Device for generating scores indicative of cognitive conditions such as attention deficit hyperactivity disorder

Abstract

The present disclosure relates to generating a score indicative of a subject having a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD). In particular, a score may be generated based on electroencephalogram or magnetic map (EEG / MEG) data by calculating one or more correlation values between a pair of time series determined based on EEG / MEG data. Devices, systems, computer-readable media, and computer-implemented methods described herein receive EEG / MEG data of a subject. A first plurality of time sequences is then determined based on the EEG / MEG data, where each time sequence of the first plurality of time sequences corresponds to a respective source location within the cranial cavity of the subject. A first correlation value may then be calculated for a first pair of time series included in the determined first plurality of time series. A score is generated based on the first correlation value that indicates that the subject has a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD). The score is output to, for example, a physician that may diagnose the subject.

Description

Background Art

[0001] The Bayesian theory of the brain assumes that probability distributions are represented throughout the cortex and that these distributions converge to estimates only when a decision needs to be made. However, the way in which this convergence process, i.e. the transition from a probabilistic brain activity to a deterministic or one-shot decision, occurs is currently unclear. Therefore, a measure remains to be derived that can reconcile stimulus-dependent surprise (Shannon surprise) and allow the calculation of the information gain between prior and posterior beliefs (Bayesian surprise). Expressing the probability of an event in terms of its information content (e.g., the -log of the above probability), i.e., Shannon surprise, implies a transition from a probabilistic to a deterministic environment; in fact, being surprised at the occurrence of an event may mean that the event has occurred, and thus the generated information bits may provide accuracy, that is, they may offset the uncertainty about the occurrence of the event. Therefore, it may be necessary to distinguish between bits of accuracy and bits of Shannon information content, and bits of Shannon entropy that express uncertainty.

[0002] In attention research, it is assumed that the saliency map generated by the subject's sensory saliency is modulated by top-down behavioral perceptual signals to generate a priority map. While neural encoding has been extensively studied (e.g., through computational modeling of attention), proper processing of top-down attention signals has not yet been achieved. Instead, most prior art modeling studies have focused on the bottom-up component of visual attention.

[0003] Furthermore, due to the complexity and heterogeneity of cognitive conditions, for example cognitive disorders such as attention deficit hyperactivity disorder (ADHD), the traditional gap between the neurobiology and symptoms of such disorders is often large and, therefore, their underlying neurocognitive mechanisms are often unknown.

[0004] Therefore, it may be beneficial to provide an improved system, corresponding computer-readable medium, and corresponding computer-implemented method for generating a score indicating that a subject has a cognitive condition (eg, a cognitive disorder, such as attention deficit hyperactivity disorder ADHD). Summary of the invention

[0005] The present disclosure relates to the generation of a score indicating that a subject suffers from a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder ADHD. In particular, a score can be generated based on electroencephalogram or magnetogram (EEG / MEG) data by calculating one or more correlation values ​​between a pair of time series determined based on EEG / MEG data. The devices, systems, computer-readable media, and computer-implemented methods described herein generate a score indicating that a subject suffers from a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD).

[0006] According to the present disclosure, a device is configured to receive first EEG / MEG data of a subject. The device is also configured to determine a first plurality of time series based on the EEG / MEG data, wherein each time series in the first plurality of time series corresponds to a corresponding source position in the cranial cavity of the subject. In addition, the device is configured to calculate a first correlation value for a first pair of time series, wherein the first pair of time series is included in the determined first plurality of time series. The device is also configured to generate a score based on the first correlation value, the score indicating that the subject has a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD). In addition, the device is configured to output a score.

[0007] According to one embodiment, the corresponding source position of the first time series in the first pair of time series may be located in the frontal lobe in the cranial cavity of the subject, such as the middle frontal gyrus (MFG) and / or the inferior frontal gyrus (IFG), and the corresponding source position of the second time series in the first pair of time series may be located in the superior parietal lobe (SPL) in the cranial cavity of the subject.

[0008] According to another embodiment, the device may also be configured to calculate a second correlation value for a second pair of time series, the second pair of time series being included in the determined first plurality of time series. A score may be generated based on the first correlation value and the second correlation value. According to this embodiment, the corresponding source position of the first time series in the second pair of time series may be located in a ventral attention network region in the subject's cranial cavity, such as the temporoparietal junction (TPJ), and the corresponding source position of the second time series in the second pair of time series may be located in a dorsal attention network region in the subject's cranial cavity, such as the superior parietal lobe (SPL).

[0009] According to an embodiment, the device may also be configured to calculate a third correlation value for a third pair of time series, the third pair of time series being included in the determined first plurality of time series. A score may be generated based on the first correlation value, the second correlation value, and the third correlation value. In addition, the corresponding source position of the first time series in the third pair of time series may be located in a ventral attention network region in the subject's cranial cavity, such as the temporoparietal junction (TPJ), and the corresponding source position of the second time series in the second pair of time series may be located in a ventral visual pathway region in the subject's cranial cavity, such as the inferior temporal gyrus (ITG).

[0010] In another embodiment, the device may also be configured to receive a second EEG / MEG data of the subject, and determine a second plurality of time series based on the second EEG / MEG data, wherein each time series in the second plurality of time series corresponds to a corresponding source position in the cranial cavity of the subject. In addition, the device may also be configured to calculate a fourth correlation value for a fourth pair of time series, the fourth pair of time series being included in the determined second plurality of time series, and the fourth pair of time series corresponding to the same corresponding source position as the first pair of time series. In addition, the device may be configured to calculate a comparison value between the fourth correlation value and the first correlation value. In addition, a score may be generated based on the comparison value.

[0011] In an embodiment, the step of determining the first plurality of time series may include filtering the first EEG / MEG data so that the first plurality of time series includes only a subset of all available time series in the first EEG / MEG data, the subset including a first pair of time series, a second pair of time series, and a third pair of time series.

[0012] In an embodiment, a system may include an apparatus according to any of the above-described embodiments, and may further include an EEG or MEG device for measuring at least first EEG or MEG data of a subject, and a task presentation device for presenting one or more tasks to the subject while the EEG / MEG device is measuring at least first EEG / MEG data of the subject.

[0013] According to the present disclosure, a computer-implemented method includes receiving first electroencephalogram or magnetoencephalogram (EEG / MEG) data of a subject. The method also includes determining a first plurality of time series based on the first EEG / MEG data, wherein each time series in the first plurality of time series corresponds to a corresponding source position within the cranial cavity of the subject. In addition, the method includes calculating a first correlation value for a first pair of time series, the first pair of time series being included in the determined first plurality of time series. The method also includes generating a score based on the first correlation value, the score indicating that the subject has a cognitive condition, such as a cognitive disorder, such as ADHD, and outputting the generated score.

[0014] According to an embodiment, the corresponding source position of the first time series in the first pair of time series may be located in the frontal lobe in the cranial cavity of the subject, such as the middle frontal gyrus (MFG) and / or the inferior frontal gyrus (IFG), and the corresponding source position of the second time series in the first pair of time series may be located in the superior parietal lobe (SPL) 340 in the cranial cavity of the subject.

[0015] According to an embodiment, the method may further include calculating a second correlation value for a second pair of time series, the second pair of time series being included in the determined first plurality of time series. A score may be generated based on the first correlation value and the second correlation value. In this embodiment, the corresponding source position of the first time series in the second pair of time series may be located in a ventral attention network region in the subject's cranial cavity, such as the temporoparietal junction (TPJ), and the corresponding source position of the second time series in the second pair of time series may be located in a dorsal attention network region in the subject's cranial cavity, such as the superior parietal lobe (SPL).

[0016] According to another embodiment, the method may further include calculating a third correlation value for a third pair of time series, the third pair of time series being included in the determined first plurality of time series. A score may be generated based on the first correlation value, the second correlation value, and the third correlation value. In addition, the corresponding source position of the first time series in the third pair of time series may be located in a ventral attention network region in the subject's cranial cavity, such as the temporoparietal junction (TPJ), and the corresponding source position of the second time series in the second pair of time series may be located in a ventral visual pathway region in the subject's cranial cavity, such as the inferior temporal gyrus (ITG).

[0017] According to an embodiment, the method may further include receiving second EEG / MEG data of the subject (and determining a second plurality of time series based on the second EEG / MEG data, wherein each time series in the second plurality of time series corresponds to a corresponding source position within the cranial cavity of the subject. In addition, the method may further include calculating a fourth correlation value for a fourth pair of time series, the fourth pair of time series being included in the determined second plurality of time series, and the fourth pair of time series corresponding to the same corresponding source position as the first pair of time series. In addition, the method may include calculating a comparison value between the fourth correlation value and the first correlation value. A score may be generated based on the comparison value.

[0018] According to an embodiment, the step of determining the first plurality of time series may include filtering the first EEG / MEG data so that the first plurality of time series includes only a subset of all available time series in the first EEG / MEG data, the subset including the first pair of time series, the second pair of time series and the third pair of time series.

[0019] The above embodiments may be combined with each other. The above embodiments may also be implemented on a computer-readable medium comprising computer-readable instructions, which, when executed by a processor, cause the processor to perform the above steps.

[0020] This summary is provided to introduce some concepts in a simplified form that are further described in the detailed description below. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] These and other aspects and features of the present disclosure will be described by the following detailed description of embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0022] Figure 1 is a diagram illustrating the Shannon brain model according to an embodiment of the present disclosure.

[0023] Figure 2 is a diagram illustrating a dorsal-ventral attention network according to an embodiment of the present disclosure.

[0024] Figure 3 is a graph illustrating conditional probabilities and information according to an exemplary task (“Task 2”) performed by a subject while EEG or MEG data is measured from the subject.

[0025] Figure 4 is a graph illustrating experimental results showing associations with significantly different PLV values ​​in a comparison between two conditions of the same task ("Task 1") measured from healthy subjects.

[0026] Figure 5 is a graph illustrating experimental results showing associations with significantly different PLV values ​​in a comparison between two conditions of two respective tasks ("Task 1" and "Task 2") measured from healthy subjects.

[0027] Figure 6 is a graph illustrating experimental results showing associations with significantly different PLV values ​​in comparison between two conditions of two respective tasks ("Task 1" and "Task 2") measured from subjects with ADHD.

[0028] Figure 7 is a graph illustrating experimental results showing connections of the central executive network where conditions for a task ("Task 1") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0029] Figure 8 is a graph illustrating experimental results showing connections of a saliency network where conditions for a task ("Task 1") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0030] Fig. 9 is a graph illustrating experimental results showing connections of the central executive network where conditions for a task ("Task 2") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0031] Fig.10 is a graph illustrating experimental results showing connections of the central executive network where conditions for a task ("Task 2") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0032] Fig.11 is a graph illustrating experimental results showing connections of a saliency network where conditions for a task ("Task 1") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0033] Fig.12 is a graph illustrating experimental results showing connections of the central executive network where conditions for a task ("Task 1") had significantly different PLV values ​​in a comparison between healthy subjects and subjects with ADHD.

[0034] Fig.13 is a diagram illustrating an example system in accordance with one or more implementations of the present disclosure.

[0035] Fig.14 is a diagram illustrating an example data analysis device according to an embodiment of the present disclosure.

[0036] Fig.15 This is a diagram illustrating the brain in the cranial cavity of a subject.

[0037] Fig.16 is a diagram illustrating a method according to an embodiment of the present disclosure.

[0038] Fig.17 is a block diagram of various exemplary components of an attention detection system for accurately detecting attention deficits according to some embodiments of the present disclosure.

[0039] FIG. 18A to FIG. 18B is a tabular representation of exemplary tasks for attention deficit detection according to some embodiments of the present disclosure.

[0040] Fig.19 Illustrated is a connection network between brain regions according to some embodiments of the present disclosure.

[0041] Fig. 20 is a flow chart depicting the operation of an exemplary method for attention deficit detection according to some embodiments of the present disclosure.

[0042] Fig.21 is a flow chart depicting the operation of an exemplary method for attention deficit detection according to some embodiments of the present disclosure. DETAILED DESCRIPTION

[0043] Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to like elements throughout.

[0044] According to one embodiment, the probability in Bayes' theorem can be converted to Shannon information content by applying a -log function on both sides of the equation. Therefore, a Shannon metric, namely transfer information content (TIC), can be defined, which can reconcile Shannon surprise and Bayesian surprise. In addition, in the field of action selection or decision making, TIC can explain how information flows (e.g., via which pathway or network) to increase or decrease uncertainty about an action depending on the behavioral relevance of the stimulus. Therefore, TIC can be the link between belief and knowledge in cognitive modeling.

[0045] The decomposition of TIC can show that two information contents can be generated by a single observation of a stimulus. One of them, called the bottom-up component, can be the information content associated with the evidence (e.g., the surprise associated with the stimulus that may cause the mentioned event to occur), which can contribute positive bits to the information flow; the other is the top-down component, which can be the information content associated with the Bayesian likelihood, which can contribute negative information bits to the information flow. Therefore, if the absolute value of the latter is smaller than the former, this may result in a net positive information measure (TIC>0); otherwise, if the negative bits introduced by the top-down component are greater than the positive bits introduced by the bottom-up component, this may result in a net negative information measure (TIC<0).

[0046] Furthermore, conditional TICs can be defined that can be derived from the conditional Bayes theorem to derive information flows conditioned on an internally maintained contextual representation (hereafter referred to as contextual TICs) or a temporal context (hereafter referred to as episodic TICs). TICs can have an additive property, so the total amount of uncertainty reduction can be computed by adding the consecutive sensorimotor, contextual, and episodic TICs generated during evidence integration, such that uncertainty about action selection can increase or decrease depending on the signs of these TICs, and a given action can be selected only if this uncertainty is converted to confidence.

[0047] According to embodiments, insights gained from TIC can be used to discover attention networks for cognitive conditions, such as cognitive disorders, such as attention deficit hyperactivity disorder (ADHD). For example, subjects with cognitive conditions (e.g., cognitive disorders such as ADHD) may need to increase activity in the prefrontal cortex (PFC) when processing exceptions from a particular condition (e.g., episodic task switching) or filtering unimportant stimuli, while healthy subjects may show better performance when processing exceptions from the condition or filtering unimportant stimuli by increasing the activity of the dorsal-ventral attention network discovered according to one or more embodiments of the present disclosure, which does not involve an increase in PFC activity. This may explain why subjects with cognitive conditions (such as ADHD) may have more difficulty staying focused, filtering out unimportant information, or processing exceptions (compared to healthy subjects), because subjects with cognitive conditions may need to use the PFC to filter out unimportant information or use the PFC when processing exceptions, which may be more tiring than using the direct dorsal-ventral attention network without involving the PFC (as healthy subjects may do). Therefore, this network may play a central role in filling the gap between neurobiology and ADHD symptoms.

[0048] As described below, at least one of the activity of the prefrontal cortex (PFC) pathway or the activity of the dorsal-ventral attention network may serve as a marker for detecting whether a subject suffers from a cognitive condition, such as a cognitive disorder such as attention deficit hyperactivity disorder (ADHD).

[0049] Shannon brain model

[0050] It is proposed in the prior art that neurons encode the logarithmic probability that a feature (e.g., a feature of an image) can take a specific value. The idea that neurons encode logarithmic probabilities is discussed in Rao, RPN, 2004, Bayesian Computation in Recurrent Neural Circuits (Neural Comput. 16, pages 1 to 38) and in Pouget, A., et al., 2013, Probabilistic brains: knowns and unknowns (Nat. Neurosci. 16 1170-1178). However, how the Bayesian brain transitions from a probabilistic brain to a deterministic or single decision (the so-called "collapse" process) is not well understood in the prior art.

[0051] According to an embodiment, event x i The Shannon information content or hereinafter also simply referred to as information content (which may be expressed in bits) may be given by:

[0052] I(x i )=-log2[p(x i )](bit) (Equation 1)

[0053] Shannon entropy, also in bits, can be expressed as a weighted sum of the information content of each event and its probability of occurrence:

[0054]

[0055] Therefore, event x i The probability of can be equated to its information content. The probability can be expressed in terms of information content according to Equation 1:

[0056]

[0057] Shannon interpreted the amount of information carried by a message as a measure of how surprised people are to learn the message. It may therefore be natural to apply the same idea to the unaveraged information content given above in equation 1, i.e., to say that this measure can represent in bits the surprise that may be caused by the occurrence of a given event. From equations 1 and 3, it can be seen that the higher / lower the probability of a given event, the lower / higher its information content, i.e., the surprise it causes.

[0058] According to one embodiment, it is shown that by expressing classical probability theorems (such as Bayes' theorem) in terms of Shannon information content, it is possible to understand how information is transmitted and how classical negative information is generated. On average, as in the expression of entropy in Equation 2, interesting statistical implications can be given, such as uncertainty about the content of the message, which is what Shannon was interested in for communication purposes, but the accuracy obtained by a given measurement may be lost. When an event occurs, probabilistic reasoning (occurring in the brain) may collapse into a deterministic environment given by Shannon information content. As further shown below, this deterministic feature of Shannon information content may be a bridge between belief and knowledge in cognitive modeling.

[0059] According to one embodiment, a Shannon metric can be derived that can reconcile stimulus-dependent surprise (such as Shannon surprise) and can allow for the calculation of information gain between prior beliefs and posterior beliefs (such as Bayesian surprise). According to this embodiment, Bayesian theorem is expressed in terms of Shannon information content by applying a -log function on both sides of the Bayesian theorem. Under Bayesian theorem, the effect of data x on an observer can be to change the prior distribution P(θ) of the model θ to the posterior distribution P(θ|x):

[0060] P(θ|x)=P(x|θ)P(θ) / P(x) (Equation 4)

[0061] Where P(x) is called the evidence and P(θ|x) is called the likelihood. Then, taking the -log2 function on both sides of Equation 4 yields:

[0062] -log2P(θ|x)=-log2[P(x|θ)P(θ) / P(x)](Equation 5)

[0063] By applying the properties of the logarithmic function, the right side of Equation 5 can be expanded to obtain:

[0064] -log2P(θ|x)=-log2P(x|θ)-log2P(θ)+log2P(x) (Equation 6)

[0065] By considering Equation 1, Equation 6 can be expressed in terms of the information content corresponding to the individual probabilities that appear in Bayes' theorem:

[0066] I(θ|x)(bit)=I(θ)(bit)-I(x)(bit)+I(x|θ)(bit) (Equation 7)

[0067] In Equation 7, each term may be the information content of a certain event, and it may represent the surprise in bits that caused the occurrence of the event. The interpretation of this last equation may allow understanding of how information bits introduced in the space of multiple models may be transferred to each of the multiple models, and concepts such as negative information may naturally emerge. It may be inferred from Equation 7 that the effect of the data on the observer may be to change the information content of a given model θ in the space of multiple models, i.e., it may change (e.g., increase or decrease) the surprise that caused the occurrence of the mentioned event. Therefore, the term I(θ|x) on the left side of Equation 7 may be referred to as the posterior information content of a single model θ, and the first term I(θ) on the right side of the equation may be referred to as the prior information content of a single model θ.

[0068] An aspect of Equation 7 is that the mentioned change in the information content of a single model θ in the space of multiple models that may be caused by a single observation x may include two information contents I(x) and I(x|θ) generated by the mentioned single observation (rather than just one information content as might be expected). Thus, in Equation 7, -I(x) (bits) + I(x|θ) (bits) may represent the information transferred to the single model θ due to the observation x.

[0069] The transfer information content (TIC) can be defined as:

[0070] TIC(x→θ)(bit)=I(x)(bit)-I(x|θ)(bit) (Equation 8)

[0071] Therefore, Equation 7 can be expressed as:

[0072] I(θ|x)(bit)=I(θ)(bit)-TIC(x→θ)(bit) (Equation 9)

[0073] In Equation 8, the first term I(x) can be the Shannon information content associated with the evidence (e.g., it is the Shannon surprise), and it can contribute with positive bits to reduce the surprise of the posterior information content I(θ|x). The second term I(x|θ) in Equation 8 can also be the Shannon information content, but is associated with the Bayesian likelihood so that it can be Bayesian at the same time. This second term can contribute with negative bits to increase the surprise of the posterior information content I(θ|x). For example, this can take into account the surprise of the evidence I(x) so that only those bits associated with a single model θ can be transmitted to it. Therefore, TIC can allow the information gain (e.g., Bayesian surprise) between prior beliefs and posterior beliefs to be calculated. Because the Shannon information content associated with the Bayesian likelihood can be a link between a probabilistic framework and a deterministic framework, this component of TIC can be a candidate for modeling Bayesian brain collapse (as further described below).

[0074] As already discussed above, TIC can describe information flow, so it is not surprising that its sign can be of interest. In particular, if the negative bits of information content of the Bayesian likelihood I(x|θ) do not cancel the positive bits of information content of the evidence I(x), then TIC is positive (i.e., there may be a positive net transfer of information); conversely, if the negative bits corresponding to I(x|θ) are larger than the bits of evidence I(x), then TIC is negative and there may be a negative net transfer of information. Thus, TIC can overcome limitations of other information gain models.

[0075] According to one embodiment, a conditional TIC may be defined as:

[0076]

[0077] The conditional TIC may represent the information transferred to a single model θ as a result of observing x given that an event y has occurred. Like the unconditional TIC of Equation 8, the conditional TIC of Equation 10 may take positive or negative values.

[0078] Using Equation 9 in the context of action selection, the information transfer that triggers or does not trigger action a when stimulus s occurs can be expressed as the so-called sensorimotor TICTIC(s→a):

[0079] I(a|s)=I(a)-TIC(s→a) (Equation 11)

[0080] Where s can represent a stimulus and a can represent an action. Equation 11 shows that in action selection, the TIC from stimulus s to action a may reduce (TIC>0) or increase (TIC<0) the uncertainty of the latter. In the extreme case where TIC>0 can offset the prior uncertainty I(a), it can be confident that the action will occur, that is, I(a)=0. In other words, when the uncertainty about the action is reduced to zero (via TIC), a decision to perform the action can be made. As described above, TIC can be the subtraction of two information contents. Adding and subtracting information content can be done on a short time scale (e.g., the time scale of brain attention operations), in contrast to multiplying and dividing probabilities, which may be the case when the brain (e.g., Bayesian brain) is not collapsed to an estimate (e.g., to Shannon information content). Therefore, it can be proposed that TIC can be a (deterministic) neural response that can reflect action selection (or decision). This can bridge the gap between belief and knowledge. Because the net information transfer due to TICs can be positive or negative, this gap can be reduced or increased, and in both cases they can provide accuracy to the uncertainty of previous estimates.

[0081] The TIC of Equation 11 can be decomposed as follows:

[0082] I(a|s)=I(a)-TIC(s→a)=I(a)-I(s)+I(s|a) (Equation 12)

[0083] The TIC decomposition can allow identification and separation of bottom-up and top-down components in action selection (or decision making). This decomposition into bottom-up and top-down components can correspond to an attention model for action selection known as the Shannon brain model.

[0084] The first component of TIC, I(s), can represent the information content associated with the evidence (e.g., the stimulus-related Shannon surprise caused by stimulus s). Therefore, I(s) can be a bottom-up component in the attention model, which can generate positive information bits and can therefore help reduce the information content of the action (e.g., it can reduce or even offset its uncertainty). The second component of TIC, I(s|a) (e.g., the information content associated with the Bayesian likelihood) can represent the negative information generated in order to calculate the Bayesian surprise (which may increase the information content of the action (e.g., increase its uncertainty)). Therefore, I(s|a) can be a top-down component in the attention model, and therefore, a correct treatment of the top-down attention signal can be given. In other words, a high top-down contribution may lead to no action being taken.

[0085] According to another embodiment, the information transfer that can trigger or not trigger action a when stimulus s occurs in context c can also be expressed in TIC. The information content corresponding to the probability of triggering action a conditional on the occurrence of stimulus s in context c can be given by the following formula:

[0086] I(a|s,c)=-log2[P(a|s,c)](Equation 13)

[0087] Applying the conditional version of Bayes' theorem to the posterior probability in Equation 13 gives:

[0088]

[0089] Applying the logarithmic rule gives:

[0090]

[0091] The first term in Equation 15 can be given by the sensorimotor TIC of Equation 11. From Equation 10, it can be seen that the second term is the conditional TIC, called the contextual conditional TIC: TIC(c→a|s) and it can represent the information bits transferred to action a when stimulus s occurs in context c. Therefore,

[0092] I(a|s,c)=I(a)-TIC(s→a)-TIC(c→a|s) (Equation 16)

[0093] Now, the sensorimotor TIC and context-conditional TIC in Equation 16 can be decomposed into their two components:

[0094] I(a|s,c)=I(a)-TIC(s→a)-[I(c|s)-I(c|a,s)]=I(a)-I(s)+I(s|a)-I(c|s)+I(c|a,s) (Equation 17)

[0095] As for sensorimotor TIC, the components I(c|s) and I(c|a, s) of the context TIC in Equation 17 can represent the bottom-up component and the top-down component in action selection, respectively. The sensorimotor TIC may have a different sign from the context TIC, so that when considered in a given context, an informative stimulus with a positive sensorimotor TIC can generate a negative context TIC that compensates for the positive sensorimotor TIC, making the stimulus uninformative for that context.

[0096] In another embodiment, stimulus s may occur in context c, and in episode (such as temporal context) e, and the information transfer that may or may not trigger action a may be calculated based on TIC. In this case, the information content may be given by:

[0097] I(a|s,c,e)=-log2[P(a|s,c,e)] (Equation 18)

[0098] Applying the conditional version of Bayes’ theorem to the posterior probability in Equation 18 yields:

[0099]

[0100] Then, applying the logarithmic rule gives:

[0101]

[0102] The first term in Equation 20 can be given by Equation 13 and Equation 16. According to Equation 10, the second term can be the episode conditional TIC: TIC(e→a|s,c) and it can represent the information bits delivered to the action when the stimulus occurs in a given context c and episode e. Therefore, the information content can be:

[0103] I(a|s,c,e)=I(a)-TIC(s→a)-TIC(c→a|s)-TIC(e→a|s,c) (Equation 21)

[0104] Furthermore, the sensorimotor TIC, contextual conditional TIC, and episodic conditional TIC in Equation 21 can be decomposed into their two components:

[0105] I(a|s,c,e)=I(a)-TIC(s→a)-TIC(c→a|s)-[I(e|s,c)-I(e|a,s,c)]

[0106] =I(a)-I(s)+I(s|a)-I(c|s)+I(c|a,s)-I(e|s,c)+I(e|a,s,c) (Equation 22)

[0107] The components I(e|s, c) and I(e|a, s, c) of the scenario-conditional TIC in Equation 22 can represent the bottom-up component and the top-down component in action selection, respectively. According to the reasoning of the context, if its scenario-conditional TIC in Equation 22 is positive and compensates for the negativity of the context-conditional TIC, then an uninformative stimulus in a given context may become informative.

[0108] As can be seen from Equation 22, in a given context and episode, the uncertainty about the action can increase or decrease when a stimulus occurs, depending on the bottom-up contribution and the top-down contribution of the information content (e.g., the contribution of the log probability, which can be added and subtracted), as further described below. In Equations 21 and 22, I(a) can represent the bits of uncertainty about the action before any stimulus occurs, and I(a|s, c, e) is the uncertainty left after the stimulus occurs depending on the sign of the respective TIC (which can be higher or lower than I(a) for example), and the subject has taken into account the context c and the episode e (if applicable). The first component of each of the three TICs (I(s), I(c|s), I(e|s, c)) can correspond to the bottom-up contribution of the stimulus s, and it can always help to reduce the uncertainty of the action (e.g., pressing the button). Because these bottom-up components appear with a negative sign in Equation 22, that is, they help to reduce the bits of uncertainty about the action (increase the probability of its occurrence). They may always appear with a positive sign in the TIC formula, which is why they are considered to be positive information components (e.g., Shannon surprises). The opposite can be said about the second components of each of the three TICs (I(s|a), I(c|a, s), I(e|a, s, c), which are top-down components). They appear with a positive sign in Equation 22, i.e., they may contribute to increasing the uncertainty about the action. Because they appear with a negative sign in their TICs, they may be negative information components, and therefore, if their absolute values ​​are greater than their respective Shannon surprises (which are I(s), I(c|s), I(e|s, c)), the corresponding TIC may be negative, and its net contribution is to increase the uncertainty about the action (reduce its probability of occurrence). Since these second components of the TIC can contribute to making the stimulus (depending on the context or plot) behaviorally relevant (e.g., when they are low or zero) or non-behavioral relevant (e.g., when they are large), they can be considered to be top-down components. For example, a subject with a cognitive condition (e.g., a cognitive disorder such as ADHD) may have difficulty distinguishing behaviorally relevant from non-behavioral relevant stimuli, and thus such a subject may process the top-down component of the TIC differently (compared to healthy subjects).

[0109] In the prior art, the respective roles of the dorsal attention network and the ventral attention network are known; on the other hand, how the two networks communicate with each other remains a subject of significant controversy.

[0110] Once the task is explained to the subject, the task can be stored in the subject's memory (i.e., brain). The dorsal attention network can then send one or more top-down signals to the ventral attention network through the middle frontal gyrus (MFG), which can correspond to non-targets, distractors, or any other stimulus different from the target, and these top-down signals are then stored in the brain's area called the temporoparietal junction (TPJ). TPJ can include the posterior region of the superior temporal sulcus (superior temporal sulcus, STS), the posterior region of the superior temporal gyrus (superior temporal gyrus, STG), and the inferior parietal lobule (Inferior parietal lobule, IPL) (as discussed in Corbetta et al., 2008, The Reorienting System of the Human Brain: From Environment to Theory of Mind, Neuron 58, pages 306-324). The inferior parietal lobule includes the angular gyrus (AG) and the supramarginal gyrus (SMG).

[0111] Refer to the following Figure 1Explain the information flow of positive and negative bits that may occur between the dorsal attention network (DAN) and the ventral attention network (VAN) (e.g., within the Shannon brain model). As further shown above, the top-down contribution in the Shannon brain model can be given by the information content associated with the Bayesian likelihood, which can generate negative information bits that can offset the bits generated by the evidence (or surprise). Therefore, the TPJ (including, for example, the IPL and STG) can act as a memory for negative information bits. This is consistent with the role of the ventral attention network (VAN), which can be activated only when the stimulus is behaviorally relevant (e.g., one or more targets or accidental distractors) in a given context or episode (e.g., if the stimulus is informative, such as when TIC>0). Then, the information content associated with the Bayesian likelihood can be low or zero, so that the bottom-up bits (e.g., generated in the visual areas of the brain) that can be transmitted by the stimulus to the dorsal attention network (DAN) may not be offset by the top-down negative bits from the ventral attention network, and the stimulus can be prioritized and enhanced in one or more visual cortical areas. On the other hand, if the stimulus is uninformative (TIC≤0), the absolute value of the information content associated with the Bayesian likelihood can be quite high, so that the ventral attention network (VAN) can send a large number of negative bits to the dorsal attention network (DAN), which can offset the positive bottom-up bits of the stimulus. Then, non-behaviorally relevant stimuli (non-targets, distractors, etc.) may lose priority, so they may be covered.

[0112] This may apply to sensorimotor TIC, contextual TIC, and episodic TIC frameworks (e.g., due to the linearity of Eqs. 21 and 22, where the contributions of these TICs are taken into account). Thus, the TPJ may act as a memory for sensorimotor, contextual, and episodic negative information bits given by I(s|a), I(c|a,s), and I(e|a,s,c), respectively (e.g., top-down contributions of TICs). In the Shannon brain model discussed above, these three information contents associated with the Bayesian likelihood may be subtracted from the corresponding bottom-up contributions of TICs (i.e., I(s), I(c|s), and I(e|s,c), respectively) in priority map brain regions. Thus, sensorimotor TIC, contextual TIC, and episodic TIC may be generated, subtracted from the prior uncertainty of the action I(a), and the posterior uncertainty I(a|s,c,e) may be derived. As described above, different situations may arise depending on the signs of these TICs. For example, the contextual TIC may be negative, making the stimulus uninformative in context, but if the episodic TIC becomes positive, it may offset the contextual TIC, making the stimulus informative in this new temporal context. Figure 1 Example in.

[0113] like Figure 1 As shown, TPJ can act as a memory for sensorimotor, contextual, and episodic negative information bits given by I(s|a), I(c|a, s), and I(e|a, s, c), respectively. These three components can be subtracted from the sensorimotor contribution I(s), contextual contribution I(c|s), and episodic bottom-up contribution I(e|s, c), respectively, in one or more of the two priority map brain regions (e.g., in the Anterior Inferior Temporal Cortex (AIT) and / or in the Intraparietal Sulcus (IPS) / Superior Parietal Lobule (SPL)). Therefore, sensorimotor TIC, contextual TIC, and episodic TIC can be generated, subtracted from the prior uncertainty of the action I(a), and the posterior uncertainty I(a|s, c, e) can be derived. The dorsal attention network (DAN) and the ventral attention network (VAN) can utilize the saliency network (SN) ( Figure 1 Communicate with the participation of others (not shown).

[0114] Not only can the VAN send signals to the DAN in the activated state (one or more bottom-up signals, e.g. for redirection), but it can also send signals to the DAN in the deactivated state (the negative bit of one or more top-down signals). Thus, there may be information flow from the VAN to the DAN not only in the activated state, but also in the deactivated state. This may be counterintuitive, as deactivation and information flow may be thought of as opposing concepts, but on the other hand, it may help to understand the equally counterintuitive concept of negative information, as this may be the type of information expected from a deactivated source.

[0115] In addition, it is known in the prior art that there is a color-biased network (CBN) located within the Visual Ventral Pathway (VVP), which may include the bilateral lingual gyrus, the fusiform gyrus, the right inferior occipital gyrus, and one or more adjacent occipito-temporo-parietal regions. The posterior regions of this network may be passive and may be responsible for processing low-level color information (e.g., color perception), while the anterior regions may be active and responsible for encoding more complex functions (e.g., color knowledge), such as combining shape and color for object recognition, storing the typical color of a given object in memory and associating it with the object, or storing the color and shape of a given behaviorally relevant stimulus in memory. Because this network is related to memory, it may be connected to hippocampus-brain structures. These associations are called object-color memories. Shape and color may be processed separately in many areas of the VVP, and indeed a shape-biased lateral occipital lobe (LO) network is known to exist. These two networks may converge in the most anterior color-biased areas (color knowledge), such as in the anterior inferior temporal cortex (AIT), where shape and color may be combined for object recognition.

[0116] As is the case with other networks such as the dorsal attention network (DAN) and the ventral attention network (VAN), the cognitive functions of the color bias network (CBN) (color perception, color naming, color knowledge) can be connected through one or more bottom-up (e.g., feedforward) back-to-front pathways, and they can also be modulated top-down (e.g., feedback) through one or more front-to-back pathways. Therefore, irrelevant stimuli in the early visual cortex can be suppressed by top-down modulation based on the encoding of object color knowledge. The top-down modulation of the color bias network (CBN) and the shape bias network (SBN), as described above for the dorsal attention network (DAN) and the ventral attention network (VAN), may come from filtering signals originating in the prefrontal cortex (PFC) and stored in the TPJ of the ventral attention network (VAN), whose activation may be limited to relevant object color knowledge. The ventral attention network (VAN) can then send this top-down signal to the AIT to regulate the expression of preferential object color knowledge by enhancing or inhibiting the influence of bottom-up inputs from the early visual cortex.

[0117] Similarly, the sensorimotor TIC, contextual TIC, and episodic TIC frameworks can find a natural place in this top-down regulatory role of the ventral attention network (VAN) with the color bias network (CBN) and shape bias network (SBN). The TPJ can store negative bits of object color knowledge corresponding to the sensorimotor, contextual, and episodic information content associated with the Bayesian likelihood in Equation 22, namely I(s|a), I(c|a,s), and I(e|a,s,c), respectively. Each of these information contents can then be subtracted from their respective bottom-up components I(s), I(c|s), and I(e|s,c) in the AIT to regulate the expression of priority object color knowledge through the corresponding sensorimotor, contextual, and episodic computations in Equation 21, also as Figure 1 As shown. If the color of the stimulus is perceived as temporally contextual or episodic, this may make it more or less behaviorally relevant, which would make the sign of the episodic TIC a contributor to increasing or decreasing uncertainty about the action. Thus, the ventral attention network (VAN) may top-down modulate (e.g., inhibit or redirect) not only the dorsal attention network (DAN), but also the color biased network (CBN) and shape biased network (SBN).

[0118] There may be anterior-posterior communication between the dorsal attention network (DAN) and the ventral attention network (VAN). In particular, frontal regions such as the prefrontal cortex (PFC) may be candidates for such interaction. In one example, the inferior frontal gyrus (IFG) and the middle frontal gyrus (MFG) may be candidates for such interaction. In addition, such interaction may be mediated by different networks (the salience network (SN; in Figure 1 The SN (not shown) is driven based on the sustained activation of its two nodes, the anterior cingulate cortex (ACC) and the anterior insula (AI) during top-down filtering or redirection. The same interaction can be applied to the shape bias network (SBN) and the color bias network (CBN) as well as the ventral attention network (VAN), that is, their communication can be driven by the SN through the IFG and MFG. During cognitive tasks, the SN may have a dual role of "set maintenance" and monitoring the decisions made to monitor the task goals. Therefore, the SN may be involved in the anterior-posterior interaction between the ventral attention network (VAN) and the other networks mentioned above.

[0119] like Figure 2 As shown, in addition to anterior-posterior communication, (direct) dorsal-ventral communication (DVAN) can be used in conjunction with the above Figure 1The front-back communication coexists in the context of the description of the dorsal-ventral attention network (DVAN). In the dorsal-ventral attention network (DVAN), the TPJ of the ventral attention network (VAN) can modulate (e.g., filter or redirect) any of the dorsal attention network (DAN), the ventral shape network (VVP), and the color bias network (CBN) top-down to regulate sensory input without involving PFC regions. Figure 2 In particular, in addition to the front-to-back communication discussed further above, Figure 2 A coexisting dorsal-ventral pathway (DVAN) is also illustrated. The dorsal-ventral attention network (DVAN) can include occipital visual areas, the shape biased network (SBN) and color biased network (CBN) of the VVP, two priority maps (IPS / SPL and AIT), and one or more of the posterior regions of the dorsal attention network and the ventral attention network. The TPJ of the ventral attention network can also be a node of the network, which can communicate dorsally and ventrally with the IPS and AIT priority maps, respectively. The network can be a fast track for processing top-down violations (or filtering irrelevant stimuli), which can calculate sensorimotor TICs, contextual TICs, and episodic TICs without the involvement of additional PFC resources. In particular, by using direct connections between IPS / SPL and TPJ and / or between TPJ and AIT (both in Figure 2 The information between the dorsal attention network (DAN) and the ventral attention network (VAN) can be exchanged without involving the MFG or IFG (e.g. Figure 2 ). These direct connections therefore ensure that information can be processed by the brain without the involvement of frontal regions.

[0120] Experimental Results

[0121] A sample of 28 young children was recruited for a magnetoencephalography (MEG) study, including 14 ADHD participants (age 10 years 5 months; SD = 1 year 11 months; females = 35.7%) and 14 control participants (age 10 years 7 months; SD = 1 year 8 months; females = 35.7%). The ADHD participants were diagnosed by a neurologist who evaluated them and issued a neurological report with a diagnosis of ADHD. The study was approved by the Ethics Committee of the Hospital Clinic San Carlos in Madrid, and informed consent was signed by each participant and their legal guardian before MEG recordings.

[0122] In one example, a 306-channel (102 magnetometers and 204 planar gradiometers) whole-head MEG Elekta Neuromag system is used to measure MEG data including brain signals. In one example, brain signals are measured using an online anti-aliasing bandpass filter between 0.1 Hz and 330 Hz and a sampling rate of 1000 Hz. In the example, a three-dimensional Fastrak digitizer (Polhemus, Colchester, Vermont) is also used to obtain the head shape of the participant, and three landmarks (nasion and left and right preauricular points) are used as reference points. In the example, four head position indicator (HPI) coils are recorded on the participant's scalp (forehead and mastoid area). In the example, two sets of bipolar electrodes are used to record eye blinks and heartbeats, respectively.

[0123] Data preprocessing was performed in several steps. In one example, to remove external noise from the raw data, the time extension of spatial signal separation (tSSS) method was applied, using a window length of 10 seconds and a correlation threshold of 0.90 as input parameters. In the example, automatic detection of eye, cardiac and muscle artifacts was performed, which was validated by MEG experts. Finally, independent component analysis was performed in one example to remove eye blinks and cardiac activity. In the example, the data were segmented into 1 second trials (300ms baseline and 700ms task-related data) according to task events, and trials marked as containing artifacts were discarded from subsequent analysis.

[0124] In one example, a Montreal Neurological Institute (MNI) template is used to create a uniform grid with 1 cm spacing between source locations within the cranial cavity of a participant, resulting in 2459 source locations located within the cranial cavity of the participant. In the example, these source locations are labeled according to the Automated Anatomical Labeling atlas (AAL), and only 1188 source locations labeled as one of the 76 cortical regions are maintained. In one example, the surface is linearly transformed together with the source model in the subject space to match the location of the participant within the MEG scanner, using the head shape as an aid. According to one example, the intracranial interface is combined with the above-mentioned source model and the sensor definition provided by the system to solve the forward model and construct the lead field based on the modified spherical solution. In the example, a linear constrained minimum variance (LCMV) beamformer is used as an inverse method, using the average of the single-trial covariance matrix of the entire segment, using the Tikhonov method and a lambda factor regularization of 10% of the average channel power, to reconstruct the source level activity. To compute common beamformer filters for magnetometers and gradiometers, in one example, the data and guidance field matrices are normalized channel by channel so that signal amplitudes of different channel types are comparable. In another example, the three-dimensional time series are projected on the spatial principal components to obtain a single source level time series for each source location.

[0125] According to the example, to calculate phase synchronization, source-level activity is calculated separately for each classical frequency band: Theta (4Hz to 8Hz), Alpha (8Hz to 12Hz), Low Beta (12Hz to 20Hz), High Beta (20Hz to 30Hz), and Low Gamma (30 to 45Hz). In the example, the band-specific data is filtered using an 1800-order FIR filter constructed using a Hamming window and applied in a two-pass process to remove any possible phase distortion. In one example, to avoid edge effects, 1 second of real data on each side is used as padding to filter the data. It can be understood that for each source location in the participant's cranial cavity, a corresponding time series can be obtained.

[0126] In one example, a correlation value (such as a functional connectivity (FC) analysis) is calculated for a pair of time series i and j, where each time series corresponds to a corresponding source location. In one embodiment, the FC analysis is calculated by a phase-locked value (PLV; e.g., under the assumption of phase synchronization). The PLV can be based on the distribution of instantaneous phase differences of two time-related signals (e.g., a pair of time series). If and are the instantaneous phases of time t and trial n of signals i and j, respectively (e.g., time series i and j measured by the MEG apparatus, which correspond to brain signals at corresponding source locations i and j within the participant's cranial cavity), then the value of the PLV can be given by:

[0127]

[0128] The PLV can take values ​​between 0 and 1, where 0 indicates absolute phase independence (such as low or no correlation) and 1 indicates absolute phase dependence (such as high correlation). In one example, the instantaneous phase of the signal can be estimated using the Hilbert analysis signal after bandpass filtering. In order to avoid edge effects on the instantaneous phase extraction, according to one example, the Hilbert analysis signal can be calculated using one second of real data on each side of the trial as fill. In one example, after stimulation, the PLV is calculated for each frequency band (theta, from 4 to 8 Hz; alpha, from 8 Hz to 12 Hz; low beta, from 12 Hz to 20 Hz; high beta, from 20 Hz to 30 Hz; and low gamma, from 30 Hz to 45 Hz) in a time window from 50 ms to 450 ms. i,j .

[0129] In one example, the PLV is calculated independently for each task (e.g., Task 1 and Task 2, as explained further below) and for each condition (Failable, Infeasible, and Red E Vowel, also explained further below). i，j In one example, the PLV is calculated for each pair of source locations i and j i，j , generating a 1188×1188 FC matrix (ie, i=1...1188, and j=1...1188). In another example, the regional level PLV i，j is calculated as the average PLV of all cortical sources contained in each of the 76 cortical regions of interest (ROIs), for example, based on the AAL atlas, to obtain a 76×76 whole-brain matrix (i.e., i=1...76, and j=1...76). Obviously, the matrix PLV i，j Other sizes of are possible. In another example, PLV is calculated for specific cortical networks such as the salience network (SN) and the overall, right, and left executive center networks (ECNs). i，j In this case, the PLV pairs constituting the ROIs of the respective networks are extracted, obtaining a subsequent matrix of 8×8 ROIs for SN and 14×14 ROIs for ECN (right / left 7×7 ROIs).

[0130] For the whole-brain analysis example (discussed above, where i=1...1188 and j=1...1188), a permutation-based T-test was used to compare the PLV values ​​for each pair of ROIs. T-test independent sample analysis can be used for group comparisons within each functional network and for each condition. For within-group comparisons, a related sample analysis can be performed between pairs of conditions of interest and between similar conditions in two tasks within each group and functional network. In the example, the obtained p-value can be corrected for multiple comparisons (the number of pairwise comparisons and the number of frequency bands) with a false discovery rate (FDR) of 0.1, thereby obtaining an FDR-corrected alpha threshold for each analysis. In the example, p-values ​​below this threshold can be reported as significant.

[0131] In one example, a go / no-go task (hereinafter referred to as Task 1) shows the participant one or more stimuli, such as one or more letters, on a task presentation device, and instructs the participant to press a button when a vowel appears on the device, and not to press a button when a consonant appears on the device. In this example, the letters may have different colors (such as red, green, blue, and white). In addition, an exception rule is added to the task, that is, when the color of the letter is red, the participant is instructed not to press the button. After completing this task, the participant may start a new task (hereinafter referred to as Task 2), which may add a temporal context or plot, such as when a red vowel letter "e" is shown, a button should be pressed (this is similar to an exception from an exception). The second task can measure cognitive flexibility and the ability to change patterns learned during the first task.

[0132]

[0133] Table 1: Task summary

[0134] Still in this example, the stimuli can be generated in the following four evenly distributed groups: feasible A = non-red vowel; infeasible A = consonant; feasible B = red letter 'e'; infeasible b = red vowel instead of 'e'. The probabilities derived from the evenly distributed stimulus groups may lead to the following prior probabilities of feasible and infeasible conditions for each task. Task 1: Feasible condition = feasible A → P (feasible) = 1 / 4; infeasible condition = feasible B + infeasible A + infeasible B → P (infeasible) = 3 / 4. Task 2: Feasible condition = feasible A + feasible B → P (feasible) = 1 / 2; infeasible condition = infeasible A + infeasible B → P (infeasible) = 1 / 2. Figure 3 The conditional probabilities of the various relevant features according to the above model are illustrated. When a probability approaches zero, the value 10 can be assumed -5 to calculate the associated information to avoid any infinities.

[0135] In Task 1, there may be no relevant context when the consonants are presented to the participants, so applying Equation 12 may result in:

[0136]

[0137] Therefore, in this sensorimotor situation, That is, there may be a negative transfer of information from the stimulus of the consonant appearance to the action a of pressing the button.

[0138] Furthermore, when participants were presented with non-red vowels under the conditions of Task 1, applying Equation 17 to the non-red context gave:

[0139]

[0140] This means that the uncertainty about the action a can be reduced to zero, and therefore the button can be pressed. In this calculation, it is assumed that the task can be performed without any errors. In real cases, the terms I(s|a) and I(c|a,s) will not be strictly zero, which can explain the errors that occur.

[0141] Furthermore, applying Equation 17 to the context of red when participants were presented with non-red vowels under the conditions of Task 1 gives:

[0142]

[0143] In this case, the component I(c|a, s) can generate negative information bits, which can offset the information bits generated by the evidence representing the top-down control exerted by the subject. To derive the value of the probability P(c|a, s), one can reason as follows: given that a vowel has occurred and the button has been pressed, what is the probability that the vowel is red. The answer is that this probability may be very low, and therefore its associated information content may be very high (e.g., a large number of negative information bits are generated to offset the positive bits generated by the evidence I(c|s). Therefore, in this context, there may be a negative transfer of information from the evidence of the occurrence of a red vowel to the action a of pressing the button.

[0144] Furthermore, in the episodic condition in Task 2, the button should be pressed when the red vowel “E” is presented to the participant, and in this case, no top-down control occurs, despite the fact that the same context as the last component may occur (the vowel is red). Applying Equation 22, the values ​​of the individual antecedent components can be the same as in the previous case, with the exception of I(a), which may be different in Task 2 because of the increase in pulsation frequency due to the new episodic condition), but two new terms may appear to take into account the episodic component information:

[0145]

[0146] Again, since the uncertainty about the action can be reduced to zero, the result is a decision to press the button. Therefore, in this scenario, That is, there may be a positive transfer of information from the evidence of the occurrence of the red vowel e to the action of pressing the button. Although the contextual TIC may be negative in this case, it can be compensated by the episodic TIC, which is positive and larger than the contextual TIC. There are multiple contributions to Equation 27:

[0147] Sensorimotor TIC: The subject sees a vowel, i.e., Shannon surprise I(s), with a probability of 3 / 4, i.e., 0.42 bits of positive information, which temporarily reduces the uncertainty to 0.58 bits; at this sensorimotor stage, the negative information generated by the other component of TIC, I(s|A), is zero and has no effect. In this case, the subject may have been informed in advance of the probability of the vowel, or he / she may have learned the probability of the vowel during the task.

[0148] Contextual TIC: Subjects can consider the context, in this case the color of the stimulus through the bottom-up component of the contextual TIC, I(c|s), which may again be a Shannon surprise of value 0.58 bits (probability of red vowels is 2 / 3). Subjects can then consider the top-down component (e.g., I(c|a,s)), which may introduce 1 bit of negative information, which may increase uncertainty about the button press (e.g., may reduce its probability).

[0149] Episodic TIC: So far, the 1-bit initial uncertainty has been reduced by exactly 1 bit (0, 42 + 0, 58) through the two Shannon surprises I(s) and I(c|s), but the contextual TIC has introduced 1 bit of negative information through its top-down component I(c|a, s), which has increased the uncertainty to 1 bit again. Therefore, when the subject has processed that the stimulus is a vowel and that it is red, he / she has 1 bit of uncertainty. If the stimulus is "e", the subject can wait to process it because this may be the episodic condition that should cause a button press in Task 2. The episodic TIC is 1 bit: Episodic TIC(e=red e→a|s,c)=I(e=red E|s=vowel,c=red)-I(e=red E|a,s=vowel,c=red)=1-0=1 bit. Therefore, as expected, the episodic TIC can be 1 bit, and when subtracted in Equation 27, the uncertainty can be reduced to 0 bits, thereby causing a button press.

[0150] MEG data measured from participants may include brain signals in the alpha, beta, and theta frequency bands. Figures 3 to 12The experimental results are discussed in the context of . On the left side of each of these figures, posterior, superior, left and right views of the brain are shown, as well as connections indicating significantly different PLV values. On the right side of each of these figures, a circular diagram shows a schematic diagram of important connections (corresponding to the diagram on the left). The labels are abbreviations of brain regions according to the automated anatomical labeling atlas shown at the end of the instructions.

[0151] In a first example, the PLV of a healthy condition (HC) group (eg, healthy participants) performing Task 1 under an infeasible condition (consonants or red vowels) can be compared to the PLV under a feasible condition (non-red vowels) in the alpha band (8 Hz to 12 Hz). Figure 4 The connections with significantly different PLV values ​​in the first example above are illustrated. It can be observed that under the infeasible condition, the occipito-temporo-parietal association cortex and the ventral prefrontal cortex are deactivated. In particular, Figure 4 The connections with significantly different PLV values ​​between the infeasible and feasible conditions when the HC group performs Task 1 are illustrated, represented by lines. On the left, the posterior, superior, left, and right views of the brain are shown. On the right, a circular diagram is shown, showing a schematic diagram of the significant connections. The lines represent the reduction in PLV values ​​in the infeasible condition compared to the feasible condition. It can be seen that there are connections with significantly different PLV values ​​between the left superior temporal gyrus (lSMG; belonging to the temporoparietal junction) and the left inferior temporal gyrus (lITG, belonging to the VVP region). This is consistent with the Shannon brain model discussed above, in which the ventral attention network (VAN) may send such top-down signals to the VVP network to regulate the expression of priority object knowledge by enhancing or inhibiting the influence of bottom-up inputs from the early visual cortex.

[0152] In this case, under the feasible condition (no red vowel), Equation 25 shows that the two information contents associated with the Bayesian likelihood, I(s=vowel|a) and I(c=non-red|a, s=vowel), are zero. In contrast, under the infeasible condition, if the stimulus is a consonant, Equation 24 shows that the information content associated with the Bayesian likelihood I(s=consonant|a) is large and the resulting sensorimotor TIC is negative; if the stimulus is a red vowel, Equation 26 shows that the contextual TIC is negative due to the non-zero value of I(c=red|a, s=vowel). Under the infeasible condition, the ventral attention network can be deactivated by these negative information contents associated with the Bayesian likelihood. Furthermore, when these negative top-down bits are subtracted from the bottom-up bits of Shannon surprise corresponding to the stimulus, a negative TIC can be generated in the anterior inferior temporal lobe, which deactivates the VVP and visual areas in the occipital cortex. This is in Figure 4As can be observed in Figure 2, in addition to the connection between the left supramarginal gyrus of the TPJ and the left inferior temporal cortex of the VVP, there are connections between temporal and occipital regions, showing the mentioned deactivation of all these regions in the infeasible condition, which can be described by a negative sensorimotor TIC in case the stimulus is a consonant or a negative contextual TIC in case the stimulus is a red vowel. Moreover, the connection between the IFG and the middle temporal gyrus could indicate an interaction between the ventral attention network and the shape network driven by the SN.

[0153] In the second example, when the stimulus was a red vowel e, the PLV in the alpha band (8 Hz to 12 Hz) in the HC group performing Task 1 under the infeasible condition was compared with Task 2 under the feasible episode condition (red vowel e). Figure 5 The connections with significantly different PLV values ​​for this second example are shown. It can be observed that in the Task 1 infeasible condition (red vowel e), the occipito-temporal parietal cortex is deactivated. In particular, Figure 5 The connections with significantly different PLV values ​​in the infeasible condition of Task 1 (red vowel e) relative to the episodic feasible condition of Task 2 (red vowel e) for the HC group are illustrated as lines. The brain is shown in the left posterior view, top view, left view and right view. The circular diagram is shown on the right side, showing a schematic diagram of the significant connections. The line indicates the reduction of PLV value in the infeasible condition of Task 1 (red vowel e) relative to the episodic feasible condition of Task 2 (red vowel e). For completeness, as shown in Figure 5 As shown, for this second example, the connections with significantly different PLV values ​​are repeated in the following table:

[0154]

[0155]

[0156]

[0157]

[0158] Table 2: Figure 5 The connections shown in

[0159] The PLV results are also consistent with the Shannon brain model discussed above. In the case of the infeasibility condition in Task 1, the information content associated with the Bayesian likelihood I(c=red|a, s=vowel) can generate negative bits, which can be stored in the TPJ of the ventral attention network, which can then be sent to the priority mapping area to compensate for the bottom-up bits from the input signal. In the case of red vowels, the Shannon brain model discussed above can predict that the negative information bits corresponding to the term I(c=red|a, s=vowel) can be stored in the TPJ of the ventral attention network, and when the red vowel is presented, they can be sent to inhibit the color-biased network. This can be seen in the Figure 5 It was observed that there were connections with significantly different PLV values ​​between TPJ regions (e.g., left superior marginal gyrus, bilateral angular gyrus) and color-biased network regions (e.g., bilateral lingual gyrus and fusiform gyrus, and right inferior occipital gyrus).

[0160] Furthermore, there are connections between the superior parietal gyrus (an area belonging to the dorsal attention network, where the priority map has been identified) and occipital visual areas, which also belong to the dorsal attention network. These connections confirm that top-down modulation of occipital visual areas by higher-level areas also occurs in the dorsal attention network. In Task 2, in contrast to the deactivation of the color-biased network in Task 1, as shown in Equation 32, the uncertainty about action a can be fully offset by the episodic TIC, which can strengthen the color-biased network. Therefore, in Task 2, the color-biased network can be strengthened. Figure 5 The differences in PLV values ​​observed in may be due not only to the deactivation of the color-biased network in Task 1, but also to the enhancement of this network in Task 2. Figure 5 This suggests that the reconfiguration of mental resources during task switching, suggested by the different processing of the red vowel e in Task 1 and Task 2, may be achieved without involving additional prefrontal cortex (PFC) recruitment. Figure 2 The hypothetical dorsal-ventral attention network exemplified in (in which the apical descending temporoparietal junction (TPJ) of the ventral attention network may modulate (e.g., filter or redirect)) modulates not only the dorsal attention network but also the ventral shape and color networks, thereby regulating sensory input, without involving prefrontal cortex (PFC) regions.

[0161] In the third example, if Figure 6 As shown, when the stimulus was the red vowel e, the PLV in the alpha band (8 Hz to 12 Hz) in the ADHD group performing Task 1 under the infeasible condition was compared with Task 2 under the feasible condition (red vowel e). Figure 6, the central executive network (CEN) of the ADHD group (e.g., participants with an ADHD diagnosis) is shown. Connections with significantly different PLV values ​​in the infeasible condition of Task 1 (red vowel e) relative to the episodic feasible condition of Task 2 (red vowel e) are shown as lines. The left, back, top, left, and right views of the brain are shown. A circular diagram is shown on the right, showing a schematic diagram of the significant connections. The line indicates that the PLV value was reduced in the infeasible condition of Task 1 (red vowel e) compared to the episodic feasible condition of Task 2 (red vowel e). In Figure 6 , connections with significantly different FC values ​​can be seen. In contrast to the HC group, in the ADHD group, the reconfiguration of mental resources for task switching, suggested by the different processing of the red vowel e in Task 1 and Task 2, can be achieved by involving additional PFC recruitment. For example, the regions of the PFC involved are the IFG (e.g., left inferior frontal triangle and opercular gyrus) and the right middle frontal gyrus, showing communication between the dorsal attention network and the ventral attention network through the SN. There are three connections between these three regions of the PFC and the nodes of the dorsal attention network, which are the superior parietal gyrus (e.g., left superior parietal gyrus; lSPG). In addition, direct dorsal-ventral communication can be observed in the connection between the left superior parietal gyrus and the right angular gyrus of the ventral attention network.

[0162] It can be considered that the above Figure 2 The discussed dorsal-ventral attention network (DVAN) (where the ventral attention network (VAN) may have the role of communicating dorsally and ventrally with the dorsal attention network (DAN) and VVP, respectively) may represent a direct pathway for filtering irrelevant stimuli or redirecting behaviorally relevant stimuli, in contrast to an anterior-posterior pathway that may involve the prefrontal cortex (PFC) and may be driven by the substantia nigra (SN) via the PFC (e.g., via the inferior frontal gyrus (IFG) and middle frontal gyrus (MFG); e.g., Figure 1 Thus, subjects performing Task 1 and Task 2 may show the inefficiency of subjects with cognitive conditions (e.g., ADHD) because such subjects may need increased activity in PFC pathways when processing task switching (e.g., with changes in temporal context or events, such as exceptions to conditions). The HC group was able to perform task switching without increasing activity in PFC pathways by increasing activity in the dorsal-ventral network. It can be argued that Figure 1 may represent ADHD brain connectivity in the reconfiguration of mental resources due to task switching, and Figure 2 It can represent HC brain connectivity.

[0163] In the fourth example, the PLV values ​​in the feasible condition for both Task 1 and Task 2 were compared. In Task 1, the feasible condition (non-red vowels), the connectivity results showed an increase in PLV values ​​in the alpha band (8 Hz to 12 Hz) in the HC group relative to the ADHD group. Therefore, the behaviorally relevant stimulus, the non-red vowel, can be modulated top-down, and at the same time, the irrelevant stimulus (consonant and red vowel) can be suppressed. Figure 7 The connections of the central executive network are shown, where the HC group has significantly different PLV values ​​relative to the ADHD group in the alpha band (8Hz to 12Hz) under the feasible condition of Task 1 (non-red vowels). The left view, back view, top view, left view, and right view of the brain are shown. The right side shows a circular diagram showing a schematic diagram of the significant connections. The line indicates that the HC group has increased PLV values ​​relative to the ADHD group under the feasible condition of Task 1 (non-red vowels). In particular, Figure 7 Connections belonging to the central executive network (CEN) are shown, which include the dorsal attention network and the ventral attention network. It can be seen that the PLV values ​​of HC subjects increased in areas of the dorsal attention network (e.g., superior parietal gyrus), the ventral attention network (e.g., bilateral angular gyrus), the IFG (e.g., eye and inferior triangular frontal gyrus), and the connections between the middle frontal gyrus compared to the ADHD group. These connections demonstrate the interaction between the dorsal attention network and the ventral attention network through the SN (for the middle frontal gyrus, the connection between the two). This may suggest that the ventral attention network in the HC group is sending stronger bottom-up redirection signals to the dorsal attention network compared to the ADHD group.

[0164] also, Figure 8 The connections with significantly different (i.e., increased) PLV values ​​between the HC and ADHD groups in the feasible condition (non-red vowels), alpha band (8Hz to 12Hz) are shown. The network shown is the salience network (SN), also known as the cingulate-opercular network. The increase in PLV values ​​of the HC relative to the ADHD group can be observed in the connections between most nodes of the cingulate-opercular network (e.g., bilateral insula, bilateral inferior triangular frontal gyrus, bilateral infraorbital frontal gyrus, left inferior opercular frontal gyrus, and right anterior cingulate cortex). The better cognitive performance of the HC group relative to the ADHD group in Task 1 can also be explained by this increase in PLV values ​​in the SN, which can have the effect of providing stable "set retention" throughout the task period and monitoring the achievement of desired goals. In particular, Figure 8The connections of the significant network are shown, where the HC group has significantly different PLV values ​​under the feasible conditions (non-red vowels) of Task 1 relative to the ADHD group in the alpha band (8Hz to 12Hz). The left view, back view, top view, left view and right view of the brain are shown. The right side shows a circular diagram showing a schematic diagram of the significant connections. The line indicates that the HC group has an increased PLV value under the feasible conditions (non-red vowels) of Task 1 relative to the ADHD group.

[0165] In a fifth example, the connections in the beta band (20 Hz to 30 Hz) with significantly reduced PLV values ​​under the feasible conditions (non-red vowels and red vowel e) of Task 2 for the HC group were compared with the ADHD group. In Task 2, under the feasible conditions, vowel e can be an episodic top-down violation that may require behavioral flexibility, which can be obtained by a reduction in the beta band. Therefore, the high beta band synchronization of the ADHD group can explain the performance deficit of this group in Task 2, which is shown as a lower percentage of correct hits (86.23% in ADHD vs. 92.63% in HC, p-value = 0.009) in both non-red vowels (89.93% in ADHD vs. 95.00% in HC, p-value = 0.017) and red vowel e (82.53% in ADHD vs. 90.27% in HC, p-value = 0.022). Fig. 9 The connections of the central executive network (CEN) are shown, where the HC group had significantly different PLV values ​​under the feasible condition of Task 2 relative to the ADHD group in the beta frequency band (20Hz to 30Hz). The left view, back view, top view, left view, and right view of the brain are shown. The right side shows a circular graph showing a schematic diagram of the significant connections. The line indicates that the PLV value under the feasible condition of Task 2 of the HC group was reduced relative to the ADHD group. From Fig. 9 The PLV connections in the HC group suggest that the HC group processed more cognitive flexibility (less beta-band synchronization) using dorsal-ventral connections between the right superior parietal gyrus of the ventral attention network and the right superior parietal gyrus of the dorsal attention network. However, there may be other connections connecting regions of the dorsal attention network (e.g., bilateral superior parietal gyrus) to the IFG and middle frontal gyrus, indicating the involvement of the SN.

[0166] In the sixth example, in the theta frequency band (4 Hz to 8 Hz), the connections with significantly different (i.e., increased) PLV values ​​in the feasible condition of Task 2 in the HC group were compared with the ADHD group. Increased PLV values ​​were observed between regions of the dorsal attention network (e.g., bilateral superior parietal lobule) and regions of the ventral attention network (e.g., left superior marginal gyrus, right angular gyrus, and inferior parietal gyrus). Two connections were shown between the inferior parietal gyrus and bilateral middle frontal gyrus of the ventral attention network. Fig.10The connections of the central executive network are shown, where the HC group has significantly different PLV values ​​relative to the ADHD group under the feasible conditions of Task 2 in the theta frequency band (4Hz to 8Hz). The left view, back view, top view, left view, and right view of the brain are shown. The right side shows a circular diagram showing a schematic diagram of the significant connections. The line indicates that the HC group has increased PLV values ​​relative to the ADHD group under the feasible conditions of Task 2.

[0167] In the seventh example, the PLV values ​​in the infeasible conditions of both Task 1 and Task 2 are considered. Fig.11 The connection between the HC and ADHD groups with significantly different (i.e., increased) PLV values ​​in the alpha band under the infeasible condition (consonant or red vowel) is shown. As in the previous case, the network shown is the SN. The increase in PLV values ​​of the HC relative to the ADHD group can be observed in the connections between the nodes of the SN (e.g., left insula, bilateral infraorbital frontal gyrus, bilateral anterior cingulate cortex). Under this infeasible condition, the negative information bits stored in the TPJ can be sent to the dorsal attention network through the SN. Therefore, the increased alpha band PLV values ​​of the HC group may mean better inhibition of irrelevant stimuli (e.g., better management of negative information). This can be observed as a higher percentage of hits in the infeasible condition of the ADHD group (6.04% in ADHD vs 4.04% in HC, p value = 0.031). Fig.11 The connections of the significant network are shown, where the HC group has significantly different PLV values ​​under the infeasible conditions (consonants or red vowels) of Task 1 relative to the ADHD group in the alpha band (8Hz to 12Hz). On the left, the posterior view, upper view, left view, and right view of the brain are shown. The right side shows a circular diagram showing a schematic diagram of the significant connections. The line indicates that the HC group has an increased PLV value under the infeasible conditions (consonants or red vowels) of Task 1 relative to the ADHD group.

[0168] also, Fig.12 Connections in the beta band (20 Hz to 30 Hz) were shown, where PLV values ​​were significantly reduced in the infeasible condition (consonant and red vowel) of Task 1 in the HC group relative to the ADHD group. In Task 1, the red vowel in the infeasible condition may represent a top-down violation, where a reduction in the beta band may be desirable to provide cognitive flexibility. This may explain the better performance of the HC group relative to the ADHD group in Task 1. Again, these connections seem to indicate that the HC group processes more cognitive flexibility (synchronous reduction in the beta band) via a direct dorsal-ventral pathway and via the SN through interactions between regions of the ventral attention network (right angular gyrus, right superior marginal gyrus) and the dorsal attention network (right superior parietal gyrus) with the involvement of IFG and middle frontal gyrus regions. In particular, Fig.12The connections of the central executive network are shown, where the HC group has significantly different PLV values ​​relative to the ADHD group in the beta frequency band (20 Hz to 30 Hz) under the infeasible condition (consonant or red vowel) of Task 1. The left view, back view, top view, left view, and right view of the brain are shown. The right side shows a circular diagram showing a schematic diagram of the significant connections. The line indicates that the HC group has a reduced PLV value relative to the ADHD group under the infeasible condition (consonant or red vowel) of Task 1.

[0169] As mentioned above, Bayes' theorem can be expressed in terms of the Shannon information content. By applying a logarithmic function on both sides of Bayes' theorem, a Shannon metric can be defined, called the transfer information content (TIC), which can measure the information conveyed by a stimulus to reduce (in the case of informative stimuli or TIC>0) or increase (in the case of uninformative stimuli or TIC<0) the uncertainty about the action selection. The informative or non-informative characteristics of a stimulus can depend on its behavioral relevance, for example, in sensorimotor cognitive tasks, as discussed above. The TIC can be a function of the Shannon information content and at the same time can allow Bayesian model updating by measuring the increase or decrease in uncertainty about the action uncertainty (e.g., it can reconcile Shannon and Bayesian surprise).

[0170] The PLV values ​​of the HC group were significantly reduced in the alpha band (8 Hz to 12 Hz) under the infeasible condition (consonant or red vowel) relative to the PLV values ​​under the feasible condition (non-red vowel) of Task 1, possibly indicating that the ventral attention network was deactivated by the negative information content associated with Bayesian likelihood. In addition, when these negative top-down bits were subtracted from the bottom-up bits corresponding to the Shannon surprise of the stimulus, negative TICs could be generated in the anterior inferior temporal region, which deactivated the VVP and visual areas in the occipital cortex. Interestingly, one of the regions with significantly different FC values ​​between the infeasible and feasible conditions was the left fusiform gyrus, which responded significantly more to letters than to numbers. This is consistent with the nature of Task 1, in which the stimuli were letters.

[0171] Furthermore, the HC group PLV values ​​were also compared in the alpha band (8 Hz to 12 Hz) in the HC group that performed Task 1 in the infeasible condition relative to the HC group that performed Task 2 in the feasible interlude condition (red vowel e) when the stimulus was the red vowel e. It can be observed that in the Task 1 infeasible condition (red vowel e), there was a deactivation of the color-biased network relative to the same stimulus in Task 2, where the uncertainty about the action a was completely offset by the episodic TIC, allowing the enhancement of the mentioned color-biased network. It can be observed that the HC group could activate not only the color-biased network but also the entire dorsal-ventral attention network discussed above to manage the top-down violation implied by the red vowel e in Task 2.

[0172] The above results may indicate inefficiency in ADHD subjects because they need to increase activity in the PFC when processing task switching (e.g., exceptions to conditions), in this case, changes in temporal context or events. In contrast, the HC group can perform in task switching by increasing the activity of the new temporoparietal dorsal-ventral attention network without increasing the activity of the PFC pathway. In the feasible and infeasible conditions in the alpha band, reduced PLV values ​​were observed in the ADHD group relative to the HC group in connections that may indicate the interaction between the dorsal and ventral attention networks through the SN. Therefore, the ventral attention network in the HC group may send stronger redirection (feasible condition) or filtering (infeasible condition) signals to the dorsal attention network than in the ADHD group. In addition, again in connections that show the interaction between the dorsal and ventral attention networks through the SN, in the beta band (20 Hz to 30 Hz), a significant increase in PLV values ​​could be observed in the ADHD group relative to the HC group under the infeasible condition (red vowel) in Task 1 and the feasible condition (red vowel e) in Task 2. These two conditions may imply a top-down violation that requires a certain cognitive flexibility (e.g., low beta band synchronization).

[0173] Systems, computer-readable media, and methods for generating a score indicative of cognitive status

[0174] In view of the above considerations and experiments, it has been recognized that a system, a computer-readable medium, and a method can be provided that can be suitable for generating a score indicating a subject (e.g., a person) having a cognitive condition (e.g., a cognitive disorder, such as ADHD). In particular, the above results show that information in the brain can be transmitted through different pathways (or networks) when the subject suffers from a cognitive condition (e.g., a cognitive disorder, such as ADHD) compared to a healthy subject. A pathway may include two locations (or nodes) in the brain, between which information is transmitted. Therefore, it is suggested to find one or more pathways (e.g., pathways that are active for healthy subjects and inactive or less active for subjects with cognitive conditions, and vice versa) that distinguish healthy subjects from subjects with cognitive conditions (e.g., ADHD) to generate a score indicating a subject with such a cognitive condition. According to one embodiment, the information transmitted via a specific pathway can correspond to a correlation value (such as the PLV discussed above) between two time series measured at corresponding two source locations within the cranial cavity of the subject (the time series is obtained from an EEG or MEG device). The two source locations can be two locations of a pathway that distinguishes healthy subjects from subjects with conditions as described above. As further shown above, subjects with cognitive conditions (e.g., cognitive disorders such as ADHD) may lack the ability to increase activity in the DVAN without increasing activity in PFC pathways (due to the DVAN discussed above, e.g., Figure 1 and 2 as well as Figure 5 and Figure 6 As shown, healthy subjects may have this ability).

[0175] This lack of ability to increase activity in the dorsal-ventral network (DVAN) may be present not only in subjects with ADHD, but may also be present in subjects with other cognitive conditions (e.g., conditions that may be associated with subjects having difficulty staying focused or having difficulty switching between mental tasks). Therefore, scores generated according to the present disclosure may also be indicative of such cognitive conditions and may help physicians diagnose subjects with such conditions. In addition, scores generated according to the present disclosure may be used when assessing whether a subject is able to work in certain occupations where staying focused or switching between mental tasks is important (e.g., pilots, truck drivers, air traffic controllers, etc.).

[0176] According to a first embodiment, a correlation value between a pair of time series of source locations within the cranial cavity of a subject (i.e., in the subject's brain) included in a PFC pathway can be used to generate a score indicating that the subject has a cognitive condition (such as a cognitive disorder, such as ADHD). As shown above, a subject with such a cognitive condition may have increased activity in the PFC pathway (e.g., compared to a subject without the cognitive condition). According to one embodiment, the PFC pathway may include the frontal lobe (e.g., the first source location of the pathway) and the superior parietal lobe (e.g., the second source location of the pathway) of the brain. In this way, a correlation value (which corresponds to information transfer) can be calculated based on a pair of time series including a first time series measured at the frontal lobe of the subject and a second time series measured at the superior parietal lobe (SPL) of the subject. In one embodiment, the frontal lobe (i.e., the first source location) may include at least one of the middle frontal gyrus (MFG) and the inferior frontal gyrus (IFG).

[0177] In a second embodiment, a correlation value between a time series of a source location (inside the cranial cavity of the subject) as part of a ventral attention network and another time series of a source location (inside the cranial cavity of the subject) as part of a dorsal attention network can be used to generate a score indicating that the subject has a cognitive condition such as a cognitive disorder (such as ADHD). As shown and discussed above, a healthy subject may have the ability to (directly) transfer information between these two source locations (i.e., a healthy subject may use the dorsal-ventral network (DVAN) discussed above), while a subject with a cognitive condition such as ADHD may lack this ability. In this way, one or more correlation values ​​between a pair of time series between these two source locations may be suitable for generating a score indicating a cognitive condition (e.g., a cognitive disorder such as ADHD). In one embodiment, the location that is part of the ventral attention network may be the temporoparietal junction (TPJ), and the location that is part of the dorsal attention network may be the superior parietal lobe (SPL). As described above, the TPJ may include the posterior region of the superior temporal sulcus (STS), the posterior region of the superior temporal gyrus (STG), and the inferior parietal lobule (IPL). The inferior parietal lobule includes the angular gyrus (AG) and the supramarginal gyrus (SMG).

[0178] In a third embodiment, a correlation value between a time series of a source location (inside the subject's cranial cavity) that is part of the ventral attention network and a location that is part of the ventral visual pathway (VVP) can be used to generate a score indicating that the subject has a cognitive condition (such as a cognitive disorder, such as ADHD). As shown and discussed above, healthy subjects may have the ability to transfer information between these two locations, while subjects with cognitive conditions (such as cognitive disorders such as ADHD) may lack this ability. In this way, the correlation value between a pair of time series at these two locations can be suitable for generating a score indicating a cognitive condition (such as a cognitive disorder, such as ADHD). In one embodiment, the location that is part of the ventral attention network can be the TPJ, and the location that is part of the ventral visual pathway can be the inferior temporal gyrus.

[0179] Obviously, not only a single correlation value can be used to generate a score. Instead, multiple correlation values ​​(such as multiple correlation values ​​from the first embodiment, the second embodiment, or the third embodiment discussed above) can be used. In addition, the first to third embodiments described above can be combined so that one or more correlation values ​​obtained according to one of these embodiments can be used together with one or more correlation values ​​obtained from another or two of these embodiments to generate a score indicating that the subject suffers from a cognitive condition (e.g., a cognitive disorder, such as ADHD).

[0180] Fig.13 An example system 100 is shown that can be used to generate a score 150 indicating that a subject 110 has a cognitive condition, such as a cognitive disorder, such as ADHD. The system 100 may include an electroencephalogram (EEG) or magnetoencephalogram (MEG) device 120 for measuring EEG or MEG data 130, and a data analysis device 140 for generating the score 150. In one embodiment, the system 100 may additionally include a task presentation device 160 for presenting one or more tasks to the subject 110 and an input device 170 for receiving input from the subject 110 as a response of the subject to the presented tasks. In yet another additional embodiment, the system 100 may also include a score presentation device 180 for presenting the score 150. In one embodiment, the score 150 is presented to a physician 190, who may diagnose whether the subject 110 has a cognitive condition, such as a cognitive disorder, such as ADHD.

[0181] The EEG / MEG device 120 may be a standard EEG or MEG device known in the prior art. The EEG / MEG device 120 may measure multiple signals at various source locations within the cranial cavity of the subject 110. Each source location may correspond to a known brain region, such as a lobe, lobule, sulcus, or gyrus. In some examples, multiple signals may be measured while the subject is performing one or more tasks that may be presented to the subject 110 on the task presentation device 160. The signal measured by the EEG / MEG device 120 may be divided into one or more frequency bands based on the frequency range of the signal. For example, the signal may include (among other things) an alpha band (e.g., in a frequency range of 8 Hz to 12 Hz), a beta band (e.g., in a frequency range of 12 Hz to 30 Hz), and / or a theta band (e.g., in a frequency range of 4 Hz to 8 Hz). In one or more embodiments, the signal may correspond to the measurement results of the electrodes of the EEG device or the measurement results of the sensor coils of the MEG device, respectively. In some embodiments, the signal may correspond to an aggregate signal (such as an average signal) over multiple measurements from multiple electrodes (in the case of EEG) or multiple sensor coils (in the case of MEG). Such an aggregate signal may correspond to a region of the brain that includes the corresponding locations of the signals aggregated in the aggregate signal.

[0182] The data analysis device 140 is configured to receive the EEG / MEG data 130 and generate a score 150 indicating that the subject 110 has a cognitive condition, such as a cognitive impairment, such as ADHD. The data analysis device 140 can be located at the same location (e.g., on-site) as the EEG / MEG device 120 (e.g., the data analysis device can be a computing system located at a hospital where the EEG / MEG device 120 is located), or the data analysis device 120 can be located elsewhere (e.g., off-site, such as in the cloud) and can communicate with the EEG / MEG device 140, the input device 170, and / or the score presentation device 180 via a network (e.g., the Internet). Further details about the data analysis device 140 will be described below with respect to Figure 2 Further discussion.

[0183] When the EEG / MEG device 120 is measuring brain signals of the subject 110, the task presentation device 160 may present one or more tasks to the subject 110. In one embodiment, the task presentation device 160 may present the first task during a first time period, during which the MEG / EEG device 120 is measuring brain signals of the subject 110. The first task may include three conditions: a first condition, a second condition, and an exception to the second condition (an exception condition). Each condition may include one or more stimuli s and a corresponding action a to be performed by the subject 110 when the stimulus in the one or more stimuli is presented to the subject 110. In an example, the subject 110 may be instructed: when the first condition is presented during the first task, not to interact with the input device 170 (such as by not pressing a button); when the second condition is presented during the first task, to interact with the input device 170 (such as by pressing a button); and when the exception to the second condition is presented, not to interact with the input device 170 (such as by pressing a button). In one embodiment, the first task may be the above (in Figure 3 ) and as shown below in Figures 18 and Fig. 20 Task 1 example is described in further detail in the context of .

[0184] Furthermore, the task presentation device 160 may present the second task during the second time period (the second time period being after the first time period), during which the MEG / EEG device 120 (still or again) measures the brain signals of the subject 110. In one embodiment, the second task may include the three conditions of the first task as discussed above and an additional fourth condition as an exception to the exception to the second condition, so that when the fourth condition occurs during the second task, the subject 110 may be instructed to interact with the input device. In an example, the second task may include a first condition and a second condition, for the first condition, the subject 110 is instructed not to interact with the input device 170, and for the second condition, the subject 110 is instructed to interact with the input device 170. Furthermore, the second task may include a third condition, which defines an exception to the second condition, for which the subject 110 is instructed not to interact with the input device 170. Furthermore, the second task may include a fourth condition, which defines an exception to the exception to the second condition, for which the subject 110 is instructed to interact with the input device 170. In one embodiment, the second task may be the Task 2 example discussed further above. The task presentation device 160 may be any device suitable for presenting one or more tasks, such as a screen (eg, a screen of a tablet or computing device).

[0185] The different conditions of the first task and the second task can occur multiple times during the first time period and the second time period, respectively, such as periodically (e.g., every five seconds). In addition, different conditions can occur at a predefined frequency for each of the different conditions (e.g., each condition can occur at the same frequency, 1 / 3 in the first task discussed above or 1 / 4 in the second task discussed above). The multiple times at which the different conditions are presented by the task presentation device 160 can be matched with the brain signals measured by the MEG / EEG device 120, so that the brain's response to the corresponding different conditions can be identified. The different conditions (i.e., the three conditions of the first task discussed above and the four conditions of the second task) can each correspond to a corresponding stimulus for the brain of the subject 110.

[0186] The input device 170 may be any device capable of receiving input from the subject 110 via the subject 110's interaction with the input device 170. In one embodiment, the input device 170 may include a button that can be pressed by the subject 110. In another embodiment, the input device may detect a gesture or facial expression of the subject 110, which may correspond to the subject 110's interaction with the input device 170. The input of the subject 110 received by the input device 170 may be used to determine whether the subject 110's interaction with the input device 170 is a correct (or incorrect) action in response to the stimulus of the presented conditions of the task presented by the task presentation device 160. As described above, the subject 110 may be instructed to interact with the input device 170 under certain conditions (e.g., the second condition and the exception of the exception of the second condition), and not to interact with the input device under other conditions (e.g., the first condition and the exception of the second condition). The input of the subject 110 may be used to determine whether the subject 110 interacts with the input device 170 (correct response) or does not interact with the input device 170 (incorrect response) according to these instructions.

[0187] The score presentation device 180 may be any device capable of presenting one or more scores 150 to a physician 190, who may diagnose whether the subject 110 suffers from a cognitive condition, such as a cognitive disorder, such as ADHD. In particular, the one or more scores 150 presented by the score presentation device 180 may assist the physician 190 in diagnosing whether the subject 110 suffers from a cognitive condition, such as a cognitive disorder, such as ADHD.

[0188] The EEG / MEG data 130 may be sent to a data analysis device 140. In one embodiment, the EEG / MEG data 130 may include brain signals measured by the EEG / MEG device 120. In one embodiment, the EEG / MEG data 130 may include first EEG / MEG data that may be measured when the subject 110 performs a task (such as the first task discussed above) and second EEG / MEG data that may be measured when the subject 110 performs another task (such as the second task discussed above). In one embodiment, the brain signals correspond to raw brain signals (e.g., brain signals measured by the EEG / MEG device 120 without preprocessing). In another embodiment, the brain signals may be preprocessed (e.g., by removing external noise or eye, heart, and / or muscle artifacts) before the EEG / MEG data 130 is sent to the data analysis device 140. In one embodiment, the EEG / MEG data 130 may include all brain signals measured by the EEG / MEG device 120. In another embodiment, the EEG / MEG data 130 may include only a subset of brain signals measured by the EEG / MEG device 120. The subset of brain signals may be brain signals from locations within the cranial cavity of the subject 110 discussed above for the first, second, and third embodiments, such as from the frontal lobe (e.g., IFG and / or MFG) and superior parietal lobe (see the first embodiment discussed above), from a location that is part of the ventral attention network (e.g., TPJ) and a location that is part of the dorsal attention network (e.g., superior parietal lobe; see the second embodiment discussed above), and / or from a location that is part of the ventral attention network (e.g., TPJ) and a location that is part of the ventral visual pathway (e.g., inferior temporal gyrus; see the third embodiment discussed above).

[0189] In one embodiment, EEG / MEG data 130 may also include information (e.g., time information) that allows matching the time at which task presentation device 160 presents task conditions to subject 110 with brain signals measured by EEG / MEG device 120. Additionally, EEG / MEG data 130 may include data from input device 170 that indicates subject 110's interaction with input device 170.

[0190] Figure 2An example of a data analysis device 140, 200 is shown. As described above, the data analysis device 140, 200 can be internal (e.g., located within the local area network of the EEG / MEG device 120) or external (e.g., located in the cloud). For example, the data analysis device can be a computing system (such as a server). The data analysis device 140, 200 may include a receiving component 210, a time series determination component 220, a correlation value calculation component 230, a score generation component 240, and a score output component 250.

[0191] The receiving component 210 may be configured to receive EEG / MEG data 130 , such as the first EEG / MEG data and / or the second EEG / MEG data discussed above. In one embodiment, the receiving component 210 may receive the EEG / MEG data 130 from the EEG / MEG device 120 .

[0192] The time series determination component 220 can be configured to determine a first plurality of time series based on the received EEG / MEG data 130 (e.g., based on first EEG / MEG data measured when the subject 110 performs a task). Each time series in the first plurality of time series can correspond to a corresponding source position within the cranial cavity of the subject 110. In addition, the time series determination component 220 can be configured to determine a second plurality of time series based on the received EEG / MEG data 130 (e.g., based on second EEG / MEG data measured when the subject 110 performs another task). Each time series in the second plurality of time series can correspond to a corresponding source position within the cranial cavity of the subject. For example, the first plurality of time series and the second plurality of time series can correspond to the same corresponding source position.

[0193] In one embodiment, determining the first (and / or second) plurality of time series may include filtering the first (and / or second) plurality of EEG / MEG data 130 sequences such that the first (and / or) second plurality of time series may include only a subset of all available time series (such as a subset of all available time series that can be obtained from the EEG / MEG device 120). The subset may include only time series between source locations (such as the frontal lobe, SPL, TPJ, and / or ITG) discussed above with respect to the first, second, and third embodiments. In addition, filtering may also include filtering out certain frequencies of the EEG / MEG data (e.g., limiting the time series to alpha, beta, and theta frequency bands, limiting to only the alpha frequency band, limiting to only the beta frequency band, and / or limiting to only the theta frequency band). Filtering the EEG / MEG data 130 may reduce the amount of data that needs to be processed by the data analysis device 140. For example, filtering the EEG / MEG data 130 may help obtain data that needs to be processed by the data analysis device 130 for analysis.

[0194] In addition, according to one embodiment, determining the first (and / or second) plurality of time series may include preprocessing the first (and / or second) EEG / MEG data 130, such as by removing external noise, removing eye, heart, and muscle artifacts from the EEG / MEG data (e.g., brain signals measured from the EEG / MEG device 120), and segmenting the EEG / MEG data 130 so that the time series may have the same length (e.g., 1 second). In one embodiment, the time series determination component 220 may be configured to match the time at which the task presentation device 160 presents the task conditions to the subject 110 with the EEG / MEG data 130 (e.g., brain signals measured from the EEG / MEG device 120) so that each of the determined plurality of time series includes a corresponding start time of the task condition. This may ensure that the subject's brain's response to the condition is included in the time series. In a specific embodiment, preprocessing may be performed according to the preprocessing example further described above with respect to the experimental results.

[0195] Obviously, the time series determination component 220 can determine not only one time series, but also multiple time series. In particular, and as described above, the time series determination component 220 is capable of determining a corresponding time series for each source location, for each condition of a specific task, and for each subject being examined. By considering only a subset of all available time series (as described above), the amount of time series to be determined and / or processed can be reduced.

[0196] The correlation value calculation component 230 can calculate the correlation value based on the time series pair included in the multiple (such as the first multiple or the second multiple) time series determined by the time series determination unit 220. In particular, in one embodiment, the correlation value calculation component 230 can calculate the first correlation value for the first pair of time series, and the first pair of time series is included in the determined first multiple time series. In addition, the correlation value calculation component 230 can calculate the second correlation value for the second pair of time series, and the second pair of time series is included in the determined first multiple time series. In addition, the correlation value calculation component 230 can calculate the third correlation value for the third pair of time series, and the third pair of time series is included in the determined first multiple time series. In addition, the correlation value calculation component 230 can calculate the fourth correlation value for the fourth pair of time series, and the fourth pair of time series is included in the determined second multiple time series. In one embodiment, the correlation value can be associated with the phase of the pair of time series (for example, the correlation value can be given by the PLV as described above or by the cyclic correlation coefficient). In another embodiment, the correlation value can be associated with the amplitude of the pair of time series (for example, the correlation value can be given by the Pearson correlation).

[0197] The score generating component 240 can generate one or more scores 150 based on the first correlation value, and the one or more scores indicate that the subject 110 has a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD). In one embodiment, each score in the one or more scores 150 can be given by the corresponding correlation value calculated by the correlation value calculating component 230. In one embodiment, the score can be the correlation value of the time series corresponding to the frontal lobe and the time series corresponding to the SPL (being the first pair of time series). Another score can be the correlation value of the time series corresponding to the SPL and the time series corresponding to the TPJ (as the second pair of time series). Another score can be the correlation value of the time series corresponding to the TPJ and the time series corresponding to the ITG (as the third pair of time series). In one embodiment, these scores can be combined into a single score, and in another embodiment, all three scores can be output.

[0198] Based on the considerations discussed above (such as the TIC, its top-down contribution, and / or the dorsal-ventral attention network (DVAN)) and experimental results, a high correlation value between a time series corresponding to the frontal lobe and a time series corresponding to the SPL may indicate that the subject suffers from a cognitive condition, such as a cognitive disorder, such as ADHD (because the PFC pathway may be increased due to a lack of dorsal-ventral communication). Similarly, a high correlation value between a time series corresponding to the TPJ and a time series corresponding to the SPL may indicate that the subject does not have a cognitive condition (because dorsal-ventral communication may be active). In addition, a high correlation value between a time series corresponding to the TPJ and a time series corresponding to the ITG may indicate that the subject does not have a cognitive condition (because dorsal-ventral communication may be active). A physician who is presented with one or more scores (such as one or more of the three scores discussed above) may use the scores to diagnose whether the subject suffers from a cognitive condition, such as a cognitive disorder, such as ADHD.

[0199] Additionally or alternatively, one or more scores 150 may be generated based on comparing the correlation value calculated by the correlation calculation component 230 with one or more other values ​​(e.g., obtained from the second EEG / MEG data). In particular, a comparison value between the correlation value and another value may be calculated, and a score may be generated based on the comparison value.

[0200] In one embodiment, the correlation value can be compared with another correlation value obtained for another pair of time series from the same source location, the same subject and the same task, but wherein the subject 110 can be presented with another condition of the task (e.g., via the task presentation device 160). For example, the condition of the correlation value can be the second condition of the first task discussed above, and the other condition of another correlation value can be an exception to the second condition of the first task discussed above. In another example, the condition of the correlation value can be an exception to the second condition of the second task discussed above, and the other condition of another correlation value can be an exception to the exception of the second task as described above. The score generating component 240 can be configured to calculate a comparison value between the two correlation values. In one example, the score 150 can be a comparison value. In this case, the score can indicate the ability of the subject to handle exceptions. Another correlation value can be calculated by the correlation value calculating component 230 (e.g., based on the time series from the second EEG / MEG data) in the same manner as the calculated correlation value.

[0201] In another embodiment, the correlation value can be compared with another correlation value obtained for another pair of time series from the same source location, the same subject, and the same condition, but where the subject 110 is performing another task (e.g., presented via the task presentation device 160). For example, the task of the correlation value can be the first task as described above (where there is no exception to the exception of the second condition), and the other task can be the second task as described above (where there is an exception to the exception of the second condition). As described above, the second task can be performed by the subject after the first task, so that the subject needs to switch from not interacting with the input device 170 in the first task to interacting with the input device 170 in the second task (although the task presentation device 160 can present the same stimulus). The score generation component 240 can be configured to calculate a comparison value between the two correlation values. In one example, the score 150 can be a comparison value. In this case, the score can indicate the ability of the subject to handle switching from one task to another task (i.e., redirection). Another correlation value can be calculated by the correlation value calculation component 230 (e.g., based on the time series from the second EEG / MEG data) in the same manner as the correlation value.

[0202] In yet another embodiment, the correlation value may be compared to another correlation value obtained for another pair of time series from the same source location, the same conditions, and the same task, but for at least one other subject known to have a cognitive condition (e.g., a cognitive disorder such as ADHD). The other correlation value may be stored in a data storage of the data analysis device 140, 200 ( Figure 2). Likewise, the correlation value may be compared to another correlation value obtained for another pair of time series from the same source location, the same conditions, and the same task, but for at least one other subject who does not have a cognitive condition (such as a cognitive disorder, such as ADHD). Likewise, the other correlation value may be stored in a data storage of the data analysis device 140, 200.

[0203] Additionally, a score may be generated based on information from input device 170 that indicates whether subject 110 correctly interacted with the input device in response to the conditions of the task. In one example, a large number of incorrect responses may indicate that the subject suffers from a cognitive condition.

[0204] Each of the scores discussed above may be output separately as one or more scores 150, for example to assist physician 190 in diagnosing subject 110. Alternatively, a combined score may be generated (e.g., based on a weighted average of the above scores) and output to physician 190.

[0205] The score output component 250 is configured to output one or more scores 150. In one example, the score output component 250 may send the one or more scores 150 to the score presentation device 180, which may present the one or more scores to the physician 190.

[0206] Fig.15 The brain within the cranial cavity 300 of the subject 110 is illustrated. The brain may include a frontal lobe 310, a parietal lobe 370, an occipital lobe 380, and a temporal lobe 390. As described above, subjects with cognitive conditions (cognitive disorders such as ADHD) may lack the ability to activate the dorsal-ventral network (DVAN), which can connect ventral attention network areas with dorsal attention network areas and / or connect ventral attention network areas with ventral visual pathway areas. In contrast, subjects with such cognitive conditions may have increased activity in the PFC pathway. Thus, according to one embodiment, one or more scores indicating that the subject suffers from such a cognitive condition may be generated based on at least one of the following: a correlation value between a time series from the frontal lobe 310 (e.g., from the MFG 320 and / or from the IFG 330) and a time series from the subject's SPL 340, a correlation value between a time series from the SPL 340 and a time series from the subject's TPJ 350, and a correlation value between a time series from the TPJ 350 and a time series from the subject's ITG 360. In addition, for any of the three correlation values ​​(which may also be combined as further discussed above), a correlation value of a pair of time series may be used, wherein the source position of the first time series (in the pair of time series) is a connection (i.e., Figures 4 to 12 , and where the source position of the second time series (in the pair of time series) is the position where the same connection ends. An example is the correlation value between the time series from the left superior marginal gyrus (lSMG) and the left inferior temporal cortex (lITG), such as Figure 4 Another example would be the correlation value between the time series from the left superior parietal gyrus (lSPG) and the left inferior frontal gyrus, part orbital (lIFGo), as Figure 6 It is obvious that Figures 4 to 12 Any other associations shown in can be used to calculate a correlation value, which is used to generate a score indicating that the subject has a cognitive condition, as described above.

[0207] Fig.16 A computer-implemented method 400 according to an embodiment of the present disclosure is shown. Fig.16 The method steps shown in and discussed below can also be implemented as a computer-readable medium, which enables a computing system (e.g., a data analysis device 140) to perform these steps. In a first step 410, EEG / MEG data 130 (such as first EEG / MEG data or second EEG / MEG data) of a subject is received. In a next step 420, multiple time series are determined based on the EEG / MEG data. For example, a first plurality of time series are determined based on the first EEG / MEG data, and a second plurality of time series are determined based on the second EEG / MEG data. Each time series in the multiple time series can correspond to a corresponding source position within the cranial cavity 300 of the subject 110. Then, at step 430, a correlation value can be calculated for a pair of time series included in the multiple time series. At step 440, a score 150 is generated based on the correlation value. The score 150 can indicate that the subject suffers from a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD). Next, at step 450, the generated score is output.

[0208] Other examples that may help understand the present invention

[0209] Systems and methods consistent with the present disclosure relate to effectively and accurately detecting ADHD in individuals. The systems and methods described below include techniques for presenting simple tasks to accurately detect the presence of ADHD and the probability of certainty of ADHD. The present invention also identifies metrics for modeling tasks and calculating the probability of certainty of ADHD. In some embodiments, the disclosed techniques include multiple tasks for presenting content with different conditions that cause reactions to the content. As described below, specialized tasks with reaction conditions can lead to various technical improvements in the accuracy of ADHD detection using underlying systems, hardware and software, and other applications executed on the underlying hardware and software.

[0210] Fig.17 1 is a block diagram illustrating various exemplary components of an attention deficit detection system (ADDS) 1700 for accurately detecting ADHD according to some embodiments of the present disclosure. In one embodiment, the ADS 1700 may be Fig.13 The system 100 shown in and described above. Detection of ADHD may include evaluating a user of ADDS 1700 to detect the presence of ADHD and a percentage of certainty of ADHD. The measure of the indication of ADHD may be based on the amount of match between the user's response and the expected outcome of the task. In some embodiments, ADDS 1700 may measure the amount of task response deviation from the task response defined by the user of ADDS 1700 as part of the initialization of ADDS 1700. The user of ADDS 100 may define a measure of the attention level of a healthy individual when configuring ADDS 1700. For example, user 1760 may configure ADDS 1700 by providing a text configuration file to be parsed by ADDS 1700.

[0211] like Fig.17 As illustrated, the ADDS 1700 may include a measurement module 1710 for assessing an individual's attention level and a database 1730 for storing the assessed attention level in a performance metric 1734. The measurement module 1710 may use data from the database 1730 to help determine the attention level. The database 1730 may help assess the attention level based on a control group 1731 associated with a user 1732 through a task definition 1733. The reporting module 1720 may help assess the attention level to determine the attention deficit of the user 1732 associated with the control group 1731. The ADDS 1700 may determine the attention level of the control group 1731 and the user 1732 using a response 1740 provided using a user device 150. The user device 1750 may be a processor or a complete computing device, such as a laptop computer, a desktop computer, a mobile device, a smart home appliance, an IoT device, etc.

[0212] The measurement module 1710 can measure the attention level of a user (e.g., user 1760) by measuring oscillation signals in different areas of the user's brain. The measurement module 1710 can measure the attention level of the user 1760 by determining the level of connectivity between different areas of the user's 1760 brain. The measurement module 1710 can measure the oscillation signal when a reaction 1740 from the user 1760 is received from the user device 1750. The ADDS 1700 can be configured to allow the measurement module 1710 to collect signals from different brain areas based on the reaction. The configuration can include whether and when to collect brain signals, and from which areas to collect brain signals. The user of the ADDS 1700 can configure the measurement module 1710 using a configuration provided by the user device 1750 via the network 1770.

[0213] The reporting module 1720 can evaluate the measured signals to determine the presence of ADHD and calculate a percentage of certainty of ADHD. The reporting module 1720 can evaluate the signals by measuring signal connectivity. In some embodiments, the reporting module 1720 can evaluate the signals by comparing the determined connectivity levels. Fig.19 and Fig. 20 A detailed description of the various evaluation techniques is provided in the description.

[0214] In various implementations, database 130 can take a number of different forms. For example, database 1730 can be a SQL database or a NoSQL database, such as that provided by MICROSOFT TM , REDIS, ORACLE TM, CASSANDRA or MYSQL developed or various other types of databases, including data returned by calling a web service, data returned by calling a computing function, sensor data, IoT devices or various other data sources. Database 1730 can store data used or generated during the operation of the application, such as data generated by measurement module 1710. For example, if measurement module 1710 is configured to evaluate reaction 1740, it can access task definition 1733 to share stimulus 1780 and measure the signal using connectivity level when receiving reaction 1740 and store it as performance metric 1734. Similarly, if reporting module 1720 is configured to determine ADHD, reporting module 1720 can access performance metric 1734 to obtain the assessed signal connectivity level associated with the user (e.g., user 1760) for the stimulus, relative to the connectivity level of the signal for the same stimulus provided to control group 1731. In some embodiments, database 1730 can be fed with data from an external source (e.g., a server, a database, a sensor, an IoT device, etc.). In some implementations, database 1730 may provide data storage for a distributed data processing system (eg, Hadoop Distributed File System, Google File System, ClusterFS, and / or OneFS).

[0215] The network 1770 can take various forms. For example, the network 1770 can include or utilize the Internet, a wired wide area network (WAN), a wired local area network (LAN), a wireless WAN (e.g., WiMAX), a wireless LAN (e.g., IEEE802.11, etc.), a mesh network, a mobile / cellular network, an enterprise or private data network, a storage area network, a virtual private network using a public network, or other types of network communications. In some embodiments, the network 1770 can include a local (e.g., LAN) network, while in other embodiments, the network 1770 can include a virtualized (e.g., AWS) network. TM , Azure TM , IBM Cloud TM In addition, in some embodiments, network 1770 can be a hybrid local and virtualized network, including components of both types of network architectures.

[0216] ADDS1700 can provide stimulus 1780 to user 1760 of user device 1750 by presenting content that matches the conditions of task definition 1733. User 1760 provides reaction 1740 to stimulus 1780 using a device attached to user device 1750. For example, user 1760 of user device 1750 can react to the task presented to user 1760 by clicking a button or touching a screen to respond to stimulus 1780. Reactions 1740 to the stimulus associated with the task can include answers to the shared task. For example, a task can include a question that provides a list of response options that can be selected by user 1760 in response to the question. In some embodiments, reactions 1740 can include the user's non-reaction to the task. Non-reaction can be recorded based on the lack of response to the stimulus after a certain period of time. Determining the time period for non-reaction to the stimulus can be configured by the user (e.g., user 1760) of ADDS1700.

[0217] The task definition 1733 includes reaction rules that can be pre-presented to the user 1760 of the user device 150 to react to the stimulus 1780 based on the task rules in the task definition 1733. The rules can include when to react and how to react. In some embodiments, the rules can include multiple conditional statements that divide and subdivide the reaction type and reaction time based on the content that meets the task definition. In some embodiments, a task can include a rule that extends the rules of another task.

[0218] FIG. 18A to FIG. 18B 18 is a tabular representation of an exemplary task for detecting ADHD and its probability according to some embodiments of the present disclosure. The tabular representation includes various response reaction conditions presented as row and column headings 1810 to 1840 and 1860 to 1890. Fig.18A As shown, row headers 1810 and 1820 are divided into user devices (e.g., Fig.17 1750) is presented to the user (e.g., Fig.17 The divided content indicates when the user should react to content presented as stimulus 1780 on user device 1750. For example, task 1800 divides content into consonants and vowels, with a rule of not reacting when consonants are presented as stimuli. Tasks may include applying restrictions to stimuli presented to the user (e.g., Fig.17 In task 1800, the additional conditions are included as column headings (e.g., column headings 1830 and 1840). The additional conditions can apply restrictions when the first condition of the reaction is met. For example, Fig.18A As shown, task 1800 includes a second condition of letter color, which constrains responses to vowels when it is a red vowel.

[0219] A task can include multiple sub-conditions that add different restrictions or permissions. For example, Fig.18B Task 1850 is presented in a table format, where a vowel may include a condition to determine whether it is the letter "E" or another vowel as one condition, and whether the color of the letter is "red" or "other color" as a second condition.

[0220] A reaction is an action by a user (e.g., user 1760) on a user device (e.g., user device 1750). Reactions can include positive reactions (e.g., clicking a button) and negative reactions (e.g., clicking a different button). In some embodiments, a negative reaction can be no reaction. For example, in task 1800, when a consonant is displayed (rule 1810), user 1760 is expected to have a negative reaction by not reacting to the displayed letter. Task conditions can divide the universe of possible content into active states and silent states. The active state and silent state of a condition indicate when user 1760 is expected to react or not react to an instance from the universe of content presented as stimulus 1780. For example, if Fig.18A As illustrated, task 1800 includes the alphabet as content, and is divided into consonants associated with negative reactions and vowels associated with positive reactions. In some embodiments, the task may include restrictions on conditions to further restrict the expectations of the active state. For example, in task 1800, condition 1820 may have a restriction of not reacting when the displayed vowel is displayed in red. In some embodiments, the second condition may apply restrictions to content belonging to both the active state and the silent state.

[0221] Fig.18B A table representation of task 1850 is illustrated with additional conditions for reversing restrictions on active state content. Fig.18B As illustrated, rules 1860 and 1870 include an active state and a silent state for a first condition, similar to rules 1810 and 1820 for the first condition in task 1800. Similar to task 1800, task 1850 includes a second condition with rules 1880 and 1890 to set restrictions when the first condition matches the active state. In addition to the second condition, a third condition can segment the active state content. The third condition segments the active state content to reverse the restrictions imposed by the second condition rules 1880 and 1890. For example, Fig.18BAs shown, task 1850 includes a third condition that splits the vowel content in the active state into the letter "E" and other vowels. When the third condition is positive, the task allows reversal of the response restriction imposed by the second condition on the active state content. For example, in task 1850, the second condition rule 1880 sets a restriction of not responding when the vowel is "red", but the rule 1875 of the third condition reverses this restriction of the letter "E" by allowing a response.

[0222] Content belonging to the population of content that satisfies the first condition, second condition, or third condition for reaction and the rules of expected reaction defined in tasks 1800 and 1850 may be presented to a user of ADDS 1700 (e.g., user 1760). ADDS 1700 may then collect reactions (e.g., Fig.17 1740) and signals generated in the brain of the user 1760. When the user 1760 reacts 1740 when the stimulus 1780 is displayed, the collected brain signals will be evaluated to determine the level of connectivity between the brain regions of the user 1760 from which the brain signals were collected. Fig.19 A detailed description of different brain regions and determination of connectivity levels are described in the description.

[0223] Figure 3 The conditional probabilities of the various outputs of the cascade model according to some embodiments of the present disclosure are illustrated. The conditional probabilities and information transfer are calculated using the transfer information content (TIC) metric given below. For example, using the following TIC metric, in task 1800, when the consonant data is received by the user (e.g., Figure 1 When observed by user 1760), the TIC metric uses the TIC metric defined below to calculate the amount of information transfer as follows:

[0224]

[0225] In the above case, That is, there is a negative transfer of information from stimulus 1780, which includes the display of consonants as part of task 1800, which elicits a response. The negative information bit is an indication that user 1760 is inhibited or that he / she is restraining himself / herself from providing an affirmative response (e.g., clicking a button). As a result, because I(a|s) is much greater than I(a), the result is that the button is not pressed.

[0226] If vowel data is presented to user 1760 as part of task 1800, user 1760 needs to check whether the color of the letter is red (e.g., a red vowel). In the case of non-red vowels, the conditional TIC metric is used to calculate information transfer as follows:

[0227]

[0228] Both the sensorimotor TIC and the contextual TIC are positive and are subtracted from the initial uncertainty, which decreases to zero. Therefore, the button is pressed.

[0229] If the red vowel data is presented to user 1760 as part of task 1800, user 1760 needs to check whether the letter is red. Applying conditional TIC to the context of red, we have:

[0230]

[0231] In this case, the component I(c|a,s) generates negative information bits, which offset the information bits generated by the evidence representing the top-down control exerted by the subject. To derive the value of the probability P(c|a,s) (whose information content we are interested in), we can reason as follows: Assume that a vowel has occurred and the user (e.g., Fig.17 ) has pressed the button, what is the probability that the vowel is red? The answer is that the probability is very low, and therefore, its associated information content is very high, i.e., a large number of negative information bits have been generated to offset the positive bits generated by evidence I(c|s).

[0232] Thus, in this context, there is a negative transfer of information from the evidence of the presence of the red vowel to the action a of pressing the button.

[0233] If the red vowel data is presented to user 1760 as part of task 1850, user 1760 needs to check whether the letter is "E". Information transfer is calculated using the episodic TIC metric as follows:

[0234]

[0235] A positive sensorimotor TIC reduces uncertainty about the action, but a negative contextual TIC increases uncertainty, and the net contribution is zero. Finally, a positive episodic TIC reduces uncertainty further, and the uncertainty is reduced to zero. Therefore, the button is pressed.

[0236] The above amounts of information bits can be used as default values ​​for comparing different users to determine attention levels. The ADDS 1700 can start with these default values ​​as expected signal values ​​and modify them as signals from more individuals considered to be part of the control group are added.

[0237] Fig.19 illustrating a network of connections between brain regions according to some embodiments of the present disclosure. Fig.17When a user 1760) provides a response 1740, ADDS 1700 can identify the presence of signals in various brain regions that represent different connection networks. ADDS 1700 can use the signals identified in the connection networks to determine the level of connectivity between different regions of the brain. ADDS 1700 can assess the level of connectivity between various regions of the brain to determine attention levels and determine whether the user indicates ADHD based on those determined attention levels.

[0238] ADDS1700 can determine and use signals of varying frequencies to determine attention levels and deficits. ADDS1700 can group signals by frequency spectrum when evaluating signals and using the connectivity of signals to determine attention levels and attention deficits. ADDS1700 groups signals in the frequency range of 8Hz to 12Hz into alpha groups, signals in the frequency range of 20Hz to 30Hz into beta groups, and signals in the frequency range of 4Hz to 8Hz into theta groups. In some embodiments, ADDS1700 can collect signals of different frequencies at different times and from different areas of the brain.

[0239] Each signal group indicates a certain aspect of task stimulus provision and response behavior. For example, alpha band oscillation amplitude can indicate inhibition of activity. In some embodiments, signal group characteristics (e.g., amplitude) can have multiple meanings. For example, the same alpha band oscillation amplitude can also indicate active processing, causing it to be considered task-related information. Any changes in signals or characteristics (such as signal amplitude) can be considered as differences in cognitive activity in the brain of the user 1760 that generated the signal.

[0240] In some embodiments, a decrease in connectivity between brain regions indicates an increase in attention. The increase in connectivity is based on the tasks used to activate different regions of the user's 1760 brain.

[0241] like Fig.19 As illustrated, disclosed embodiments of ADDS 1700 assess communication between various known regions of the brain used to assess attention (e.g., regions within the dorsal attention network (DAN) 1910 or the ventral attention network (VAN) 1920). Disclosed embodiments of ADDS 1700 may utilize FIG. 18A to FIG. 18B The ADDS 1700 generates the information flow of positive and negative bits occurring between the DAN 1910 and the VAN 1920 by using the tasks defined in the ADDS 1700 and identifies the pathways involved in the information flow. The ADDS 1700 can also identify connectivity networks and alternative pathways that can include signals when different individuals respond to the same stimulus. For example, the ADDS 1700 can detect signals in the dorsal-ventral attention network (DVAN) 1930.

[0242] The ADDS 1700 determines the user's ( Fig.17 The disclosed embodiment of ADDS 1700 uses tasks to provide stimulation as a way to trigger activity within the brain network and generate signals by passing information (positive and negative bits), and then uses the generated signal connectivity to test the user for ADHD (e.g., Fig.17 For example, ADDS 1700 may use tasks 1800 and 1850 to activate and test the presence of signals in DVAN 1930 by presenting an instance of the entirety of content presented for user response as defined by tasks 1800 and 1850. Use of DVAN 1930 is tested by comparing different types of measured signals assessed in user 1760 with a previously identified group of healthy users without ADHD (e.g., control group 1731). In the following Fig. 20 A detailed description of the comparative analysis of signal groups between different users is presented in the description of .

[0243] In some embodiments, attention deficits can also be determined by comparing changes in connectivity when switching between tasks with different rules. In this case, instances of the content population are selected where opposite responses are defined in the task. For example, a red "E" letter presented as part of task 1800 would result in an infeasible condition where the active state is restricted by a negative second condition. Similarly, a red "E" letter presented as part of task 1850 would result in a restricted second condition that is reversed by a third condition. A detailed description of attention network analysis in task switching scenarios is provided below. Fig.21 Description.

[0244] The disclosed embodiments of ADDS 1700 can help model the probability of ADHD level based on the tasks presented for the action. The probability of ADHD level helps to accurately predict the presence of ADHD and the probability of ADHD. The model for identifying the probability of ADHD level can use the above Figure 3 The example tasks 1800 and 1850 presented in the description are implemented.

[0245] Bayesian brain theory establishes a probability distribution for the entire cortex representation, which collapses to an estimate only when a decision is needed. Under Bayesian theorem, the effect of data x as stimulus 1780 on user 1760 is presented as follows by changing the prior distribution P(θ) to the posterior distribution P(θ|x):

[0246] P(θ|x)=P(x|θ)P(θ) / P(x)

[0247] Here, P(x) is called evidence and P(x|θ) is the likelihood of the Bayesian formula.

[0248] When data x is presented as stimulus 1780, the same formula presented using the -log function for representing the information content transmitted in the brain is as follows:

[0249] -log2[P(θ|x)]=-log2[P(x|θ)P(θ) / P(x)]

[0250] -log2[P(θ|x)]=-log2P(x|θ)-log2P(θ)+log2P(x)

[0251] The information content (in bits) of an event x_i, also called self-information, is given by:

[0252] I(x i )=-log2[p(x i )]Bit

[0253] Therefore, the logarithmic equation expressed as information bits is as follows:

[0254] I(θ|x)(bits)=I(θ)(bits)-I(x)(bits)+I(x|θ)(bits)

[0255] One aspect of the above equation is that the change in the information content of a single model θ in the space of multiple models resulting from a single observation x occurring on the rendered data x is not a compound, but rather the two information contents I(x) and I(x|θ) generated by the single observation. Thus, in the above equation, -I(x) (bits) + I(x|θ) (bits) represents the information transferred to the single model θ due to the observation x. The ADDS1700 calculates this information transfer using a novel metric called Transfer Information Content (TIC), defined as:

[0256] TIC(x→θ)(bit)=I(x)-I(x|θ)

[0257] I(θ) can also be interpreted as the prior uncertainty about the model θ, and I(θ|x) as the posterior uncertainty about the model θ after a single observation x.

[0258] When the TIC metric is applied to the information content equation above, it is transformed as follows:

[0259] I(θ|x)(bit)=I(θ)(bit)-TIC(x→θ)(bit)

[0260] The ADDS1700 uses the TIC metric to describe information flow. If the negative bits of information content of the Bayesian likelihood I(θ|x) do not offset the positive bits of information content of the evidence I(x), then the TIC metric is positive, that is, there is a positive net transfer of information; conversely, if the negative bits corresponding to I(θ|x) are greater than the bits of evidence I(x), then the TIC is negative and there is a negative net transfer of information. In the following Fig. 20 A detailed description of using this gain or loss of bits in the formal signal transfer between two regions to detect ADHD is provided in detail in the description of .

[0261] Using the TIC metric defined above, when using stimuli (e.g. Fig.17 of stimulation 1780) and actions (e.g., Fig.17 When the reaction 1740) is expressed, the information content is transformed into

[0262] I(a|s)=I(a)-TIC(s→a)(bit)

[0263] I(a) can also be interpreted as the a priori uncertainty about the action a, and I(a|s) as the uncertainty about the stimulus s after it occurs, which is determined by the user (e.g., Fig.17 The a posteriori uncertainty of action a taken by user 1760).

[0264] When stimulus s occurs in context c defined in tasks 1800 and 1850 using different conditions, the information content of triggering or not triggering action a is conveyed as follows:

[0265] I(a|s,c)=-log2[P(a|s,c)]

[0266]

[0267] ADDS1700 can use the conditional TIC metric to describe the contextual information content transfer. The conditional metric is defined as follows:

[0268] I(a|s,c)=I(a)-TIC(s→a)-TIC(c→a|s)

[0269] I(a|s,c)=I(a)-TIC(s→a)(bits)-[I(c|s)-I(c|a,s)]

[0270] =I(a)-I(s)+I(s|a)-I(c|s)+I(c|a,s)(bits)

[0271] I(a) can also be interpreted as the prior uncertainty about action a, and I(a|s,c) as the posterior uncertainty about action a after stimulus s occurs in context c.

[0272] When additional conditions are presented to user 1760, such as the second condition in task 1800 defined using rules 1830 and 1840, ADDS 1700 may use conditional TICs to describe the information bits being passed.

[0273] In some embodiments, ADDS 1700 can consider the occurrence of stimulus s in context c and episode (temporal context) e to calculate the TIC metric of information transfer that can trigger action a. Information transfer is as follows:

[0274] I(a|s,c,e)=-log2[P(a|s,c,e)]

[0275] Similar to the context case, ADDS1700 can then apply a conditional version of Bayes’ theorem to the posterior probabilities as follows:

[0276]

[0277] Then, applying the logarithmic rule, we have:

[0278]

[0279] The ADDS1700 can use the conditional TIC metric to describe the information content of the episode transfer. The episode metric is defined as follows:

[0280] I(a|s,c,e)=I(a)-TIC(s→a)-TIC(c→a|s)-TIC(e→a|s,c)

[0281] Now, the sensorimotor TIC, contextual conditional TIC, and episodic conditional TIC can be decomposed into their two components:

[0282] I(a|s,c,e)=I(a)-TIC(s→a)-TIC(c→a|s)-[I(e|s,c)-I(e|a,s,c)]

[0283] =I(a)-I(s)+I(s|a)-I(c|s)+I(c|a,s)-I(e|s,c)+I(e|a,s,c)

[0284] The components I(e|s,c) and I(e|a,s,c) of the episode-conditional TIC represent the bottom-up contribution and top-down contribution in action selection, respectively. After reasoning about the contextual situation, an uninformative stimulus in a given context may become informative if the episode-conditional TIC of the uninformative stimulus in the given context is positive and compensates for the negativity of the context-conditional TIC. ADDS1700 can also interpret I(a) as the prior uncertainty about action a, and I(a|s,c,e) as the posterior uncertainty about action a after stimulus s occurs in context c and episode e.

[0285] When additional conditions (such as the second condition in task 1850 defined using rules 1880 and 1890) are presented to user 1760, ADDS 1700 can use conditional TIC to describe the information bits being passed. Figure 3 The description of Figure 1 A detailed description of the conditional probabilities and information bits transferred when various stimuli are presented to user 1760).

[0286] Fig. 20 2000 is a flowchart depicting the operation of an exemplary method for attention deficit detection according to some embodiments of the present disclosure. For illustrative purposes, the steps of method 2000 may be performed by ADDS 1700. It should be understood that the illustrated method 2000 may be altered to modify the order of the steps and to include additional steps.

[0287] In step 210, ADDS 1700 may provide a user (eg, Fig.17 The ADDS 1700 can be configured to provide a stimulus to induce a response by providing a stimulus to induce a response at a user device (e.g., Fig.17 ADDS 1700 may present a task (e.g., tasks 1800 and 1850 of FIG. 18 ) on a user device 1750 of the ADDS 1700. ADDS 1700 may present a task by displaying the rules of the task and then content for review and reaction. A user 1760 of the user device 1750 may react to the provided stimulus, which is an example of possible content based on the rules of the task. ADDS 1700 may continuously provide stimulus by presenting content that matches the task. In some embodiments, ADDS 1700 may cycle through different tasks to provide stimulus to a user 1760 of ADDS 1700.

[0288] In step 220, the ADS 1700 may obtain signals from sensors in response to the presented stimulus. The sensors may be probes from an electroencephalogram (EEG) or magnetoencephalogram (MEG) device, or other functional neuroimaging techniques for mapping brain activity, such as functional magnetic resonance imaging (fMRI) and functional near infrared spectroscopy (fNIRS). The sensors are placed on the head, matching different known regions (e.g., Fig.19 1910 to 1930 of the brain region). ADDS1700 can obtain signals based on certain triggering events. In some embodiments, the triggering event may include a time period. The time period may include a waiting time since the stimulation was provided in step 2010. ADDS1700 can allow the configuration of triggering events by allowing the selection of triggering events to obtain signals and parameters of the triggering events. For example, ADDS1700 can allow the amount of waiting time after providing stimulation before obtaining signals from different brain regions. The EEG / MEG device can be a standard EEG or MEG device. The EEG / MEG device can measure multiple signals of corresponding source locations located in the cranial cavity of the user 1760. Each source location can correspond to a known brain region, such as a lobe, a lobule, a sulcus, or a gyrus. In some examples, multiple signals can be measured when the subject performs one or more tasks that can be presented to the user 1760 by ADDS1700. In one or more embodiments, the signal can correspond to the measurement results of the electrodes of the EEG device or the measurement results of the sensor coils of the MEG device. In some embodiments, the signal may correspond to an aggregate signal (such as an average signal) over multiple measurements from multiple electrodes (when EEG is used) or multiple sensor coils (when MEG is used). Such an aggregate signal may correspond to a region of the brain that includes the corresponding locations of the signals aggregated in the aggregate signal.

[0289] MEG signals can be measured using a 306-channel (102 magnetometers and 204 planar gradiometers) whole-head MEG Elekta neuroimaging system. In some embodiments, brain signals can be measured using an online anti-aliasing bandpass filter between 0.1 Hz and 330 Hz and a sampling rate of 1000 Hz. In some embodiments, in addition to three reference points (nasion and left and right preauricular points) as landmarks, a 3D Fastrak digitizer (Polhemus, Colchester, Vermont) can be used to acquire the user's (e.g., Fig.17 In another embodiment, four head position indicator (HPI) coils may be recorded on the scalp (forehead and mastoid) of the user 1760. In some embodiments, two sets of bipolar electrodes may be used to record eye blinks and heartbeats, respectively.

[0290] In step 2030, ADDS1700 can evaluate the signals of each brain region from the brain signals obtained in step 2020. ADDS1700 can evaluate the signal quality to remove artifacts from the signal and reconstruct brain activity using the signals obtained from each sensor on the user's head that approximate each brain region. ADDS1700 can improve the data by removing external noise (such as eye, heart or muscle artifacts) from the measured EEG / MEG data and segmenting the EEG / MEG data so that the time series can have the same length (e.g., 1 second).

[0291] In step 2031, ADDS1700 may filter the signal extracted in step 2030 into a plurality of frequency bands by using, for example, a pass filter. The frequency bands may be functionally related signal frequency bands. Fig.19 As provided in the description of , the signal can be filtered into different frequencies to indicate different attention behaviors of the user of ADDS1700. For example, ADDS1700 extracts the signal into frequency bands: Theta (4Hz to 8Hz), Alpha (8Hz to 12Hz), Low Beta (12Hz to 20Hz), High Beta (20Hz to 30Hz), and Low Gamma (30Hz to 45Hz). In some embodiments, the frequency ranges describing theta, alpha, low beta, high beta, and low gamma bands may be different. In some embodiments, there may be additional signal groups between the example frequency bands. ADDS1700 may allow multiple frequency band ranges to be configured as part of a signal group. ADDS1700 extracts signals by frequency band by filtering and removing signals that do not fall into one of the selected frequency bands.

[0292] In step 2032, ADDS1700 can determine the connectivity level between two brain regions based on the dependency measure between the signals of the two brain regions of each frequency band. For example, ADDS1700 can measure the phase synchronization between the signals grouped under the frequency band to calculate the connectivity value. The phase synchronization measure includes a phase lag index or a phase lock value (PLV) and its derivative, an imaginary part of the PLV (iPLV) and a corrected imaginary part of the PLV (ciPLV). ADDS1700 can use other different connectivity measures, including information theory measures, such as mutual information or transfer entropy, Granger causality measures, and classical measures, such as correlation, cross-correlation, coherence or phase slope index. ADDS1700 can measure the connectivity level between two brain regions based on the metric derived from TIC. In some embodiments, the dependency measure can be based on the connectivity between brain regions between different frequency bands.

[0293] In step 2033, ADDS1700 can check whether the signal type matches the frequency of the theta band signal by filtering the theta band signal. If the answer is "no", jump to step 2038.

[0294] At step 2034, ADDS 1700 may check whether the first condition of the second task is satisfied. If the answer to the question is "yes", then jump to step 2035. If the answer to the question is "no", then jump to step 2056.

[0295] In step 2035 , ADDS 1700 may check whether the second condition of the second task is satisfied. If the answer to the question is “no”, then jump to step 2037 .

[0296] At step 2036, ADDS 1700 may check whether the third condition of the second task is satisfied. If the answer to the question is "yes", proceed to step 2037. If the answer to the question is "no", jump to step 2056.

[0297] In step 2037, ADDS1700 can check whether the connectivity level of the user calculated in step 2032 is less than the connectivity level of the healthy control group. If the answer to the question is "yes", jump to step 2070. ADDS 1700 can obtain the connectivity level of the healthy control group stored in the population database 1730. ADDS1700 can receive the control group type user as input and store the result in the control group 1731. ADDS1700 can compare the connectivity levels by checking whether the calculated connectivity level is less than the range of possible connectivity levels for the healthy control group. ADDS1700 can determine the range of values ​​for the connectivity level of the healthy control group based on the connectivity level of individuals identified as having ADHD as evaluated by method 2000.

[0298] At step 2038, ADDS100 may check if the signal type matches the beta band signal frequency by filtering the beta band signal. If the answer is "no", jump to step 2049.

[0299] In step 2039, ADDS1700 can check whether the stimulus is from the first task. If the answer to the question is "no", jump to step 2044.

[0300] At step 2040 , ADDS 1700 may check whether the first condition of the first task is satisfied. If the answer to the question is “no”, then jump to step 2048 .

[0301] At step 2041, ADDS 1700 may check whether the second condition of the first task is satisfied. If the answer to the question is "yes", then jump to step 2048. If the answer to the question is "no", then jump to step 2056.

[0302] At step 2044, ADDS 100 may check whether the first condition of the second task is satisfied. If the answer to the question is "yes", then proceed to step 2045. If the answer to the question is "no", then jump to step 2056.

[0303] At step 2045, ADDS 1700 may check whether the second condition of the second task is satisfied. If the answer to the question is "yes", proceed to step 2046. If the answer to the question is "no", jump to step 2047.

[0304] At step 2046, ADDS 100 may check whether the third condition of the second task is satisfied. If the answer to the question is "yes", then jump to step 2048. If the answer to the question is "no", then jump to step 2056.

[0305] At step 2047, ADDS 1700 may check whether the third condition of the second task is satisfied. If the answer to the question is "no", proceed to step 2048. If the answer to the question is "yes", jump to step 2056.

[0306] In step 2048, ADDS1700 can check whether the connectivity level of the user calculated in step 2032 is greater than the connectivity level of the healthy control group. If the answer to the question is "yes", jump to step 2070. ADDS 1700 can obtain the connectivity level of the healthy control group stored in the population database 1730. ADDS1700 can receive the control group type user as input and store the result in the control group 1731. ADDS1700 can compare the connectivity levels by checking whether the calculated connectivity level is less than the range of possible connectivity levels for the healthy control group. ADDS1700 can determine the range of values ​​for the connectivity level of the healthy control group based on the connectivity level of individuals identified as having ADHD as evaluated by method 2000.

[0307] At step 2049, ADDS1700 can check whether the type of signal matches the alpha band signal frequency by filtering the alpha band signal. If the answer to the question is "yes", then go to step 2050. If the answer to the question is "no", then jump to step 2056.

[0308] In step 2050, ADDS1700 can check whether the stimulus is from the first task. If the answer to the question is "no", jump to step 2054.

[0309] In step 2051, ADDS1700 may check whether the first condition of the first task is met. If the answer to the question is "no", proceed to step 2051.2; otherwise, jump to step 2052.

[0310] In step 2051.2, if the connectivity level of user 1760 calculated in step 2032 is stored in container A1, then jump to step 2055. Container A1 can be a data structure, such as a list or array. In some embodiments, container A1 can be persisted on disk in a flat file or database. In step 2052, ADDS 1700 can check whether the second condition of the first task is met. If the answer to the question is "no", then go to step 2052.2; otherwise, jump to step 2052.3.

[0311] In step 2052.2, if the user's connectivity level calculated in step 2032 is stored in container A2, jump to step 2055. Container A2 can be a data structure, such as a list or array. In some embodiments, container A2 can be persisted on disk in a flat file or database. When the first condition of the first task is met and the second condition is not met, container A2 will contain alpha band connectivity values ​​between brain regions.

[0312] In step 2052.3, the connectivity level of user 1760 as calculated in step 2032 is stored in container A1.

[0313] In step 2053, ADDS1700 can check whether the third condition of the first task is met, that is, it is a red vowel "E". If the answer to the question is "yes", go to step 2053.2; otherwise, jump to step 2055.

[0314] In step 2053.2, if the user's connectivity level calculated in step 2032 is stored in container A3, jump to step 2055. Container A3 can be a data structure, such as a list or array. In some embodiments, container A3 can be persisted on disk in a flat file or database. When the first condition, the second condition, and the third condition of the first task are met, that is, it is a red vowel "E", container A3 will contain the alpha band connectivity values ​​between brain regions.

[0315] In step 2054, ADDS1700 may check whether the first condition, the second condition, and the third condition of the second task are satisfied. If the answer to the question is "yes", proceed to step 2054.2; otherwise, jump to step 2056.

[0316] In step 2054.2, if the user's connectivity level calculated in step 2032 is stored in container A4, proceed to step 2055. Container A4 can be a data structure, such as a list or array. In some embodiments, container A4 can be persisted on disk in a flat file or database. When the first condition, the second condition, and the third condition of the second task are met, that is, it is a red vowel "E", container A4 will contain the alpha band connectivity values ​​between brain regions.

[0317] In step 2055, ADDS1700 can check whether the connectivity level of the user calculated in step 2032 is less than the connectivity level of the healthy control group. If the answer to the question is "yes", jump to step 2070. ADDS 1700 can obtain the connectivity level of the healthy control group stored in the population database 1730. ADDS1700 can receive the control group type user as input and store the result in the control group 1731. ADDS1700 can compare the connectivity levels by checking whether the calculated connectivity level is less than the range of possible connectivity levels for the healthy control group. A range of values ​​including the connectivity level of the healthy control group can be determined based on the connectivity level of the individual identified as having ADHD.

[0318] In step 2056, ADDS1700 may check whether all connectivity values ​​calculated in step 2032 have been checked for ADHD identification by going through steps 2033 to 2055. If the answer to the question is "yes", proceed to step 2057. If the answer to the question is "no", check the next connectivity value by jumping to step 2033.

[0319] At step 2057, ADDS1700 may retrieve the connectivity level values ​​in containers A1 and A2 from steps 2051.2, 2052.2, and 2052.3. ADDS1700 may compare the changes in connectivity level. If the decrease in connectivity level is less than the decrease in connectivity level of the healthy control group, then jump to step 2070.

[0320] At step 2058, ADDS1700 may retrieve the connectivity level values ​​in containers A3 and A4 from steps 2053.2 and 2054.2, respectively. ADDS1700 may compare the changes in connectivity levels. If the changes in connectivity levels between brain regions of the user are different from the decreases in connectivity levels between the same regions of the healthy control group, then jump to step 2070.

[0321] In some embodiments, ADDS1700 can compare the change in connectivity level between two regions based on the connectivity level of the signal determined in steps 2033, 2038, and 2049 of method 2000. ADDS 1700 can calculate the probability of ADHD based on the difference in the change in connectivity level between the user of ADDS1700 and the healthy control group. ADDS1700 can collect multiple connectivity levels of the same user or a group of users to save them as healthy control group 1731 data or for those users 1732 identified as ADHD. ADDS 1700 can use the values ​​of the stored connectivity levels to determine the change in connectivity level. In some embodiments, the change in connectivity level can only include the connectivity level observed for the same data for different conditions. For example, the same data can be presented to the user of ADDS1700, and then the task rules are presented to determine the connectivity level of the same data across multiple tasks. In step 2060, ADDS1700 can mark the user (e.g., user 1760) whose brain signal is evaluated for connectivity level as having an indication of ADHD. ADDS1700 completes (step 2099 ) execution of method 2000 upon completion of step 2060 .

[0322] Fig.21 21 is a flowchart depicting the operation of an exemplary method for detecting ADHD according to some embodiments of the present disclosure. For illustrative purposes, the steps of method 2100 may be performed by ADDS 1700. It should be understood that the illustrated method 2100 may be altered to modify the order of the steps and to include additional steps.

[0323] In step 2110, ADDS 1700 may provide a user (eg, Fig.17 The user 1760) provides a trigger associated with the first task (e.g., Fig.18A Task 1800) matches the response of the stimulus. ADDS1700 can be selected for the test user (e.g., Fig.17 ADDS 1700 can be configured to provide a set of tasks for ADHD of a user 1760) that can satisfy rules (e.g., rules 1820 and 1830 of task 1800, rules 1875 and 1880 of task 1850). ... Fig.17 Content satisfying the rules of task 1800 is displayed on user device 1750) to provide stimulation.

[0324] In step 2120, ADDS 1700 may obtain a first set of signals from the brain regions. ADDS 1700 may obtain signals from different brain regions by collecting charges from probes attached to the head of the user 1760 being tested for ADHD. ADDS 1700 may obtain signals when the user responds to the stimulus provided in step 2110 by performing an action. Action performance may include clicking a button or using a pointing device to select content presented on a screen.

[0325] In step 2130, ADDS 1700 may provide the same stimulus as a second task (e.g., Figure 2 ADDS 1700 may select tasks where the first task and the second task satisfy the rules. ADDS 1700 may select tasks only from task definitions 1733 where the same content satisfies different rules requiring opposite user reactions. For example, when a red letter "E" is presented as a stimulus under task 1800, rules 1820 and 1830 are satisfied and the user is expected to not react, but under task 1850, rules 1875 and 1880 are satisfied and the user is expected to react.

[0326] In step 2140, ADDS 1700 may obtain a second set of signals from the brain region. Similar to step 2120, ADDS 1700 may obtain signals by collecting brain images or charges from the brain of user 1760 while viewing the stimulus in step 2130. The signals may be generated as part of a reaction performed by user 1760.

[0327] At step 2150, ADDS 1700 may use the first signal set and the second signal set to assess the level of connectivity. ADDS 1700 may measure the connectivity of certain brain regions (e.g., Fig.19 dorsal attention network (DAN) 1910) and measure signals in other brain regions (e.g. Fig.19 The connectivity level is assessed by using signals in the ventral attention network (VAN) 1920) to determine the amount of transmitted signals indicative of the connectivity level.

[0328] At step 2160, ADDS 1700 may determine a change in the connectivity level of the first signal set and the second signal set. ADDS 1700 may calculate the connectivity level by providing a plurality of stimuli as part of the first task and the second task in steps 2110 and 2130 and obtaining signals of the user 1760's response in steps 2120 and 2140 to determine the change in the connectivity level.

[0329] In step 2170, ADDS1700 may compare the change in connectivity level with a known change in connectivity level to identify ADHD. ADDS1700 may compare the change in connectivity level calculated in step 2160 with the change in connectivity level of a control group (e.g., control group 1731). ADDS1700 may obtain the change in connectivity level of control group 1731 from previously generated performance metrics 1734. In some embodiments, ADDS 1700 may need to evaluate the change in connectivity level from the previously calculated connectivity level stored in performance metrics 1734. If the change in connectivity level of user 1760 is different from the change in connectivity level of control group 1731, ADDS1700 may indicate that user 1760 suffers from ADHD. If the ranges of values ​​of connectivity level are different, ADDS1700 may consider the change in connectivity level to be different. In some embodiments, if the ranges of values ​​do not overlap, the change in connectivity level is considered to be different. ADDS1700 may store the attention deficit results in database 130 under user 1732. ADDS1700 completes (step 2199 ) execution of method 2100 upon completion of step 2170 .

[0330] Example 1: A non-transitory computer-readable medium comprising instructions that, when executed by at least one processor, cause the at least one processor to perform operations for automatically detecting attention deficit hyperactivity disorder (ADHD), the operations comprising: providing one or more stimuli to a user to activate multiple brain regions representing an attention network, wherein the one or more stimuli include displaying information having a first condition that causes a response and a second condition that is an exception to the first condition; obtaining multiple signals from the multiple brain regions for each of the one or more stimuli, wherein each of the multiple signals is accessed within a time period, wherein the time period begins when the information is displayed and ends when the user's response is captured; and evaluating the multiple signals based on a level of connectivity within the attention network to detect ADHD when each of the multiple signals corresponds to the one or more stimuli.

[0331] Example 2: A non-transitory computer-readable medium according to Example 1, wherein displaying information having a first condition that causes a reaction and a second condition that is an exception to the first condition also includes: presenting an instance of data matching a first task, wherein the first task includes the first condition for expectations of an active state and a silent state for a first reaction, and the second condition limits the expectation of the active state of the first reaction.

[0332] Example 3: A non-transitory computer-readable medium according to Example 2, wherein the first reaction also includes: displaying information on the screen that satisfies the first condition expected for the active state and the silent state for the first reaction; and a threshold period of time to wait for any reaction shared by the user.

[0333] Example 4: The non-transitory computer-readable medium of example 2, wherein the first reaction is at least one of clicking a pointing device, pressing a button, or taking no action within a time threshold.

[0334] Example 5: A non-transitory computer-readable medium according to Example 2, wherein evaluating the multiple signals based on the connectivity level within the attention network to detect ADHD also includes: filtering the alpha band oscillation signals of two or more of the multiple brain regions of the user upon receiving the first reaction, and using the alpha band oscillation signals to determine the connectivity between the two regions of the brain; and when the first condition of the first task satisfies the active state and the second condition is negative, comparing the connectivity level between the two regions of the user's brain with the connectivity level of a healthy control group user, wherein if the user's connectivity level is less than the connectivity level of the healthy control group user, the user is indicated to have ADHD.

[0335] Example 6: A non-transitory computer-readable medium according to Example 2, wherein evaluating the multiple signals based on the connectivity level within the attention network to detect ADHD also includes: when the first reaction is received, filtering the alpha band oscillation signals of two or more of the multiple brain regions of the user, and using the alpha band oscillation signals to determine the connectivity between the two regions of the brain; and when the first condition of the first task satisfies the silent state and the second condition is affirmative, comparing the connectivity level between the two regions of the user's brain with the connectivity level of a healthy control group user, wherein if the user's connectivity level is less than the connectivity level of the healthy control group user, the user is indicated to have ADHD.

[0336] Example 7: According to the non-transitory computer-readable medium described in Example 2, the operation also includes: when a first response is received, filtering the beta band oscillation signals of two or more brain regions, and using the beta band oscillation signals to determine the connectivity between the two brain regions; and when the first condition of the first task satisfies the silent state and the second condition is affirmative, comparing the connectivity level between the two regions of the user's brain with the connectivity level of a healthy control group user, wherein if the user's connectivity level is greater than the connectivity level of the beta band oscillation signal of the healthy control group user, the user is indicated to have ADHD.

[0337] Example 8: According to the non-transitory computer-readable medium described in Example 2, the operation also includes: filtering alpha band oscillation signals of two or more brain regions when the first reaction is received, and using the alpha band oscillation signals to determine the connectivity between the two regions of the brain; determining the connectivity level between the two regions of the user's brain when the first condition of the first task satisfies the silent state and the first condition of the first task satisfies the active state; and comparing the decrease in the user's connectivity level between when the first task satisfies the silent state and when the first task satisfies the active state with the decrease in the connectivity level of a healthy control group user between when the first task satisfies the silent state and when the first task satisfies the active state, wherein if the decrease in the user's connectivity level is less than the decrease in the connectivity level of the healthy control group user, the user is indicated to have ADHD.

[0338] Example 9: A non-transitory computer-readable medium according to Example 1, wherein evaluating the multiple signals based on the connectivity level within the attention network to detect ADHD also includes: combining signal data of each of the multiple signals, wherein the combined signal data includes determining an increase in the connectivity level compared to an existing connectivity level between regions from the multiple regions or a decrease in the connectivity level from the existing connectivity level of the region among the multiple regions.

[0339] Example 10: The non-transitory computer-readable medium of Example 1, wherein displaying the information having the first condition comprises displaying a category of text or graphics.

[0340] Example 11: The non-transitory computer-readable medium of Example 10, wherein the second condition includes displaying the category of the text or graphic in a specific color.

[0341] Example 12: A non-transitory computer-readable medium according to Example 1, wherein the one or more stimuli also include: displaying second information having the first condition that causes a reaction, the second condition that is an exception to the first condition, and a third condition that is an exception to the second condition.

[0342] Example 13: A non-transitory computer-readable medium according to Example 12, wherein displaying second information having the first condition that causes a reaction, the second condition that is an exception to the first condition, and the third condition that is an exception to the second condition further comprises: presenting an instance of data matching a second task, wherein the second task comprises the first condition for expectations of an active state and a silent state for a second reaction, the second condition restricts the expectation of the active state for the second reaction, and the third condition, which is an exception to the second condition for the expectation of the silent state for the second reaction.

[0343] Example 14: A non-transitory computer-readable medium according to Example 13, wherein the second reaction further includes: displaying information on the screen that satisfies the expected first condition for the second reaction to be in an active state and a silent state; and waiting for a threshold period of time for any input shared by the user.

[0344] Example 15: The non-transitory computer-readable medium of Example 13, wherein the third condition comprises: selecting a subcategory of a text or graphic category that satisfies the first condition; and displaying the subcategory of text or graphics.

[0345] Example 16: The non-transitory computer-readable medium of example 13, wherein the second reaction is at least one of clicking a pointing device, pressing a button, or taking no action for a threshold time.

[0346] Example 17: According to the non-transitory computer-readable medium described in Example 13, the operation also includes: when the second reaction is received, filtering the beta band oscillation signals of two or more areas of the multiple brain areas of the user, and using the beta band oscillation signals to determine the connectivity between the two areas of the brain; and when the first condition of the second task satisfies the activity state, the second condition is negative and the third condition is negative, or when the first condition of the second task satisfies the activity state, the second condition is positive and the third condition is positive, comparing the connectivity level between the two areas of the user's brain with the connectivity level of a healthy control group user, wherein if the user's connectivity level is greater than the connectivity level of the healthy control group user, then the user indicates that he has ADHD.

[0347] Example 18: According to the non-transitory computer-readable medium of Example 13, the operation further includes: filtering theta band oscillation signals of two or more of the multiple brain regions of the user upon receiving the second reaction from the user, and determining the connectivity between the two regions of the brain using the theta band oscillation signals; and comparing the connectivity level between the two regions of the user's brain with the connectivity level of a healthy control group user when the first condition of the second task satisfies the activity state, the second condition is negative, and the third condition is negative, or when the first condition of the second task satisfies the activity state, the second condition is positive, and the third condition is positive. Wherein, if the signal data of the connectivity level of the user is less than the connectivity level of the healthy control group user, the user indicates that he or she suffers from ADHD.

[0348] Example 19: According to the non-transitory computer-readable medium of Example 13, the operation also includes: filtering the alpha band oscillation signals of two or more of the multiple brain regions of the user when the second reaction is received from the user, and using the alpha band oscillation signals to determine the connectivity between the two regions of the brain; determining the change in the connectivity level between the two regions of the user's brain when an instance of data for a second task is presented (wherein the first condition of the second task satisfies an active state, the second condition of the second task is affirmative and the third condition of the second task is affirmative) to the connectivity level between the two regions of the user's brain when an instance of data for the first task is presented (wherein the first condition of the first task satisfies an active state and the second condition of the first task is affirmative); and comparing the change in the user's connectivity level with the change in the connectivity level of a user in a healthy control group, wherein if the change in the user's connectivity level is different from the change in the connectivity level of the user in the healthy control group, the user indicates ADHD.

[0349] Example 20: According to the non-transitory computer-readable medium described in Example 13, alpha, beta, and theta band oscillation signals of two or more of the user's multiple brain regions are filtered when a first reaction and a second reaction are received, and the alpha, beta, and theta band oscillation signals are used to determine the connectivity between the brain regions; and when the first condition, the second condition of the first task and the first condition, the second condition, and the third condition of the second task are changed, the probability of detecting ADHD is evaluated based on the comparison of the connectivity level, the reduction in connectivity level, and the change in connectivity level between the user's brain regions with the connectivity level of users in a healthy control group.

[0350] Example 21: A computer-implemented method for automatically detecting attention deficit hyperactivity disorder (ADHD), the method comprising the following steps: providing one or more stimuli to a user to activate multiple brain regions representing an attention network, wherein the one or more stimuli include displaying information having a first condition that causes a response and a second condition that is an exception to the first condition; obtaining multiple signals from the multiple brain regions for each of the one or more stimuli, wherein each of the multiple signals is accessed within a time period, wherein the time period begins when the information is displayed and ends when the user's response is captured; and evaluating the multiple signals based on a level of connectivity in the attention network to detect ADHD when each of the multiple signals corresponds to the one or more stimuli.

[0351] Example 22: An attention deficit detection system, comprising: one or more memory devices storing processor-executable instructions; and one or more processors configured to execute the instructions to cause the attention deficit detection system to perform operations, the operations comprising: providing one or more stimuli to activate multiple brain regions representing an attention network, wherein the one or more stimuli include displaying information having a first condition that causes a response and a second condition that is an exception to the first condition; obtaining multiple signals from the multiple brain regions for each of the one or more stimuli, wherein each of the multiple signals is accessed within a time period, wherein the time period starts when the information is displayed and ends when a user's response is captured; and when a signal of the multiple signals corresponds to the one or more stimuli, evaluating the multiple signals based on a level of connectivity within the attention network to detect ADHD.

[0352] Example 23: An attention deficit detection system, comprising: one or more memory devices storing processor-executable instructions; and one or more processors configured to execute the instructions to cause the attention deficit detection system to perform operations, the operations comprising: receiving first signal data from a user; determining a first plurality of time series based on the first signal data, wherein each time series in the first plurality of time series corresponds to a corresponding source position located within the cranial cavity of the user; calculating a first correlation value for a first pair of time series, the first pair of time series being included in the determined first plurality of time series; generating a score based on the first correlation value, the score indicating a patient with a cognitive disorder (such as attention deficit or ADHD); and outputting the generated score.

[0353]

[0354]

[0355]

[0356] Table 3: Automated anatomical labeling atlas (e.g., for Figures 4 to 12 ).

[0357] N.Tzourio-Mazoyer; B.Landeau; D.Papathanassiou; F.Crivello; O.Etard; N.Delcroix; Bernard Mazoyer&M.Joliot (January 2002). "Automated Anatomical Labeling of activations in SPM using a Macroscopic Anatomical Parcellation of the MNI MRI single-subject brain".NeuroImage.15(1):273–289.doi:10.1006 / nimg.2001.0978.PMID 11771995.

Claims

1. A device (140, 200), the device being configured to: Receiving (410) first electroencephalogram (EEG) or magnetoencephalogram (MEG) data (130) of a subject (110); A first plurality of time series is determined (420) based on the first EEG / MEG data, wherein: Each time series in the first plurality of time series corresponds to a respective source location within a cranial cavity (300) of the subject; calculating (430) a first correlation value for a first pair of time series, the first pair of time series being included in the determined first plurality of time series; generating (440) one or more scores (150) based on the first correlation value, the one or more scores indicating that the subject has a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD); as well as The generated one or more scores are output (450).

2. The device (140, 200) according to claim 1, in, The corresponding source position of the first time series in the first pair of time series is located in the frontal lobe (310) in the cranial cavity (300) of the subject (110), such as the middle frontal gyrus MFG (320) and / or the inferior frontal gyrus IFG (330), and Wherein, the corresponding source position of the second time series in the first pair of time series is located in the superior parietal lobe SPL (340) in the cranial cavity of the subject.

3. The device (140, 200) according to claim 1 or 2, further configured to: calculating a second correlation value for a second pair of time series, the second pair of time series being included in the determined first plurality of time series; as well as wherein the one or more scores are generated based on the first correlation value and the second correlation value; and / or wherein the corresponding source position of the first time series in the second pair of time series is located in a ventral attention network region in the cranial cavity (300) of the subject (110), such as the temporoparietal junction (TPJ) (350), and Wherein, the corresponding source position of the second time series in the second pair of time series is located in the dorsal attention network area in the cranial cavity of the subject, such as the superior parietal lobe SPL (340).

4. The device (140, 200) according to any one of claims 3, further configured to: calculating a third correlation value for a third pair of time series, the third pair of time series being included in the determined first plurality of time series; and in, generating the one or more scores based on the first correlation value, the second correlation value, and the third correlation value; and / or wherein the corresponding source position of the first time series in the third pair of time series is located in the ventral attention network region located in the cranial cavity (300) of the subject (110), such as the temporoparietal junction (TPJ) (350), and Wherein, the corresponding source position of the second time series in the second pair of time series is located in the ventral visual pathway area in the cranial cavity of the subject, such as the inferior temporal gyrus ITG (360).

5. The device (140, 200) according to any one of claims 1 to 4, further configured to: receiving second EEG / MEG data of the subject (110); A second plurality of time series is determined based on the second EEG / MEG data, wherein: Each time series in the second plurality of time series corresponds to a respective source location within a cranial cavity (300) of the subject; calculating a fourth correlation value for a fourth pair of time series, the fourth pair of time series being included in the determined second plurality of time series and the fourth pair of time series corresponding to the same respective source locations as the first pair of time series; calculating a comparison value between the fourth correlation value and the first correlation value; as well as Therein, the one or more scores are generated based on the comparison value (150).

6. The device according to any one of claims 1 to 5, in, Determining (420) the first plurality of time series includes filtering the first EEG / MEG data (130) so that the first plurality of time series includes only a subset of all available time series in the first EEG / MEG data, the subset including the first pair of time series, the second pair of time series, and the third pair of time series.

7. A system (100), comprising: The device (140) according to any one of claims 1 to 6; an electroencephalogram (EEG) or magnetoencephalogram (MEG) device (120) for measuring at least the first EEG / MEG data (130) of the subject (110); and A task presenting device (160) is used to present one or more tasks to the subject while the EEG / MEG device is measuring at least the first EEG / MEG data (130) of the subject.

8. A computer-implemented method (400), the method comprising the steps of: Receiving (410) first electroencephalogram (EEG) or magnetoencephalogram (MEG) data (130) of a subject (110); determining (420) a first plurality of time series based on the first EEG / MEG data, wherein each time series in the first plurality of time series corresponds to a respective source location within a cranial cavity (300) of the subject; calculating (430) a first correlation value for a first pair of time series, the first pair of time series being included in the determined first plurality of time series; generating (440) one or more scores (150) based on the first correlation value, the one or more scores indicating that the subject has a cognitive condition, such as a cognitive disorder, such as attention deficit hyperactivity disorder (ADHD); and The generated one or more scores are output (450).

9. The method (400) according to claim 8, in, The corresponding source position of the first time series in the first pair of time series is located in the frontal lobe (310) in the cranial cavity (300) of the subject (110), such as the middle frontal gyrus MFG (320) and / or the inferior frontal gyrus IFG (330), and Wherein, the corresponding source position of the second time series in the first pair of time series is located in the superior parietal lobe SPL (340) in the cranial cavity of the subject.

10. The method (400) according to claim 8 or 9, further comprising: calculating a second correlation value for a second pair of time series, the second pair of time series being included in the determined first plurality of time series, and wherein the one or more scores are generated based on the first correlation value and the second correlation value; and / or wherein the corresponding source position of the first time series in the second pair of time series is located in a ventral attention network region in the cranial cavity (300) of the subject (110), such as the temporoparietal junction (TPJ) (350), and Wherein, the corresponding source position of the second time series in the second pair of time series is located in the dorsal attention network area, for example, in the superior parietal lobe SPL (340) in the cranial cavity of the subject.

11. The method (400) of claim 10, further comprising: calculating a third correlation value for a third pair of time series, the third pair of time series being included in the determined first plurality of time series; as well as wherein the one or more scores are generated based on the first correlation value, the second correlation value and the third correlation value; and / or wherein the corresponding source position of the first time series in the third pair of time series is located in the ventral attention network area, for example, in the temporoparietal junction TPJ (350) in the cranial cavity (300) of the subject (110), and Wherein, the corresponding source position of the second time series in the second pair of time series is located in the ventral visual pathway area in the cranial cavity of the subject, such as the inferior temporal gyrus ITG (360).

12. The method (400) according to any one of claims 8 to 11, further comprising: - receiving second EEG / MEG data of the subject (110), determining a second plurality of time series based on the second EEG / MEG data, wherein each time series in the second plurality of time series corresponds to a respective source location within a cranial cavity (300) of the subject (110), calculating a fourth correlation value for a fourth pair of time series, the fourth pair of time series being included in the determined second plurality of time series and the fourth pair of time series corresponding to the same respective source locations as the first pair of time series, calculating a comparison value between the fourth correlation value and the first correlation value, and Therein, the one or more scores are generated based on the comparison value (150).

13. The method (400) according to any one of claims 8 to 12, in, Determining (420) the first plurality of time series includes filtering the first EEG / MEG data (130) so that the first plurality of time series includes only a subset of all available time series in the first EEG / MEG data, the subset including the first pair of time series, the second pair of time series, and the third pair of time series.

14. A computer readable medium comprising instructions which, when executed by a processor, perform the method (400) according to any one of claims 8 to 13.

Images