelet packet primitive subset by:

a. obtaining, by the specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

b. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, Coiflet Daubehies and Mayer wavelet families;

c. deconstructing, by the specially programmed processor, the at least one plurality of electrical signal data into a plurality of wavelet packet atoms using the selected mother wavelet;

d. storing the plurality of wavelet packet atoms in at least one computer data object;

e. determining an optimal original subset of wavelet packets using the predetermined mother wavelets and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determining is via utilizing an optimal basis algorithm; and

f. applying, by the specially programmed processor, wavelet de-noising to the plurality of wavelet packet atoms in the optimal set;

obtaining the predetermined ordering of wavelet packet atoms by:

a. projecting, by the specially programmed processor, the data representing brain activity for each time window of the at least one plurality of electrical signal data onto the predetermined representative subset of wavelet packet elements;

b. storing the projection in at least one computer data object;

c. determining, by the specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

d. storing the line length data in at least one computer data object; and

e. reordering, by the specially programmed computer, the stored projections to minimize statistical values of the line length values across each time window, across all individuals within the plurality of individuals, and across the projections; and

obtaining a predetermined set of normalization factors by:

a. determining, by the specially programmed computer, a mean and a standard deviation of the values of the stored projections.

2. The computer-implemented method of claim 1, wherein each time window is a four second time window.

3. The computer-implemented method of claim 1, wherein the reordered statistical value is selected from the group consisting of: the average of the sum of absolute differences of the wavelet packet atoms and the average of the (1-correlation) sum of the wavelet packet atoms.

4. The computer-implemented method of claim 1, wherein the electrical signal data representative of brain activity of a particular individual is recorded using two electrodes positioned on a forehead of the particular individual.

5. The computer-implemented method of claim 1, wherein the visual indication of at least one personalized mental state of the particular individual is used to identify an underlying mental state, an underlying neurological condition, or a combination of an underlying mental state and an underlying neurological condition in the particular individual, wherein the specially programmed computer utilizes at least one machine learning algorithm selected from the group consisting of: logistic regression modeling, support vector machine modeling, and deep learning modeling to assign at least one particular brain state to the visual indication of at least one personalized mental state for the particular individual, wherein the at least one particular brain state is associated with a mental state, a neurological condition, or a combination of a mental state and a neurological condition.

6. The computer-implemented method of claim 5, wherein assigning at least one particular brain state to the visual indication of at least one personalized mental state of the particular individual identifies an abnormality in at least one neural network in the particular individual's brain associated with a particular neurological condition.

7. The computer-implemented method of claim 6, wherein the abnormality in at least one neural network in the brain of the particular individual is used to diagnose the particular individual with a neurological condition.

8. The computer-implemented method of claim 6, wherein the neurological disorder is selected from the group consisting of: alzheimer's disease, dementia, stress, fatigue, anxiety, epilepsy, traumatic brain injury, loss of cognitive function, coma, lack of or inappropriate response to external stimuli associated with autism or autism spectrum disorders, inattention and sleep disorders.

9. The computer-implemented method of claim 5, wherein the at least one particular brain state is used to determine an emotional state of the particular individual.

10. The computer-implemented method of claim 1, wherein the particular individual is receiving treatment and the visual indication of at least one personalized mental state of the particular individual is used to determine the effectiveness of the treatment.

11. The computer-implemented method of claim 1, wherein the particular individual is receiving treatment and the visual indication of at least one personalized mental state of the particular individual is used to determine the nature of the treatment to be administered.

12. The computer-implemented method of claim 1, wherein the particular individual is receiving treatment and the visual indication of at least one personalized mental state of the particular individual is used to determine a duration of the treatment.

13. The computer-implemented method of claim 1, wherein the particular individual is receiving treatment and the visual indication of at least one personalized mental state of the particular individual is used to determine a dosing regimen for the treatment.

14. The computer-implemented method of claim 10, wherein the treatment is an anesthetic and the utility of the anesthetic is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of at least one personalized mental state of the particular individual.

15. The computer-implemented method of claim 10, wherein the treatment is a migraine treatment, and the utility of the migraine treatment is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of at least one personalized mental state of the particular individual.

16. The computer-implemented method of claim 1, wherein the particular individual is performing a particular cognitive task.

17. The computer-implemented method of claim 16, wherein the particular cognitive task is selected from the group consisting of: short-term and/or long-term memory recall, e-learning, meditation and concentration.

18. The computer-implemented method of claim 1, wherein the particular individual has a particular brain state at a particular time.

19. A specially programmed computer system comprising:

a. at least one special purpose computer comprising:

i. a non-transitory memory electronically storing specific computer executable program code; and

at least one computer processor that, when executing the specific program code, becomes a specially programmed computer processor configured to perform at least the following:

1. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

2. processing in real-time the electrical signal data representing brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictor associated with a particular brain state comprises:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original set of wavelet packets generated from the predetermined mother wavelet,

a predetermined ordering of wavelet packet atoms, and

a predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing, by the specially programmed processor, the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms using the predetermined representative wavelet packet atomic set,

wherein the time window of electrical signal data is projected onto the predetermined representative original set of wavelet packets,

wherein the projection is via convolution or inner product, an

Wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing the plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering said stored plurality of predetermined deconstructed wavelet packet atoms within said computer data object based on utilizing a predetermined order by said specially programmed processor;

obtaining a statistical measure of activity for each of said reordered plurality of predetermined deconstructed wavelet packet atoms; and

v. normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

3. outputting a visual indication of at least one personalized mental state of the particular individual, a visual indication of at least one personalized neurological condition of the particular individual, or both based on the processing,

wherein an individual predetermined predictor associated with a particular brain state within the plurality of predetermined predictors is generated by:

i. obtaining the predetermined representative wavelet packet primitive subset by:

1. obtaining, by the specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

2. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

3. deconstructing, by the specially programmed processor, the at least one plurality of electrical signal data into a plurality of wavelet packet atoms using the selected mother wavelet,

4. storing the plurality of wavelet packet atoms in at least one computer data object;

5. determining an optimal original subset of wavelet packets and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determining is via utilizing an optimal basis algorithm; and

6. applying, by the specially programmed processor, wavelet de-noising to a number of wavelet packet atoms in the optimal set;

obtaining the predetermined ordering of the wavelet packet atoms by:

1. projecting, by the specially programmed processor, the at least one plurality of electrical signal data representing brain activity of a time window of the data onto the predetermined representative subset of wavelet packet elements;

2. storing the projection in at least one computer data object;

3. determining, by the specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures for the projections of different channels with their neighboring channels;

4. storing the line length data in at least one computer data object; and

5. reordering, by the specially programmed computer, the stored projections to minimize statistics of line length values across each time window, across all individuals within the plurality of individuals, and across projections; and

obtaining the predetermined set of normalization factors by:

1. determining, by the specially programmed computer, a mean and a standard deviation of the stored projected values.

20. The system of claim 19, wherein each time window is a four second time window.

21. The system of claim 19, wherein the statistical value is selected from the group consisting of: the average of the sum of absolute differences of the wavelet packet atoms and the average of the (1-correlation) sum of the wavelet packet atoms.

22. The system of claim 19, wherein the electrical signal data representative of brain activity of a particular individual is recorded using two electrodes located on a forehead of the particular individual.

23. The system of claim 19, wherein the visual indication of at least one personalized mental state for the particular individual is used to identify an underlying mental state, an underlying neurological condition, or a combination of an underlying mental state and an underlying neurological condition in the particular individual, wherein the specially programmed computer utilizes at least one machine learning algorithm selected from the group consisting of: logistic regression modeling, support vector machine modeling, and deep learning modeling to assign at least one particular brain state to the visual indication of at least one personalized mental state for the particular individual, wherein the at least one particular brain state is associated with a mental state, a neurological condition, or a combination of a mental state and a neurological condition.

24. The system of claim 23, wherein assigning at least one particular brain state to the visual indication of at least one personalized mental state of the particular individual identifies an abnormality in at least one neural network in the particular individual's brain associated with a particular neurological condition.

25. The system of claim 24, wherein the abnormality in at least one neural network in the brain of the particular individual is used to diagnose the particular individual with a neurological condition.

Technical Field

Background

Electroencephalography (EEG) is a method of monitoring brain electrical activity. It is generally non-invasive, with the electrodes placed along the scalp, however, invasive electrodes may be used in certain applications. EEG measures voltage fluctuations caused by ionic currents within neurons of the brain. However, the sensitivity of the EEG electrodes limits detection to a small area of the brain near each electrode, thereby limiting the spatial resolution of the EEG.

Functional magnetic resonance imaging (fMRI) is another method of monitoring brain activity. However, a magnetic resonance imager is a large and expensive clinical device that can be used neither outside a clinic nor in a continuous manner.

Disclosure of Invention

In one embodiment, the invention provides a computer-implemented method comprising:

a. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

b. processing in real time electrical signal data representative of brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictors associated with a particular brain state include:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original set of wavelet packets generated from a predetermined mother wavelet,

a predetermined ordering of wavelet packet atoms, and

a predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms by a specially programmed processor using a predetermined representative original subset of wavelet packets,

wherein a time window of electrical signal data is projected onto a predetermined representative original subset of wavelet packets

Wherein the projection is via convolution or inner product, an

Wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing a plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering the stored plurality of predetermined deconstructed wavelet packet atoms within the computer data object based on utilizing the predetermined order by a specially programmed processor;

obtaining a statistical measure of activity for each of the reordered plurality of predetermined deconstructed wavelet packet atoms; and

normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

c. outputting, based on the processing, a visual indication of at least one personalized mental state for the particular individual, a visual indication of at least one personalized neurological condition for the particular individual, or both,

wherein an individual predetermined predictor associated with a particular brain state of a plurality of predetermined predictors is generated by:

i. obtaining a predetermined representative primitive subset of wavelet packets by:

a. obtaining, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

b. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

c. deconstructing the at least one plurality of electrical signal data into a plurality of wavelet packet atoms by a specially programmed processor using the selected mother wavelet,

d. storing a plurality of wavelet packet atoms in at least one computer data object;

e. determining an optimal original subset of wavelet packets using a predetermined mother wavelet and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determination is via utilizing an analytical optimal basis algorithm; and

f. applying wavelet de-noising, by a specially programmed processor, to a number of wavelet packet atoms in the optimal set;

obtaining a predetermined ordering of wavelet packet atoms by:

a. projecting, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity every 4 second data window onto a predetermined representative primitive subset of wavelet packets;

b. storing the projections in at least one computer data object;

c. determining, by a specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

d. storing the line length data in at least one computer data object; and

e. reordering, by a specially programmed computer, the stored projections to minimize statistics of line length values across each time window, across all individuals within a plurality of individuals, and across the projections; and

obtaining a predetermined set of normalization factors by:

a. the mean and standard deviation of the stored projected values are determined by a specially programmed computer.

In one embodiment, the invention provides a specially programmed computer system comprising:

a. at least one special purpose computer comprising:

i. a non-transitory memory electronically storing specific computer executable program code; and

at least one computer processor that, when executing specific program code, becomes a specially programmed computer processor configured to perform at least the following:

1. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

2. processing in real time electrical signal data representative of brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictors associated with a particular brain state include:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original subset of wavelet packets,

a predetermined ordering of wavelet packet atoms generated from a predetermined mother wavelet, an

A predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms by a specially programmed processor using a predetermined representative original subset of wavelet packets,

wherein a time window of electrical signal data is projected onto a predetermined representative original subset of wavelet packets

Wherein the projection is via convolution or inner product, and wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing a plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering the stored plurality of predetermined deconstructed wavelet packet atoms within the computer data object based on utilizing the predetermined order by a specially programmed processor;

obtaining a statistical measure of activity for each of the reordered plurality of predetermined deconstructed wavelet packet atoms; and

normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

3. outputting, based on the processing, a visual indication of at least one personalized mental state for the particular individual, a visual indication of at least one personalized neurological condition for the particular individual, or both,

wherein an individual predetermined predictor associated with a particular brain state of a plurality of predetermined predictors is generated by:

i. obtaining a predetermined representative primitive subset of wavelet packets by:

1. obtaining, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

2. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

3. deconstructing the at least one plurality of electrical signal data into a plurality of wavelet packet atoms by a specially programmed processor using the selected mother wavelet,

4. storing a plurality of wavelet packet atoms in at least one computer data object;

5. determining an optimal original subset of wavelet packets using a predetermined mother wavelet and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determination is via utilization of an optimal basis algorithm; and

6. applying wavelet de-noising, by a specially programmed processor, to a number of wavelet packet atoms in the optimal set;

obtaining a predetermined ordering of wavelet packet atoms by:

1. projecting, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity every 4 second data window onto a predetermined representative primitive subset of wavelet packets;

2. storing the projections in at least one computer data object;

3. determining, by a specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

4. storing the line length data in at least one computer data object; and

5. reordering, by a specially programmed computer, the stored projections to minimize statistics of line length values across each time window, across all individuals within a plurality of individuals, and across the projections; and

obtaining a predetermined set of normalization factors by:

1. the mean and standard deviation of the stored projected values are determined by a specially programmed computer.

In one embodiment, two electrodes located on the forehead of a particular individual are used to record electrical signal data representing brain activity of the particular individual.

In one embodiment, each time window is a four second time window.

In one embodiment, the reordered statistical values are selected from the group consisting of: the average of the sum of absolute differences of wavelet packet atoms and the average of the sum of wavelet packet atoms (1-correlation).

In one embodiment, the visual indication of at least one personalized mental state of the particular individual is used to identify an underlying mental state, an underlying neurological disorder, or a combination of an underlying mental state and an underlying neurological disorder in the particular individual, wherein the specially programmed computer utilizes at least one machine learning algorithm, including but not limited to logistic regression modeling, support vector machine modeling, and deep learning modeling, to assign at least one particular brain state to the visual indication of at least one personalized mental state of the particular individual, wherein the at least one particular brain state is associated with the mental state, neurological disorder, or a combination of the mental state and neurological disorder.

In one embodiment, assigning at least one particular brain state to the visual indication of at least one personalized mental state of the particular individual identifies an abnormality in at least one neural network in the brain of the particular individual associated with the particular neurological condition.

In one embodiment, an abnormality in at least one neural network in the brain of a particular individual is used to diagnose the particular individual with a neurological condition.

In one embodiment, the neurological disorder is selected from the group consisting of: alzheimer's disease, dementia, stress, fatigue, anxiety, epilepsy, traumatic brain injury, loss of cognitive function, migraine, chronic pain, coma, lack of or inappropriate response to external stimuli associated with autism or autism spectrum disorders, inattention and sleep disorders.

In one embodiment, at least one particular brain state is used to determine an emotional state of a particular individual.

In one embodiment, a particular individual is receiving treatment and the effectiveness of the treatment is determined using at least one visual indication of the personalized mental state of the particular individual.

In one embodiment, a particular individual is receiving treatment and the nature of the treatment to be administered is determined using at least one visual indication of the personalized mental state of the particular individual.

In one embodiment, a particular individual is receiving treatment and the duration of treatment is determined using a visual indication of at least one personalized mental state of the particular individual.

In one embodiment, a particular individual is receiving treatment and a dosage regimen for the treatment is determined using at least one visual indication of the personalized mental state of the particular individual.

In one embodiment, the treatment is an anesthetic and the utility of the anesthetic is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of the at least one personalized mental state of the particular individual.

In one embodiment, the treatment is migraine treatment, and the utility of the migraine treatment is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of at least one personalized mental state of the particular individual.

In one embodiment, a particular individual is performing a particular cognitive task.

In one embodiment, the specific cognitive task is selected from the group consisting of: short-term and/or long-term memory recall, e-learning, meditation and concentration.

In one embodiment, a particular individual has a particular brain state at a particular time.

Drawings

Some embodiments of the invention are described herein, by way of example only, with reference to the accompanying drawings. Referring now in detail to the drawings in detail, it should be emphasized that the details shown are by way of example and for purposes of illustrative discussion of embodiments of the invention. In this regard, the description taken with the drawings make apparent to those skilled in the art how the embodiments of the invention may be embodied.

FIG. 1 illustrates a screen shot of a recording electrode according to some embodiments of the invention.

Fig. 2 illustrates a screenshot of an example of a recorded representation of a subject's brain activity characteristics, according to some embodiments of the invention.

Fig. 3A and 3B illustrate screenshots of examples of recorded representations of brain activity characteristics of a subject, according to some embodiments of the invention.

Fig. 4 illustrates a screenshot of an example of a Brain Activity Feature (BAF) representation of a subject according to some embodiments of the invention.

Fig. 5 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 6 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 7 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 8 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 9 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 10 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

Fig. 11 illustrates a screenshot of an example of a recorded representation of a subject's brain activity characteristics, according to some embodiments of the invention.

Fig. 12 illustrates a screenshot of an example of a plurality of recorded representations of brain activity characteristics of a subject according to some embodiments of the invention.

Fig. 13 illustrates a screenshot of an example of a BAF representation of a subject according to some embodiments of the invention.

FIG. 14 illustrates a screenshot of an example of a specially programmed computer architecture according to some embodiments of the invention.

FIG. 15 illustrates a screenshot of an example of a specially programmed computer architecture according to some embodiments of the invention.

FIG. 16 illustrates a screenshot of an example of a specially programmed computer architecture according to some embodiments of the invention.

Detailed Description

Among the benefits and improvements that have been disclosed, other objects and advantages of this invention will become apparent from the following description taken in conjunction with the accompanying drawings. Detailed embodiments of the present invention are disclosed herein; however, it is to be understood that the disclosed embodiments are merely illustrative of the invention that may be embodied in various forms. In addition, each of the examples given in connection with the various embodiments of the invention are intended to be illustrative, and not restrictive.

Throughout the specification and claims, the following terms have the meanings explicitly associated herein, unless the context clearly dictates otherwise. The phrases "in one embodiment" and "in some embodiments" as used herein do not necessarily refer to the same embodiment, but they may. Moreover, the phrases "in another embodiment" and "in some other embodiments" as used herein do not necessarily refer to a different embodiment, but they may. Thus, as described below, various embodiments of the invention may be readily combined without departing from the scope or spirit of the invention.

Further, as used herein, the term "or" is an inclusive "or" operator, and is equivalent to the term "and/or," unless the context clearly dictates otherwise. The term "based on" is not exclusive and allows for being based on other factors not described, unless the context clearly dictates otherwise. Further, throughout the specification, the meaning of "a", "an" and "the" includes plural references. The meaning of "in … …" includes "in … …" and "on … …".

It should be appreciated that at least one aspect/function of the various embodiments described herein may be performed in real-time and/or dynamically. As used herein, the term "real-time" refers to an event/action that may occur instantaneously or nearly instantaneously at the time another event/action has occurred. In some embodiments, the terms "instant," "instantaneously," "instant," and "real-time" refer to a situation in which the time difference between a first time at which a search request is sent and a second time in response to a received request does not exceed 1 second. In some embodiments, the time difference between the request and the response is between less than 1 second and a few seconds.

As used herein, the term "dynamically" means that events and/or actions can be triggered and/or occur without any human intervention. In some embodiments, events and/or actions according to the present invention may be in real time and/or based on a predetermined period of at least one of: nanoseconds, milliseconds, seconds, minutes, hours, days, weeks, months, etc.

The decomposition of EEG signals into different components is an effective tool for studying brain activity and brain state, and in a given brain state, the role of certain functional areas of the brain or neural networks in the brain is inferred. Without being bound by any particular theory, a particular brain state is associated with a particular mental state, a particular neurological condition, or a particular combination of mental state and neurological condition.

Without being bound by any particular theory, brain activity detected by conventional EEG is associated with a number of frequency bands from about 0.5Hz (delta waves) to gamma waves above 32Hz, between which are theta waves, α waves, β waves, etc.

In some embodiments of the invention, the role of certain regions or neural networks within the brain need not be determined for a given cognitive function or mental state. In some embodiments, electrical brain activity of the subject is recorded using two electrodes (e.g., Fp1 and Fp2) located on the forehead of the subject. In some embodiments, the Fp1 or Fp2 electrode is used as the reference electrode, and the electrical activity recorded is the difference between the Fp1 electrode and the Fp2 electrode. Alternatively, in some embodiments, the Fp2 electrode may be used as a reference electrode or a recording electrode.

In some embodiments, the invention provides a computer-implemented method comprising:

a. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

b. processing in real time electrical signal data representative of brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictors associated with a particular brain state include:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original set of wavelet packets generated from a predetermined mother wavelet,

a predetermined ordering of wavelet packet atoms, and

a predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms by a specially programmed processor using a predetermined representative original subset of wavelet packets,

wherein a time window of electrical signal data is projected onto a predetermined representative original subset of wavelet packets

Wherein the projection is via convolution or inner product, an

Wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing a plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering the stored plurality of predetermined deconstructed wavelet packet atoms within the computer data object, optionally based on utilizing a predetermined order by a specially programmed processor;

obtaining a statistical measure of activity for each of the reordered plurality of predetermined deconstructed wavelet packet atoms; and

normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

a. outputting a visual indication of at least one personalized mental state of the particular individual, a visual indication of at least one personalized neurological condition of the particular individual, or both based on the processing,

wherein an individual predetermined predictor associated with a particular brain state of a plurality of predetermined predictors is generated by:

i. obtaining a predetermined representative primitive subset of wavelet packets by:

a. obtaining, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

b. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

c. deconstructing the at least one plurality of electrical signal data into a plurality of wavelet packet atoms by a specially programmed processor using the selected mother wavelet,

d. storing the plurality of wavelet packet atoms in at least one computer data object;

e. determining an optimal original subset of wavelet packets using a predetermined mother wavelet and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determination is via utilization of an optimal basis algorithm; and

f. applying wavelet de-noising, by a specially programmed processor, to a number of wavelet packet atoms in the optimal set;

obtaining a predetermined ordering of wavelet packet atoms by:

a. projecting, by a specially programmed processor, the at least one plurality of electrical signal data representing brain activity every 4 second window of data onto a predetermined representative primitive subset of wavelet packets;

b. storing the projections in at least one computer data object;

c. determining, by a specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

d. storing the line length data in at least one computer data object; and

e. optionally reordering the stored projections by a specially programmed computer to minimize statistics of line length values across all individuals within each time window, across multiple individuals, and across projections; and

obtaining a predetermined set of normalization factors by:

a. the mean and standard deviation of the stored projected values are determined by a specially programmed computer.

In some embodiments, electrical signal data representing brain activity of a particular individual is recorded using two electrodes located on the forehead of the particular individual.

An example of a recording electrode according to some embodiments of the present invention is shown in FIG. 1. In some embodiments, between one and three electrodes located on the forehead of the subject are used to record electrical brain activity of the subject. In some embodiments, between one and three electrodes located in the head region that are not on or in addition to the forehead of the subject are used to record electrical brain activity of the subject. For example, in some embodiments, at least one electrode is located on or at or in the back of the subject's ear. For example, in some embodiments, the at least one electrode is part of an article positioned on the head of the subject, wherein the article is configured for at least one additional function in addition to carrying the at least one electrode. For example, in some embodiments, the article is headwear (e.g., a hat). For example, in some embodiments, the article is an electronic device (e.g., a headset).

In some embodiments, the brain electrical activity of the subject is recorded according to the method disclosed in the genomics frontier (Frontiers genetics), volume 5, pages 1 to 12 (2014) of g.castellani et al.

In some embodiments, electrical signal data representing brain activity of a particular individual is recorded at a sufficiently large sampling rate above 250 and a dynamic range configured to detect sufficient cortical activity at a desired location of the brain. For example, it is contemplated that a greater dynamic range may detect more cortical activity than a smaller dynamic range. In one embodiment, the dynamic range is a modulus (A/D) resolution of 15 bits and more.

In some embodiments, a particular individual is performing a particular cognitive task.

In some embodiments, the specific cognitive task is selected from the group consisting of: short-term and/or long-term memory recall, e-learning, meditation and concentration.

In one embodiment, a particular individual has a particular brain state at a particular time.

Methods according to some embodiments of the invention process a recorded electrical signal number representing brain activity of a particular individual According to

Deconstructing recorded electrical signal data representing brain activity of a particular individual: in some embodiments, recorded electrical signal data representing brain activity of a particular individual is recorded in real time over a particular time period. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 1 hour. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 50 minutes. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 40 minutes. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 30 minutes. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 20 minutes. In some embodiments, electrical signal data representing brain activity of a particular individual is recorded for up to 10 minutes.

In some embodiments, real-time recorded electrical signal data representing brain activity of a particular individual is deconstructed into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined representative original subset of wavelet packets. Each individual predetermined deconstructed wavelet packet atom within the plurality of predetermined deconstructed wavelet packet atoms corresponds to a brain activity feature ("BAF").

In some embodiments, the predetermined representative original subset of wavelet packets is generated from a predetermined mother wavelet selected from a family of wavelets including, but not limited to: haar, coifletdubehies and Mayer wavelet families. Html at website http:// www.mathworks.com/help/wavelet/ref/wavelet families? Other families of wavelets suitable for the mother wavelet are described in refresh true according to some embodiments of the present invention.

In some embodiments, recorded electrical signal data representing brain activity of a particular individual is deconstructed into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined representative wavelet packet primitive subset according to the optimal basis algorithm disclosed in Coifman, r.r., & Wickerhauser, m.v., IEEE Transactions on Information Theory, 38(2), 713-.

In some embodiments, electrical signal data representing a record of brain activity of a particular individual is deconstructed into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined set of representative wavelet packet atoms according to shannon entropy and a combination of Stainvas, I and integrator, n.

In some embodiments, recorded electrical signal data representing brain activity of a particular individual is deconstructed into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined set of representative wavelet packet atoms according to shannon entropy and a combination of another suitable optimal basis algorithm disclosed in Intrater, N, Neural Computation 5, 443-455(1993), the specific disclosure of which is incorporated herein by reference.

In some embodiments, recorded electrical signal data representing brain activity of a particular individual is deconstructed into a plurality of predetermined deconstructed wavelet packet atoms using a predetermined set of representative wavelet packet atoms according to shannon entropy and a combination of another suitable optimal basis algorithm disclosed in intra, N, Neural Computation 4, 98-1-7(1992), the specific disclosure of which is incorporated herein by reference.

Reordering a plurality of predetermined deconstructed wavelet packet atoms: in some embodiments, the plurality of predetermined wavelet packet atoms are reordered according to a predetermined order. In some embodiments, the reordered plurality of predetermined wavelet packet atoms are normalized with a predetermined normalization factor. In some embodiments, the plurality of predetermined wavelet packet atoms are only normalized.

In some embodiments, an exemplary specially programmed processor of the present invention is programmed with electrical data recorded by standard EEG recording electrodes to first determine a distribution of representative values for each wavelet packet atom output by at least one band pass filter, respectively, based on a data set of data collected from a plurality of individuals, respectively, by normalizing each wavelet packet atom. In some embodiments, at least one band pass filter has 2 to 36 channels. In some embodiments, at least one band pass filter has at least 12 channels. In some embodiments, at least one band pass filter has at least 16 channels. In some embodiments, at least one band pass filter has at least 32 channels.

Visual indication of at least one personalized mental state of a specific individual: in some embodiments, the statistical measures of the plurality of normalized, reordered projections on the predetermined wavelet packet atoms are combined into a visual representation, where each individual normalized predetermined wavelet packet atom of the plurality corresponds to a BAF and is arranged in the representation according to a predetermined order. As used herein, "BAF representation" refers to a visual representation of a plurality of normalized, reordered predetermined projections on a wavelet packet atom. An example of a BAF representation of a subject according to some embodiments of the invention is shown in figure 2.

In some embodiments, a BAF representation of a particular individual has 121 individual BAFs. Alternatively, in some embodiments, BAF representations for a particular individual have up to 200 individual BAFs. Alternatively, in some embodiments, a BAF representation of a particular individual has from 10 to 200 individual BAFs. Alternatively, in some embodiments, a BAF representation for a particular individual has from 1 to 1000 individual BAFs. Alternatively, in some embodiments, a BAF representation for a particular individual has 30 to 1000 individual BAFs. Alternatively, in some embodiments, a BAF representation of a particular individual has at least 30 individual BAFs. Alternatively, in some embodiments, a BAF representation of a particular individual has an individual BAF number that is a multiple (e.g., 2 times, 3 times, 4 times, 5 times, 6 times, etc.) of the recorded BAF number.

In some embodiments, the BAF representation of the subject has 121 individual BAFs. Alternatively, in some embodiments, a BAF of a subject represents having more than 200 individual BAFs. Alternatively, in some embodiments, a BAF of a subject is represented as having 10 to 200 individual BAFs. Alternatively, in some embodiments, a BAF representation of a subject has 1 to 1000 individual BAFs. Alternatively, in some embodiments, a BAF representation of a subject has 30 to 1000 individual BAFs. Alternatively, in some embodiments, BAF of the subject means having at least 30 individual BAFs. Alternatively, in some embodiments, a subject's BAF indicates having a number of individual BAFs that is a multiple (e.g., 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, etc.) of the number of neural networks analyzed. In some embodiments, the BAF comprises a conventional EEG recording.

Referring to fig. 2 as an example, each line perpendicular to the y-axis represents projected activity on a single predetermined wavelet packet atom, (also referred to herein as a BAF). For example, activity may be represented by at least one suitable statistical measure of the projection of a single wavelet packet atom or a group of wavelet packet atoms, where suitable statistical measures may be, but are not limited to, averages, standard deviations, and the like. In some embodiments, the BAF representation may be color-coded. For example, as shown in fig. 2, various active regions on the intensity spectrum may be presented, such as, but not limited to, by presenting high activity regions as darker shaded regions of at least one particular color ("hot") to lower activity-prone regions as lighter shaded regions of at least one color or at least one other color ("cold") based on the respective activity level, and any continuous shading therebetween. Each column perpendicular to the x-axis represents a vector of brain activity state (BAF representation) at a particular time or a particular time period. Thus, the x-axis is measured in time (e.g., milliseconds, seconds, minutes, hours, days, etc.). In some embodiments, the image is normalized by a suitable non-linear transformation (e.g., histogram equalization) prior to color coding of each Brain Activity (BAF) of the plurality of BAFs.

In some embodiments, an exemplary specially programmed processor of the present invention is programmed to cluster electrical signal data representing brain activity of a particular individual prior to determining the predetermined predictor. For example, an exemplary specially programmed processor of the present invention is programmed to generate a set of m-dimensional vectors from projections on m predetermined deconstructed wavelet packet atoms, which may be further clustered into different brain states. In some embodiments, an exemplary specially programmed processor of the present invention is programmed to determine the number of brain states by using at least one machine learning technique. For example, an exemplary specially programmed processor of the present invention is programmed to analyze cluster data using hierarchical clustering and determine which clusters are grouped together based on relative distances between their members.

In some embodiments, an exemplary specially programmed processor of the present invention is programmed to construct a plurality of predetermined predictors using cluster membership based at least in part on:

1) distances to the center of the cluster or to different members of the cluster, and/or

2) A sequence of cluster membership preceding the current frame.

For example, an exemplary specially programmed processor of the present invention is programmed to determine a particular predictor of a predictor library using at least one temporal model (e.g., without limitation, Markov chain, hidden Markov model, other similar suitable models) based on cluster members.

In some embodiments, the exemplary specially programmed processor of the present invention is programmed to generate at least one temporal structure probability model after assigning a cluster membership to each windowed frame. For example, in text analysis, starting with data, the specially programmed processor of the present invention is programmed to: constructing a vocabulary of letters (specific clusters); words are recognized based on letter segmentation, word vocabularies are built from the recognized words, and sentences are created from the words by interpreting specific grammar rules. For example, the first step is to construct a probability matrix that moves from one letter to another.

Methods according to some embodiments of the invention identify potential mental states, potential neurological disorders, or potential neurological conditions In a combination of mental states and neurological disorders

In some embodiments, the visual indication of the at least one personalized mental state of the particular individual is used to identify the potential mental state, the potential neurological condition, or the combination of the potential mental state and the potential neurological condition in the particular individual, wherein the specially programmed computer utilizes at least one machine learning algorithm selected from the group consisting of logistic regression modeling, support vector machine modeling, and deep learning modeling to assign the at least one particular brain state to the personalized mental state of the at least one particular individual, wherein the at least one particular brain state is associated with the mental state, the neurological condition, or the combination of the mental state and the neurological condition.

In some embodiments, the exemplary specially programmed processor of the present invention is programmed to identify potential mental states, potential neurological disorders, or a combination of potential mental states and potential neurological disorders in a particular individual using at least one machine learning algorithm, such as, but not limited to, logistic regression modeling, support vector machine modeling, and deep learning modeling. Specifically, in some embodiments, the specially programmed processors of the present invention are programmed to perform at least the following steps:

1) separating electrical signal data representative of brain activity of a particular individual into training, validation and test data sets;

2) generating a group of models based on the training set, and optimizing the verification set;

3) testing the performance of each model against the test set;

4) repeating steps 1 through 3 for different parameters of a particular AI model (e.g., regularization parameters in a ridge regression model; the number of hidden units in the feedforward neural network; weight attenuation parameters in a feed-forward neural network; kernel models such as kernel type and number in support vector machines; a combination of gaussian and regularization parameters in a support vector machine; a combination of gaussian models, etc.); and

5) after the set of model parameters has been determined, a prediction is obtained for the new data set, and steps 1 to 4 are repeated for different orthogonal decomposition families and other model parameters obtained from recorded electrical signal data representing brain activity of a particular individual.

In some embodiments, electrical signal data representative of brain activity of a particular individual is recorded when the particular individual has a particular mental state. In some embodiments, the particular mental state is unknown, and methods according to some embodiments of the invention are used to identify the particular mental state.

Examples of particular mental states include, but are not limited to, seizures, fear, anxiety, pain, sleep states (e.g., Rapid Eye Movement (REM) sleep), arousal, alertness, fatigue, paralysis, meditation, stress, other emotions, different brain states associated with dementia, lack of or inappropriate response to external stimuli associated with autism or autism spectrum disorders, and the like. An example of a BAF representation of a subject with a particular mental state is shown in fig. 3.

Referring to fig. 3A, a brain activity representation of an experienced yoga teacher who is performing three types of meditation (indicated via horizontal brackets) is shown. After the first two meditations and the third meditation, the yoga teacher explains what he is doing. Ranges in BAF are indicated by vertical brackets. This set of BAFs is associated with awareness. It is clear that a meditation only emphasizes these awareness channels, which is apparent in the event that the yoga master explains what meditation he is doing.

Referring to fig. 3B, a BAF representation of an individual in a vegetative human state is shown. BAF bright spots are shown in fig. 3A and are inactive. However, management of medical brain stimulation of an individual can activate cognitively associated BAF in a short time. This provides an example of the ability to determine the utility of a drug using BAF representation.

In some embodiments, electrical signal data representative of brain activity of a particular individual is recorded while the particular individual performs a particular cognitive task. In some embodiments, methods according to some embodiments of the invention identify an underlying mental state, an underlying neurological condition, or a combination of an underlying mental state and an underlying neurological condition based, at least in part, on electrical signal data representative of brain activity of a particular individual recorded while the particular individual performs a particular cognitive task.

Examples of specific cognitive tasks include, but are not limited to: short-term and long-term memory recall, stimulus recognition, meditation, learning, movie watching, image observation, strong concentration during exercise, response to sensory stimuli, and the like. An example of a BAF representation of a subject performing a particular cognitive task is shown in figure [6 ].

In some embodiments, the sensory stimulus may be auditory, tactile, olfactory, visual, and the like.

In some embodiments, assigning at least one particular brain state to the visual indication of at least one personalized mental state of the particular individual identifies an abnormality in at least one neural network in the brain of the particular individual associated with the particular neurological condition.

In some embodiments, an abnormality in at least one neural network in the brain of a particular individual is used to diagnose the particular individual with a neurological condition.

In some embodiments, the neurological disorder is selected from the group consisting of: alzheimer's disease, dementia, stress, fatigue, anxiety, epilepsy, traumatic brain injury, loss of cognitive function, coma, lack of or inappropriate response to external stimuli associated with autism or autism spectrum disorders, inattention and sleep disorders. An example of a BAF representation of a subject with a neurological disorder is shown in fig. 4.

In some embodiments, the neurological disorder of a particular individual is unknown, and the method according to some embodiments of the invention identifies the neurological disorder.

In some embodiments, at least one particular brain state is used to determine an emotional state of a particular individual.

In some embodiments, a particular individual is receiving treatment and the effectiveness of the treatment is determined using at least one visual indication of the personalized mental state of the particular individual.

For example, by way of illustration, early intervention in patients with autism at about 6 months of age may improve treatment of autism. In another example, early detection of abnormal activity in at least one neuronal network associated with epilepsy in the brain of an individual may improve treatment of epilepsy or alert the individual or a caregiver of the individual to the occurrence or imminent occurrence of a seizure. In another example, early detection of abnormal activity in at least one neuronal network associated with a migraine in the brain of an individual may improve treatment of the migraine, or alert the individual or a caregiver of the individual to the occurrence or future occurrence of the migraine. In another example, early detection of abnormal activity in at least one neuronal network associated with an ischemic event in the brain of an individual may improve treatment of ischemic injury, or alert the individual or a caregiver of the individual that an ischemic event occurred, will occur, or has occurred, such as a transient ischemic event or stroke. In some embodiments, brain activity of a particular individual may be recorded while the subject is asleep, in the case of certain ischemic conditions, when these conditions are more likely to occur.

In some embodiments, a particular individual is receiving treatment and the nature of the treatment to be administered is determined using at least one visual indication of the personalized mental state of the particular individual.

In some embodiments, the particular individual is receiving treatment and the duration of treatment is determined using the visual indication of the at least one personalized mental state of the particular individual.

In some embodiments, a particular individual is receiving treatment and the dosage regimen for treatment is determined using at least one visual indication of the personalized mental state of the particular individual.

In some embodiments, the treatment is an anesthetic, and the utility of the anesthetic is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of the at least one personalized mental state of the particular individual.

In some embodiments, the treatment is a migraine treatment, and the utility of the migraine treatment is determined by the ability of the particular individual to feel pain and/or the level of pain perceived by the individual and the correlation to changes in the visual indication of the at least one personalized mental state of the particular individual.

In some embodiments, the migraine treatment is a drug. Alternatively, in some embodiments, migraine treatment is directed. Alternatively, in some embodiments, the migraine treatment is hypnosis. Alternatively, in some embodiments, the migraine treatment is meditation.

In some embodiments, in contrast to analysis of fetal brain activity using amplitude-integrated EEG, which typically detects both brain conditions of sleep and awake states, and the dynamics of transitions from one state to another to identify the extent of brain injury, an exemplary specially programmed processor of the present invention is programmed to perform a cohort analysis of the brain states of a group of infants and determine the brain states of the infants at a time.

In another example, in neuromedings, in some embodiments, an exemplary specially programmed processor of the present invention is programmed to perform a cohort analysis of a set of brain states in an individual receiving stimulation at each time frame, and to determine the proportion of cohorts that are in the same brain state at a given window frame. This enables the measurement of cohort engagement with stimulation, since when a larger part of the cohort is found in the same brain state, it is considered to be engaged and reacting to stimulation, as this may happen with stimulation. The particular brain state in which a portion of the cohort is located may correspond to a particular response to a stimulus, thereby allowing feedback training.

In some embodiments, the methods of the present invention determine the mental state of a particular individual at a first point in time. In some embodiments, the brain of a particular individual transitions from one mental state to another, but remains in the first mental state most of the time. In some embodiments, the first mental state is adverse. In some embodiments, the system may provide stimulation via neurofeedback that stimulates the brain of a particular individual to enter a second, more favorable mental state. For example, by way of illustration, a particular individual may be in a coma and the first mental state may be an unresponsive state. The system may provide a stimulus that stimulates the brain of a particular individual to enter a second, more responsive mental state.

In another example, the patient may have an autism spectrum disorder, and the first mental state may be that the individual refuses or fails to maintain eye contact with another person. The system may provide a stimulus that stimulates the brain of a particular individual to enter a second mental state, where the individual is able to more easily maintain eye contact.

A plurality of predetermined predictors

In some embodiments, the individual predetermined predictors associated with a particular brain state within the plurality of predetermined predictors are generated by:

i. obtaining a predetermined representative primitive subset of wavelet packets by:

a. obtaining, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

b. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

c. deconstructing, by a specially programmed processor, the at least one plurality of electrical signal data into a plurality of wavelet packet atoms;

d. storing a plurality of wavelet packet atoms in at least one computer data object;

e. determining an optimal original subset of wavelet packets and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determination is via utilization of an optimal basis algorithm; and

f. applying wavelet de-noising, by a specially programmed processor, to a number of wavelet packet atoms in the optimal set;

obtaining a predetermined ordering of wavelet packet atoms by:

a. projecting, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity every 4 second data window onto a predetermined representative primitive subset of wavelet packets;

b. storing the projections in at least one computer data object;

c. determining, by a specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

d. storing the line length data in at least one computer data object; and

e. optionally reordering the stored projections by a specially programmed computer to minimize statistics of line length values across all individuals within each time window, across multiple individuals, and across projections; and;

obtaining a predetermined set of normalization factors by:

a. the mean and standard deviation of the stored projected values are determined by a specially programmed computer.

To generate a plurality of predetermined predictor libraries requires an illustrative library of at least one plurality of electrical signal data representing brain activity for a particular brain state. The generation of an illustrative library of at least one plurality of electrical signal data representing brain activity for a particular brain state requires obtaining a sufficient set of electrical signal data (e.g., 100 records; 1,000 records; 10,000 records; 100,000 records; 1,000,000 records, etc.) representing brain activity for the particular brain state. Typically, a 3000 second recording of each event is sufficient to detect the brain state event robustly. The more observation times, the more robust the detection. Electrical signal data representing brain activity for a particular brain state may be from a sufficient number of individuals (e.g., 100; 1,000; 10,000; 100,000; 1,000,000, etc.) and may be recorded during various (e.g., different types, different intensities, etc.) activities, cognitive tasks, and neurological disorders, resulting in various brain states.

In some embodiments, multiple predetermined pools of predictors may be customized for one or more particular objectives. For example, if emphasis is needed to detect specific brain activity events, e.g., the detection of abnormalities present in a certain cortical location that occur before a seizure occurs or at an early stage of migraine headache, then emphasis should be placed on recording at these times. Such emphasis is given by recording from a single subject at the time of such an event, or from multiple subjects at those times. Another example may be a record of a subject who is performing an attention test, such as the t.o.v.a. test (TOVA, los alamitos, CA). Then, a record is obtained of a plurality of subjects performing the same task, rather than a record of many subjects exhibiting a certain brain abnormality, such as a certain type of epilepsy. In another example, based on goals, in some embodiments, multiple subjects may be required to perform a particular cognitive task. Examples of specific cognitive tasks include, but are not limited to, memory recall, stimulus recognition, performing attention tasks, meditation, learning, watching movies, observing images, a strong set during motor maneuvers, and the like.

Deconstructing the at least one plurality of electrical signal data: in some embodiments, the at least one plurality of electrical signal data is recorded over a specified time period. In some embodiments, the at least one plurality of electrical signal data is recorded for up to one hour. In some embodiments, the at least one plurality of electrical signal data is recorded for up to 50 minutes. In some embodiments, the at least one plurality of electrical signal data is recorded for up to 40 minutes. In some embodiments, the at least one plurality of electrical signal data is recorded for up to 30 minutes. In some embodiments, the at least one plurality of electrical signal data is recorded for up to 20 minutes. In some embodiments, the at least one plurality of electrical signal data is recorded for up to 10 minutes.

In some embodiments, the recorded at least one plurality of electrical signal data is deconstructed into a plurality of deconstructed wavelet packet atoms. Each individual deconstructed wavelet packet atom within the plurality of deconstructed wavelet packet atoms corresponds to a brain activity characteristic ("BAF").

In some embodiments, an exemplary specially programmed processor of the present invention is programmed to deconstruct at least one plurality of electrical signal data into a plurality of deconstructed wavelet packet atoms using different mother wavelets and other orthogonal decompositions, such as, but not limited to, orthogonal cosine transforms and wavelet transforms. In some embodiments, an exemplary specially programmed processor of the present invention is programmed to utilize a particular orthogonal decomposition to minimize a decomposition processing time proportional to an n log (n) time, where n is the number of samples in a windowed frame.

In some embodiments, the mother wavelet is selected from a family of wavelets including, but not limited to, the group of: haar, coifletdubehies and Mayer wavelet families. Html at website http:// www.mathworks.com/help/wavelet/ref/wavelet families? Other families of wavelets suitable for the mother wavelet are described in refresh true according to some embodiments of the present invention.

In some embodiments, the specially programmed processor of the present invention is programmed to obtain a set of n-dimensional vectors based on an orthogonal decomposition algorithm, wherein each vector represents a BAF.

In some embodiments, an exemplary specially programmed processor of the present invention is programmed to perform the decomposition to achieve at least one predetermined objective. For example, the at least one predetermined objective may be based on identifying a common optimal basis that achieves a particular distinction under a particular coefficient distribution (unsupervised/supervised hybrid objective) and that may be common for data analysis for a group of people.

In some embodiments, an exemplary specially programmed processor of the present invention is programmed to determine projections (convolutions) on a selected basis function or some statistics of these projections to produce an output interpretation of a particular brain activity associated with a particular BAF. For example, a specially programmed processor of the present invention is programmed to determine a particular BAF based on activity in each such projection. In some embodiments, an exemplary specially programmed processor of the present invention is programmed to estimate the energy of each projection (e.g., the variance of the signal), the maximum of the orthogonal distribution, or other suitable statistical measure, such as, but not limited to, the value of negative entropy.

In some embodiments, the at least one plurality of recorded electrical signal data is solved into a plurality of deconstructed wavelet packet atoms according to the optimal basis algorithm disclosed in Coifman, r.r., & Wickerhauser, m.v., IEEE transfactisonon Information Theory, 38(2), 713- & 718(1992), which is incorporated herein by reference, in particular based on the description of orthogonal decomposition of shannon's equations, detailed in section three, the entropy of vectors.

In particular, the exemplary specially programmed processor of the present invention identifies a minimum entropy basis to be used in orthogonal decomposition of a particular at least one of the plurality of electrical signal data. In some embodiments, an exemplary specially programmed processor of the present invention performs shannon entropy analysis on at least a plurality of electrical signal data to obtain a joint optimal basis. When considering at least one plurality of electrical signal data to obtain a joint optimal basis, in one embodiment of the patent, it is possible to select the mapping M to include additional characteristics that focus on specific properties of the at least one electrical signal data that is joined. For example, if M (1) and M (2) satisfy the definition of the mapping as an additional information cost function, then an optimal basis that depends on the sum of the two functions results. In some embodiments, an exemplary specially programmed processor of the present invention is programmed to add a new additional cost function that measures the coefficient distribution at each node in a particular wavelet packet tree to identify a function M that searches for wavelet packet coefficients with minimum shannon entropy or a modified additional optimization function (across wavelet decomposition) that is averaged across all data observations.

In some embodiments, the recorded at least one plurality of electrical signal data is deconstructed into a plurality of deconstructed wavelet packet atoms according to another suitable optimal basis algorithm disclosed in Stainvas, I and integrator, n.

In some embodiments, the recorded at least one plurality of electrical signal data is deconstructed into a plurality of deconstructed wavelet packet atoms according to another suitable optimal basis algorithm disclosed in Intrater, N, Neural Computation 5, 443, 455(1993), the specific disclosure of which is incorporated herein by reference.

In some embodiments, the recorded at least one plurality of electrical signal data is deconstructed into a plurality of deconstructed wavelet packet atoms according to another suitable optimal basis algorithm disclosed in Intrater, N, Neural Computation 4, 98-1-7(1992), the specific disclosure of which is incorporated herein by reference.

For example, in some embodiments, an exemplary specially programmed processor of the present invention is programmed to utilize moving window frames along a time series to obtain different data observations. In one example, an exemplary specially programmed processor of the present invention is programmed to utilize a particular window frame and overlap for analyzing data segments. In one example, an exemplary specially programmed processor of the present invention is programmed to utilize 4 second window frames with 75% overlap between successive window frames. In some embodiments, an exemplary specially programmed processor of the present invention is programmed to utilize a window of length index 2, so, for example, if the sampling rate is 256Hz, a 4 second window will result in 1024 samples. In another example, if the sampling frequency is 250Hz, an exemplary specially programmed processor of the present invention is programmed to utilize a window frame that is slightly more than 4 seconds (e.g., 4.05 to 4.2). In another example, if the sampling frequency is 496Hz, then an exemplary specially programmed processor of the present invention is programmed to utilize a window frame that is slightly more than 4 seconds (e.g., 4.05 to 4.2). In another example, if the sampling frequency is 496Hz, then an exemplary specially programmed processor of the present invention is programmed to utilize a window frame that is slightly more than 4 seconds (e.g., 4.05 to 4.2).

In another example, an exemplary specially programmed processor of the present invention is programmed to use windowed frames of 1 second between adjacent frames to obtain vector updates every second to generate a projection matrix of size 121 × N (seconds of data) -3 (since the first frame is 4 seconds, then 1 second per frame). In some embodiments, an exemplary specially programmed processor of the present invention is programmed to rescale the full matrix to obtain the maximum dynamic range of the visual mapping of the data.

Determination of the best set: in some embodiments, the optimal set of wavelet packets is determined according to the optimal basis algorithm disclosed in Coifman, r.r., & Wickerhauser, m.v., ieee transactions on Information Theory, 38(2), 713-718(1992), which is incorporated herein by reference, particularly the description of orthogonal decomposition.

In some embodiments, the optimal set of wavelet packets is determined according to another suitable optimal basis algorithm disclosed in Stainvas, I and integrator, n., in.j.appl.mathematics, 4(J06), 1-22(2006), the specific disclosure of which is incorporated herein by reference.

In some embodiments, the optimal wavelet packet primitive subset is determined according to another suitable optimal basis algorithm disclosed in Intrater, N, Neural Computation 5, 443-.

In some embodiments, the optimal wavelet packet primitive subset is determined according to another suitable optimal basis algorithm disclosed in Intrater, N, Neural Computation 4, 98-1-7(1992), the specific disclosure of which is incorporated herein by reference.

In some embodiments, the number of best focused wavelet packet atoms is reduced by applying the wavelet de-noising algorithm disclosed in Donoho D.L., IEEE Transactions on information theory, 41(3), 613-627 (1995).

In some embodiments, the number of best focused wavelet packet atoms is reduced by applying an LI denoising method.

In some embodiments, the number of best focused wavelet packet atoms is reduced by applying an L2 denoising method.

In some embodiments, the number of best focused wavelet packet atoms is reduced by applying a hard threshold method.

Reordering the plurality of deconstructed wavelet packet atoms: in some embodiments, the de-noised original subset of best wavelet packets is reordered such that more physiologically relevant BAFs based on analysis of the total signal data appear visually closer geographically/spatially, e.g., as shown in fig. 5.

In some embodiments, reordering is optional.

In some embodiments, the de-noised optimal primitive subset of wavelet packets is reordered by a specially programmed computer that performs the steps of:

1. determining a line length for each data point in the projection by determining an average or sum of the absolute distances of the statistical measures of the projections of the different channels and their adjacent channels;

2. storing the line length data in at least one computer data object; and

3. the stored projections are reordered to minimize the value that attenuates the average or sum of the line lengths across the projections, across the windows of every 4 seconds, and across all individuals within the plurality of individuals.

In some embodiments, the reordered statistical value is selected from the group consisting of: the average of the sum of absolute differences of wavelet packet atoms and the average of the sum of wavelet packet atoms (1-correlation).

Obtaining a predetermined set of normalization factors: in some embodiments, the predetermined set of normalization factors is obtained by determining a mean and standard deviation of the stored projected values.

In some embodiments, brain activity is represented by the energy of BAF in the individual. In some embodiments, the energy is determined based on a variance of the signal. In some embodiments, the energy is a maximum of the energy of the individual BAF. In some embodiments, the energy is a negative entropy of an energy coefficient of the individual BAF, such as Coifman and Wickerhauser.

In some embodiments, the BAF representation of the subject is used to determine the contribution of each BAF to the total energy of the signal being recorded. For example, the BAF representation of the subject is used to determine the contribution of each BAF to the total energy of the signal being recorded based at least in part on:

1) at least one orthogonality condition for orthogonal decomposition and/or

2) Summation of the orthogonal components using the Parseval equation applicable to the BAF representation.

In some embodiments, the BAF of the subject represents the contribution of each BAF to the total length of the virtual line resulting from the line segmentation from which the BAF peak was obtained; wherein the virtual line identifies at least one communication channel used by the individual brain subsystems of the subject to communicate with each other. In some embodiments, the contribution of each BAF to the total length of the virtual line measures the smoothness of brain activity in different, geographically proximate BAFs.

In some embodiments, the contribution of each BAF to the total energy of the signal and the contribution of each BAF to the total length of each virtual line produced by the line segmentation that obtains the peak of BAF activity are used to determine which BAF is present in the final BAF representation. In some embodiments, the particular BAFs present in the BAF representation of the subject are those BAFs that contribute significantly to the variance, and are assumed to contribute less to the total virtual line length.

In some embodiments, the present invention provides a system capable of having an underlying mental state, an underlying neurological condition, or a combination of an underlying mental state and an underlying neurological condition in a particular individual. Thus, in some embodiments, the system may be used by a physician.

In some embodiments, the device for recording brain electrical activity of a subject may be worn continuously and non-invasive or unobtrusive. Thus, in some embodiments, identification of a nerve injury or determination of a first mental state of a subject may be accomplished at an earlier time or may be accomplished more efficiently than other methods, as the subject is monitored in a more natural or less clinical setting. In some embodiments, the system of the present invention enables earlier detection, identification or diagnosis of a mental state and/or neurological condition of an individual.

In some embodiments, the present invention provides a specially programmed computer system comprising:

a. at least one special purpose computer comprising:

i. a non-transitory memory electronically storing specific computer executable program code; and

at least one computer processor that, when executing specific program code, becomes a specially programmed computer processor configured to perform at least the following:

1. obtaining, by a specially programmed processor in real time, electrical signal data representative of brain activity of a particular individual;

2. processing in real time electrical signal data representative of brain activity of a particular individual based on a predetermined predictor associated with a particular brain state, the predictor selected from a predictor library comprising a plurality of predetermined predictors, wherein each individual predetermined predictor is associated with a unique brain state,

wherein the predetermined predictors associated with a particular brain state include:

i. a predetermined mother wavelet is set to a predetermined mother wavelet,

a predetermined representative original subset of wavelet packets,

a predetermined ordering of wavelet packet atoms generated from a predetermined mother wavelet, an

A predetermined set of normalization factors,

wherein the processing comprises:

i. deconstructing the electrical signal data into a plurality of predetermined deconstructed wavelet packet atoms by a specially programmed processor using a predetermined representative original subset of wavelet packets,

wherein a time window of electrical signal data is projected onto a predetermined representative original subset of wavelet packets

Wherein the projection is via convolution or inner product, and wherein each predetermined representative wavelet packet atom corresponds to a particular predetermined brain activity feature from a plurality of predetermined brain activity feature libraries;

storing a plurality of predetermined deconstructed wavelet packet atoms in at least one computer data object;

reordering the stored plurality of predetermined deconstructed wavelet packet atoms within the computer data object based on utilizing the predetermined order by a specially programmed processor;

obtaining a statistical measure of activity for each of the reordered plurality of predetermined deconstructed wavelet packet atoms; and

normalizing the reordered plurality of predetermined wavelet packet atoms based on utilizing a predetermined normalization factor; and

3. outputting, based on the processing, a visual indication of at least one personalized mental state for the particular individual, a visual indication of at least one personalized neurological condition for the particular individual, or both,

wherein an individual predetermined predictor associated with a particular brain state within the plurality of predetermined predictors is generated by:

i. obtaining a predetermined representative primitive subset of wavelet packets by:

1. obtaining, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity of a particular brain state from a plurality of individuals;

2. a mother wavelet is selected from a plurality of mother wavelets,

wherein the mother wavelet is selected from a wavelet family selected from the group consisting of: haar, coifletdubehies and Mayer wavelet families;

3. deconstructing, by a specially programmed processor, the at least one plurality of electrical signal data into a plurality of wavelet packet atoms using the selected mother wavelet;

4. storing a plurality of wavelet packet atoms in at least one computer data object;

5. determining an optimal original subset of wavelet packets using a predetermined mother wavelet and storing the optimal original subset of wavelet packets in at least one computer data object,

wherein the determination is via utilization of an optimal basis algorithm; and

6. applying wavelet de-noising, by a specially programmed processor, to a number of wavelet packet atoms in the optimal set;

obtaining a predetermined ordering of wavelet packet atoms by:

1. projecting, by a specially programmed processor, at least one plurality of electrical signal data representing brain activity every 4 second data window onto a predetermined representative primitive subset of wavelet packets;

2. storing the projections in at least one computer data object;

3. determining, by a specially programmed processor, a line length for each data point in the projection by determining an average absolute distance of the statistical measures of the projections of the different channels from their neighbors;

4. storing the line length data in at least one computer data object; and

5. reordering, by a specially programmed computer, the stored projections to minimize statistics of line length values across each time window, across all individuals within a plurality of individuals, and across the projections; and

obtaining a predetermined set of normalization factors by:

1. the mean and standard deviation of the stored projected values are determined by a specially programmed computer.

Reference is now made to the following examples, which together with the above descriptions illustrate some embodiments of the invention in a non limiting manner.

Examples of the invention

Example 1: BAF representations obtained from a plurality of subjects performing specific cognitive tasks.

Figure 5 shows BAF representations from 7 subjects. Each row (Y-axis) represents activity of a single BAF. Color coding ("hot") means that high level activity tends to be red and low level activity tends to be blue. Each column (X-axis) represents an activity vector for a particular time frame. Thus, the X-axis is measured in time (minutes or seconds). Fig. 5 is measured in minutes. The specific figure shows the activity of 6 subjects during the cognitive task. Thus, different recorded EEG's may be connected together such that one figure (or BAF matrix) represents the activity of different tasks or different subjects performing (different) tasks. In this particular figure, the arrows indicate two subjects with relatively similar activity patterns during the task. Thus, the figure represents different cognitive strategies performed during the same cognitive task performed by different subjects.

Example 2: BAF representations obtained from subjects watching three movies eliciting different emotional responses.

Brain Activity (BAF) during three movies (recorded at different times each). Referring to fig. 6, it can be seen that the activity of the second and third movies is more similar in terms of active BAFs than the activity in the first movie. The first movie, the "over-the-track pursuit" (Derailed), involves violence and terrorism, while the two other movies, the "steal life" (StolenLife) and the "flesh life" (Skin), are more associated with feelings of sadness and warmth. The arrows mark the time of strong positive emotional feelings in both movies. It is clear that the active BAF is the same at that time. They are also active during meditation in figure 7 below.

Example 3: BAF representations obtained from seven subjects performing e-learning tasks.

According to some embodiments of the invention, EEG recordings were obtained from 7 subjects, while the subjects performed a specific e-learning task. BAF representation is shown in fig. 7. The task involves gathering information and then answering the questions for two difficulty levels (this is the same representation as in fig. 5). Next are two experienced medinders performing the gayantri meditation and one inexperienced person performing the Japa meditation followed by listening to the music. During the gayacri meditation, active BAF is similar to that during the positive mood portion of the movie (fig. 6). It is evident that experienced meditatics exhibit increased mPFC, co-activation of the isle and temporal lobes, while reducing the activity of the default mode network. Happiness was also shown.