Systems, methods, and apparatus for neurological activity data analysis

Abstract

Electroencephalography (EEG) data may be analyzed to calculate various metrics, such as maximum amplitude projection, node visitation frequency, node transition frequency, and / or node transition polarity. In some examples, the calculated metrics may be provided graphically. In some examples, the metrics may be provided graphically in combination with other data, such as raw EEG traces.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims priority to U.S. Provisional Application No. 63 / 156,040, filed Mar. 3, 2021, which is hereby incorporated by reference in its entirety for any purpose.

[0002] Technical Field Examples described herein generally relate to the processing, analysis, and provision of neurological activity data.

[0003] Background Data regarding neurological activity can be collected from a subject in various ways. For example, electroencephalogram (EEG) data can be obtained by applying a plurality of probes (e.g., electrodes) to the subject's scalp. In some applications, the number of electrodes (e.g., channels) may be 19. However, in other applications, other numbers of probes may be used. Electrical signals within the subject's brain may be measured by the probes. The electrical signals can indicate neurological activity. In some applications, the electrical signals may be measured over time.

[0004] EEG data may be plotted in two - dimensions (2D). For example, the electrical signals measured by each probe may be plotted over time. Each plot may be referred to as a trace. An example of a 2D EEG trace is shown in Image A of FIG. 1. Reading a 2D EEG trace (e.g., a plot of the magnitude of electrical signals over time) to recognize seizures and determine seizure characteristics, for example, to estimate the neural activity underlying the seizure, requires years of training. Even experts typically cannot accurately identify the exact location of neural activity (e.g., a seizure) from only EEG data. This limits the diagnosis and test values of EEG traces.

[0005] Summary According to examples disclosed herein, electroencephalogram (EEG) data may be analyzed to calculate various metrics (also referred to as parameters), such as maximum amplitude projection, node visit frequency, node transition frequency, and / or node transition polarity. These metrics can be displayed, stored, and / or used to diagnose a patient, determine or adjust treatment plans, and / or take other actions.

[0006] According to examples of the present disclosure, the system may include at least one processor and memory accessible to at least one processor, the memory being encoded in computer-readable instructions, which, when executed, cause the system to calculate from neurological activity data at least one metric comprising maximum amplitude projection, node visit frequency, node transition frequency, node transition polarity, or a combination thereof. In some examples, the neurological activity data comprises electroencephalogram (EEG) data.

[0007] In some examples, the system may further include a display communicatively coupled to at least one processor, and when computer-readable instructions are executed, the system further causes the system to provide a graphic representation of at least one metric on the display. In some examples, the graphic representation of at least one metric comprises nodes of three-dimensional source localization overlaid on an image of the brain.

[0008] In some examples, the system may further include a display communicatively coupled to at least one processor, and when a computer-readable instruction is executed, the system may further cause the system to provide a report on the display, where the report comprises multiple contents having one or more graphic representations of at least one metric, further comprising a timeline of neurological activity data, one or more statistics, one or more additional parameters, or a combination thereof. In some examples, the system may further include an input device configured to receive user input, which indicates which of the multiple contents are included in the report.

[0009] In some examples, the method may involve receiving neurological activity data and calculating from the neurological activity data at least one metric having maximum amplitude projection, node visit frequency, node transition frequency, node transition polarity, or a combination thereof. In some examples, the neurological activity data has a three-dimensional (3D) source localization with multiple nodes, where each of the multiple nodes corresponds to a part of the brain.

[0010] In some examples, calculating the maximum amplitude projection of one node among several nodes may involve finding the number of times this node among several nodes was labeled as a local maxima within the analysis time window, and determining the maximum value of the local maxima from the number of times this node among several nodes was labeled as a local maxima.

[0011] In some examples, calculating the frequency of node visits for one node among several nodes involves counting the number of times this node among the multiple nodes is labeled as a local maxima within the analysis time window, and then dividing this count by the duration of the analysis time window.

[0012] In some examples, calculating node transition frequencies involves counting the number of transitions between a first local maximal node and a second local maximal node of multiple nodes within an analysis time window, and then dividing this number by the duration of the analysis time window.

[0013] In some examples, calculating the node transition polarity for one node in a group of nodes involves counting a first number of local maxima of transitions from one or more nodes in the group to this node, counting a second number of local maxima of transitions from this node to one or more nodes in the group, and taking the difference between the first and second counts. In some examples, this node in the group of nodes has inward polarity when the first count is greater than the second count, and outward polarity when the second count is greater than the first count.

[0014] In some examples, this method may further include preprocessing the neurological data by comparing the values ​​of the neurological activity data to a threshold, thereby comparing, discarding, or ignoring one or more of the values ​​of the neurological activity data, and by calculating at least one local maxima from the remaining values ​​of the neurological activity data.

[0015] In some examples, this method may further include providing a graphic representation of at least one metric on a display. In some examples, the graphic representation of at least one metric comprises a node of three-dimensional (3D) source localization overlaid on a brain image. In some examples, the size of the node is proportional to the value of the maximum amplitude projection or the value of the node visit frequency relative to the node. In some examples, the shading, color, or combination thereof of the node indicates the value of the node's maximum amplitude projection, node visit frequency, or node transition polarity. In some examples, the graphic representation further comprises a second node of 3D source localization and an arrow between the aforementioned node and the second node, the direction of which the arrow indicates a local maximal transition between the aforementioned node and the second node. In some examples, the shading, color, or combination thereof of the arrow indicates the frequency of transitions across an analysis time window.

[0016] In some examples, this method may further include providing a report on a display, which comprises multiple contents having one or more graphic representations of at least one metric, and further comprising a timeline of neurological activity data, one or more statistics, one or more additional parameters, or a combination thereof. [Brief explanation of the drawing]

[0017] [Figure 1A] This figure shows an example of EEG data visualization using the examples provided in this disclosure. [Figure 1B] This figure shows an example of EEG data visualization using the examples provided in this disclosure. [Figure 1C] This figure shows an example of EEG data visualization using the examples provided in this disclosure. [Figure 2] This figure shows two exemplary 3D source localizations from the same subject at different points in time, as illustrated by the examples of this disclosure. [Figure 3] This figure shows an example of 3D source localization with a set threshold, according to the examples of this disclosure. [Figure 4] A diagram showing an example of a local maximum identified in 3D source localization according to an example of the present disclosure. [Figure 5] A diagram illustratively showing an exemplary calculation of the maximum amplitude projection of nodes according to an example of the present disclosure. [Figure 6] A diagram showing an example of the result of the maximum amplitude projection of seizures according to an example of the present disclosure. [Figure 7] A diagram illustratively showing an exemplary calculation of the node visit frequency of nodes according to an example of the present disclosure. [Figure 8] A diagram showing the result of an exemplary node visit frequency of seizures according to an example of the disclosure. [Figure 9] A diagram illustratively showing an exemplary calculation of the node transition frequency for three nodes according to an example of the present disclosure. [Figure 10] A diagram showing the result of an exemplary node transition frequency of seizures according to an example of the disclosure. [Figure 11A] A diagram illustratively showing an exemplary calculation of the node transition polarity for a node according to an example of the present disclosure. [Figure 11B] A diagram illustratively showing an exemplary calculation of the node transition polarity for a node according to an example of the present disclosure. [Figure 12] A diagram showing the result of an exemplary node transition polarity of seizures according to an example of the disclosure. [Figure 13] A diagram showing an exemplary user interface for displaying source localization data presented with EEG data according to an example of the present disclosure. ]> [Figure 14] A diagram showing an exemplary user interface for displaying the maximum amplitude projection presented with EEG data according to an example of the present disclosure. [Figure 15] A diagram showing an exemplary user interface for displaying the node visit frequency presented with EEG data according to an example of the present disclosure. [Figure 16] A diagram showing an exemplary user interface for displaying the node transition frequency presented with EEG data according to an example of the present disclosure. [Figure 17]This figure shows an exemplary user interface for displaying a seizure analysis summary in a report, as illustrated by the examples in this disclosure. [Figure 18] This diagram schematically illustrates a system configured according to the examples in this disclosure. [Figure 19] This diagram shows a flowchart of the method according to the example in this disclosure.

[0018] Detailed explanation To provide a full understanding of the described embodiments, certain details are described below. However, it will be obvious to those skilled in the art that the embodiments can be carried out without these specific details. In some cases, well-known EEG techniques and systems, circuits, control signals, timing protocols, and / or software operations are not shown in detail to avoid unnecessarily obscuring the described embodiments.

[0019] As described above, electroencephalography (EEG) data can be obtained by applying multiple probes (e.g., electrodes) to the scalp of a subject. In some applications, the locations of these probes relative to the subject's brain and / or other probes may be known. The electrical signals recorded as EEG data by these probes may indicate neurological activity (e.g., electrical or electrochemical activity in the brain). The collected EEG data can then be visualized using various techniques.

[0020] Figure 1 shows an example of EEG data visualization according to an example of the present disclosure. First, EEG data is acquired from multiple probes as described above. The EEG data includes electrical signals acquired over time by each probe. An example of a 2D plot 100 of the electrical signals detected and recorded by each probe is shown in Image A of Figure 1. However, as described above, reading the EEG trace of plot 100 to understand the underlying neurological activity is difficult and time-consuming.

[0021] The electrical signals recorded by each probe may be used to generate a contour map of the scalp showing the intensity of the electrical signals in various regions. An exemplary contour map 104 is shown in Image B of Figure 1. Contour map 104 was generated from EEG data in plot 100 at points indicated by vertical lines 102. Points 106 indicate the locations of the probes on the subject's scalp. Contour lines 108 show the gradient of change in electrical signal intensity, as do contours on a topographical map showing the gradient of change in height. In some contour maps, such as contour map 104, shading in various tones and / or colors may be used to show the intensity and / or polarity of the recorded electrical signals. Contour map 104 provides an intuitive view of the data in plot 100 extrapolated to the surface of the subject's head, but will only provide a general understanding of the parts of the brain involved in electrical activity.

[0022] Based at least partially on electrical signals (e.g., the electrical signals shown in plot 100) and locations, electrical signals may be localized to one or more locations in the brain (e.g., the location of the origin of the brain signals may be determined and / or estimated). These locations may be referred to as source localization nodes or simply nodes. For example, a node may relate to a part of the brain. Nodes can be constructed using parts of any size. These nodes may form a three-dimensional (3D) grid corresponding to locations in the subject's brain. In some examples, the 3D grid may be a Cartesian grid where locations are described by referring to three axes (e.g., x-axis, y-axis, z-axis). However, in other examples, other grids / coordinate systems may be available. The size of the nodes in the 3D grid, the number of nodes, and / or the distances between nodes may be based on one or more factors. Illustrative factors include, but are not limited to, the sensitivity and / or specificity of the EEG probes, the number of EEG probes, and the signal-to-noise ratio of the EEG data. Each node in the 3D grid may have one or more values ​​associated with an electrical signal (e.g., magnitude, polarity) that belongs to that node, in order to form a 3D (volume) EEG dataset.

[0023] 3D datasets may be rendered as images for visualization. The 3D dataset and / or the resulting images may be referred to as 3D source localization. In some examples, 3D source localization may be rendered as an overlay on an image of the brain. This can provide a visual guide to which nodes correspond to which regions of the brain. In some examples, the brain may be rendered semi-transparently to enable visualization of the 3D source localization. In some examples, the brain image may be rendered from a model or from an image obtained from a subject (e.g., computed tomography, MRI). An example of source localization 110 is shown in image C of Figure 1. Source localization 110 was generated from EEG data in plot 100 at the point indicated by the vertical line 102. Nodes 112 are indicated by individual cubes. However, in other examples, nodes 112 may be visualized using other shapes (e.g., spheres, irregularly shaped volumes). In some examples, different tones and / or colors may be used for node 112 to indicate the intensity and / or polarity of the recorded electrical signals. The source localization 110 shown in Figure 1 is a top view, but the source localization 110 may be a 3D dataset that can be viewed from multiple angles and / or locations (e.g., views generated closer or further from the 3D dataset, views generated from within the 3D dataset, views of "intersections" of the volume of the 3D dataset, etc.).

[0024] Four-dimensional (4D) EEG data relates to the temporal observation of 3D EEG data. For example, a 3D image may show the electrical signals at each node for a given time point, and 4D EEG data may be a series of 3D images, each representing electrical activity at different time points. In some examples, each time point may be a period (e.g., microseconds, milliseconds, seconds) over which the electrical signals at each node are summed, averaged, or otherwise combined. 4D EEG data can be used to visualize source localization (e.g., 4D source localization), such as seizure source localization, which describes where a seizure begins and travels within the brain. Figure 2 shows two exemplary 3D source localizations from the same subject at different time points, according to an example of this disclosure. The two 3D source localizations 200 and 202 at time points 0 and 1 may represent a 4D seizure source localization or a portion of a 4D seizure source localization, respectively. In the example shown in Figure 2, various shades are used to represent the electrical signals belonging to individual nodes. The shading of multiple nodes changes from time 0 to time 1, illustrating how electrical activity in different locations within the brain changes over time.

[0025] 4D source localization can be more intuitive and easier to understand than 2D EEG tracing. However, generating understandable 4D visualizations requires appropriate values ​​for time range, field of view, viewing distance, and display threshold. In some applications, users may need to empirically determine these appropriate values. This can make preparing and loading 4D source localization data time-consuming.

[0026] According to embodiments of this disclosure, one or more metrics summarizing the 4D seizure source localization may be calculated. In some applications, this can reduce the time required to generate and / or interpret the 4D seizure source localization. According to examples disclosed herein, EEG data obtained from a subject (e.g., traces of electrical signals recorded over time by one or more electrodes) may be analyzed to calculate one or more metrics such as maximum amplitude projection, node visit frequency, node transition frequency, and / or node transition polarity. These metrics may then be provided in various formats, such as text, charts, generated images, or a combination thereof. The metrics may be used for diagnosis, to determine and / or adjust treatment, and / or to take other measures.

[0027] The examples disclosed herein relate to EEG data, but these are for illustrative purposes only, and embodiments are not limited to EEG data. Some or all of the techniques disclosed herein can be applied to other neurological activity data that can be associated with locations within the brain, such as functional magnetic resonance imaging (fMRI) data, cortical electroencephalography data, magnetoencephalography data, near-infrared spectroscopy data, and event-related optical signal data. Furthermore, while the examples disclosed herein relate to seizure localization, some or all of the techniques disclosed herein can be applied to the localization of other neural activities (e.g., responses to stimuli, responses to treatment).

[0028] In some examples, the 4D source localization data, i.e., the 3D dataset for each time point, may be preprocessed. In some examples, the preprocessing may involve calculating local maxima for individual source localizations at multiple time points, and in some examples, there may be multiple local maxima for individual source localizations at individual time points. In some examples, the local maxima may be based on the amplitude, magnitude, polarity, frequency, or combination thereof of electrical signals at a node. In some examples, the user may set a threshold for the minimum value of the local maxima. In other words, the neurological data may be preprocessed by comparing the values ​​of the neurological data with a threshold. Based on this comparison, some of the neurological data may be discarded or ignored. From the remaining (not discarded or ignored) data, local maxima of the neurological data can be calculated.

[0029] Figure 3 shows an example of thresholded 3D source localization according to the present disclosure. Thresholded 3D source localizations 300 and 302 are generated from 3D source localizations 200 and 202 shown in Figure 2, respectively, based on local maxima minimums. In the example shown in Figure 3, 3D source localization 300 includes region 304 that satisfies or exceeds a local maxima minimum, and 3D source localization 302 includes regions 306 and 308 that satisfies or exceed a local maxima minimum indicated by the shaded portion. In some examples, data related to local maxima that do not satisfy or exceed a local maxima minimum may be discarded or ignored in further calculations.

[0030] Figure 4 shows an example of local maxima identified in 3D source localization according to an example of the present disclosure. 3D source localizations 400 and 402 are generated from thresholded 3D source localizations 300 and 302, respectively, as shown in Figure 3. The highlighted node 404 indicates a local maxima in region 304, and the highlighted nodes 406 and 408 indicate local maxima in regions 306 and 308, respectively.

[0031] In some examples, various metrics may be calculated after preprocessing, such as the preprocessing described with reference to Figures 3 and 4. The graphical representation of the results of the metric calculations may be displayed as one or more images. For example, one or more nodes corresponding to nodes in 4D source localization may be provided as a sphere, cube, or other graphical representation in an image. The nodes may be located in the image at locations corresponding to the node locations when displayed together with the entire 3D grid of source localization. In some examples, the nodes may be overlaid on an image of a brain. The image of the brain may be a rendering of a subject's brain or a brain model.

[0032] In some examples, a maximum amplitude projection may be calculated for 4D source localization. The maximum amplitude projection can indicate brain regions with large amplitude electrical signals, for example, across different points in time during a seizure. The maximum amplitude projection may be calculated by analyzing individual source localization nodes of the 3D source localization over time. The period for analyzing the 3D source localization (e.g., the analysis time window) may extend to the entire period covered by the 4D source localization (e.g., the entire period over which the EEG trace was acquired) or a subset of this period. Examples of analysis time windows include 1 millisecond, 5 milliseconds, 10 milliseconds, 20 milliseconds, 50 milliseconds, 100 milliseconds, 1 second, and 30 seconds. Other analysis time windows may be used in other examples.

[0033] If a node is labeled as a local maximal at least once within the analysis time window (e.g., nodes 404, 406, 408), the maximum amplitude projection will output the maximum source localization amplitude for this node within the analysis window. If a node is not labeled as a local maximal at any point within the analysis time window, no output will be provided for this node. In some embodiments, the output of the maximum amplitude projection may have the same units as those used for 3D source localization (e.g., arbitrary units for standardized low-resolution brain electromagnetic tomography (sLORETA) source localization, picoamperes-meters (pA-m) for minimal norm estimation (MNE) source localization).

[0034] Figure 5 illustrates an exemplary calculation of the maximum amplitude projection of a node according to an example of the present disclosure. Node 500 is labeled as a local maxima at time points 0 milliseconds, 10 milliseconds, 15 milliseconds, and 20 milliseconds within the analysis time window. The values ​​of node 500 for the local maxima are 0.5, 1.5, 0.9, and 1.2, respectively. The maximum value of node 500 within the analysis time window is 1.5, occurring at time point 10 milliseconds. Therefore, the output of the maximum amplitude projection for node 500 is 1.5. Although the exemplary calculation in Figure 5 is for one node, it will be understood that the procedure shown is performed at least once for all nodes, or for all nodes labeled as local maxima, within the analysis time window.

[0035] Figure 6 shows an example of the results of seizure maximum amplitude projection according to an example of the present disclosure. Result 600 is shown in left, right, posterior, anterior, inferior, and superior views of the same brain (L: left side of the brain, R: right side of the brain). In the example shown in Figure 6, the size of node 602 is quadratically proportional to its value (e.g., maximum amplitude), and the shading of node 602 corresponds to its value indicated by legend 604. However, in other examples, other visualization techniques may be used to convey the value of the maximum amplitude projection (e.g., color, shape, text indicating numerical values).

[0036] While the maximum amplitude is described, other amplitude projections can be calculated similarly (for example, based on a threshold selected by the user). The maximum amplitude projection can provide information about the intensity and / or most responsive brain regions during a particular neurological event (e.g., a seizure).

[0037] In some examples, node visit frequency can be calculated for 4D source localization. Node visit frequency can indicate brain regions that are frequently active during a given period. Node visit frequency can provide insight into which brain regions are most frequently active during neurological events. Node visit frequency may be calculated over a given analysis time window for source localization nodes, such as source localization nodes associated with local maxima. For individual source localization nodes, the number of times this node is labeled as a local maxima can be counted. In some examples, the output may be a number of counts (e.g., visits). In some examples, this number may then be divided by the duration of the analysis time window. In these examples, the output may have units of counts per unit time (e.g., milliseconds, seconds).

[0038] Figure 7 illustrates an exemplary calculation of node visit frequency for a node according to an example of the present disclosure. Node 700 is labeled four times as a local maxima at time points 0, 10, 15, and 20 milliseconds within an analysis time window with a duration of 20 milliseconds. The count (4) is obtained by dividing the node visit frequency of 200 counts / second by the duration of the analysis time window (0.02 seconds).

[0039] Figure 8 shows the results of exemplary node visit frequency for seizures, according to an example of disclosure. The seizure used to generate exemplary result 800 is the same seizure used to generate exemplary result 600. Result 800 is shown in left, right, posterior, anterior, inferior, and superior views of the same brain. The size of node 802 is quadratically proportional to its value (e.g., counts / second), and the shading of node 802 corresponds to its value, indicated by legend 804. However, in other examples, other visualization techniques may be used to convey the node visit frequency values. For example, text indicating the value may be provided within or next to the node.

[0040] In some cases, node transition frequencies can be calculated for 4D source localization. Node transition frequencies can indicate the order in which nodes, particularly frequently active nodes, become active. The sequence in which nodes become active can provide information about how various brain regions interact and / or are connected. For example, it can provide insights into potential therapeutic goals, such as disrupting communication between two regions for suppressing or reducing the intensity of seizures and / or other undesirable neurological events. Node transition frequencies may be calculated over an analysis time window for source localization nodes, such as local maxima.

[0041] A node transition occurs when a characteristic changes from one node to another between multiple time points in time. For example, if a local maximal node at one time point is different from a local maximal node at a different time point, the activity associated with this local maximal is called a "transition" from one node to another. A node transition connection is a transition from one node at a first time point (e.g., one local maximal node) to another node at a second time point (e.g., another local maximal node). This connection is indicated by an arrow that starts at the node at the first time point and ends at the node at the second time point. If there are multiple nodes (e.g., multiple local maximals) at one or more time points, the two nodes with the shortest Euclidean distance will be connected. For example, nodes A and B are local maximals at time point 0, and nodes C and D are local maximals at time point 1. If node A is closer to node C than node D (based on Euclidean distance), and node A is closer to node C than node B, then an arrow will be drawn between node A and node C. If the same node exists at both time points (for example, if it has multiple local maxima at both time points), no arrow will be drawn between this node and the other node.

[0042] In addition to providing arrows to indicate transitions, numerical results of node transition frequencies may be provided. After analyzing all points in the analysis time window, the number of times a transition from one node to another occurred within the analysis time window is counted. The result may be the number of counts (e.g., transitions). In some examples, this count may be divided by the duration of the analysis time window, and the result may be the number of counts divided by time (counts / second).

[0043] Figure 9 illustrates an exemplary calculation of node transition frequencies for three nodes according to an example of the present disclosure. While the example in Figure 9 involves three nodes, other examples may use a different number of nodes (e.g., two, four, five, etc.). The first node, 900, is labeled as a local maximum at time points 0 and 10 milliseconds. A second node, 902, corresponding to a different brain region than node 900, is labeled as a local maximum at time point 15 milliseconds. A third node, 904, corresponding to a different brain region than nodes 900 and 902, is labeled as a local maximum at time point 20 milliseconds. Both graphical and numerical results may be provided for node transition frequencies. As shown in Figure 910, arrow 906 indicates the transition of a local maxima from node 900 to node 902 between time points 10 and 15 milliseconds, and arrow 908 indicates the transition of a local maxima from node 902 to node 904 between time points 15 and 20 milliseconds. The direction of the arrows indicates the temporal order of the transitions. For example, the initial local maxima was at node 900, and subsequently, the local maxima was at node 902. Therefore, the origin of the arrow is at node 900, and the tip of the arrow is at node 902.

[0044] Additionally, the number of transitions between node 900 and node 902 is counted (1), and this is divided by the analysis time window (0.02 seconds) to provide a node transition frequency value (100 counts / second). This value is also provided for the node transition frequency between node 902 and node 904, which, coincidentally, is the same as the node transition frequency between node 900 and node 902 in this example (100 counts / second). However, in another example, the node transition frequencies between different nodes may be different.

[0045] Figure 10 shows the results of exemplary node transition frequencies for a seizure, according to an example of disclosure. The seizure used to generate exemplary result 1000 is the same seizure used to generate exemplary result 600. Result 1000 is shown in left, right, posterior, anterior, inferior, and superior views of the same brain. Arrows 1002 indicate nodes involved in one or more transitions, and the direction of the arrows indicates the temporal order of the node transitions. In the example shown in Figure 10, the shading of the arrows indicates numerical values ​​of the node transition frequencies indicated by legend 1004. However, in other examples, other visualization techniques may be used to convey the values ​​of the node transition frequencies. For example, the frequencies may be provided as text above the arrows.

[0046] In some cases, node transition polarity may be calculated for 4D source localization. Node transition polarity may be based at least partially on node transition frequency. Node transition polarity may provide information about the mean inward / outward polarity of a brain region during activity transitions. That is, node transition polarity can indicate whether a brain region is more often a "receiver" of input from other brain regions or more often a "sender" of output to other brain regions. In some cases, when this brain region is a sender, activity may be triggered in other brain regions, and when this brain region is a receiver, activity may be shown in response to input from other brain regions. It may also provide insights into potential therapeutic goals, for example, preventing a brain region from sending and / or receiving signals for suppressing or reducing the intensity of seizures and / or other undesirable neurological events.

[0047] For each node connected to another node by an arrow during the calculation of node transition frequency, as described with reference to Figures 9 and 10, the number of arrows to this node (e.g., activities transitioning from another node to this node) and the number of arrows from this node (e.g., activities transitioning from this node to another node) may be counted. Positive or negative values ​​can be assigned to the direction of the arrows. For example, arrows to this node may be positive, and arrows from this node may be negative. The total number of arrows (e.g., connections) can be summed up. If the sum of all connections to this node is positive (e.g., there are more arrows to this node), the polarity of this node may be inward. If the sum of all connections to this node is negative (e.g., there are more arrows from this node), the polarity of this node may be outward.

[0048] Figure 11 illustrates an exemplary calculation of node transition polarity for two nodes according to an example of the present disclosure. It was found that node 1100 transitioned 12 times to a node (not shown) during the previous calculation of node transition frequencies, and transitioned 1, 5, and 10 times from three nodes (not shown) during the previous calculation of node transition frequencies. Thus, node 1100 had a total of 12 departure connections and 16 arrival connections. Subtracting 16 from 12 gives -4. Thus, node 1100 has an inward polarity of magnitude 4.

[0049] It was found that node 1102 transitioned 12 times to one node (not shown) and 3 times to another node (not shown) during the previous calculation of node transition frequencies, and transitioned once and 4 times from two other nodes (not shown) during the same period. Thus, node 1102 had a total of 15 departure connections and 5 arrival connections. Subtracting 5 from 15 gives 10. Therefore, node 1102 has an outward polarity of magnitude 10.

[0050] Figure 12 shows the exemplary node transition polarity results of a seizure according to an example of the present disclosure. The seizure used to produce exemplary result 1200 is the same seizure used to produce exemplary result 600. Result 1200 is shown in left, right, posterior, anterior, inferior, and superior views of the same brain. In the example shown in Figure 12, the shading of the nodes indicates their polarity and size, according to legend 1204. However, in other examples, other visualization techniques may be used to convey the values ​​of node transition polarity. For example, different colors may be used for different polarities and / or sizes. In another example, different shapes may be used for different polarities (e.g., "×" for inward polarity and "○" for outward polarity).

[0051] Various metrics, such as maximum amplitude projection, node visit frequency, node transition frequency, and / or node transition polarity, can be calculated from 4D source localization, and the results can be provided numerically and / or as images, as described with reference to Figures 5 to 12. Data from metrics and their visualizations may be provided via a user interface. In some examples, metric data and visualizations may be displayed on the user interface simultaneously with other data such as 2D EEG plots (e.g., plot 100), contour maps (e.g., contour map 104), 3D source localizations (e.g., 3D source localization 110), 4D source localizations (e.g., source localizations 200, 202), or combinations thereof. The user interface may be included with and / or communicate with the computing device. For example, the user interface may be provided on the screen or touchscreen of the computing device.

[0052] Figure 13 is an exemplary user interface for displaying source localization data shown together with EEG data, as illustrated by an example of the present disclosure. User interface 1300 provides a plot 1302 of an EEG trace over time and a graphic drawing 1304 of the subject's scalp, which is overlaid with a color gradient showing the maximum deviation and / or standard deviation of electrical signals on the scalp over various time periods (e.g., every second). This graphic drawing 1304 is similar to the contour map 104 in Figure 1. In box 1306, a 3D rendering of a model brain is shown with a 3D source localization overlaid, which has a color gradient to show a source localization estimate of the brain's electrical activity on the surface of the brain, in addition to various user controls. This is similar to the 3D source localization 110 in Figure 1 and the source localizations 200 and 202 in Figure 2. The color gradient may change at various points in time as the electrical activity in the brain changes. Therefore, the 3D rendering may, in some cases, be a series of images over time (e.g., a sequence, a video) rather than a static image to provide 4D source localization. User controls allow the user to control various display functions, such as the speed of the video, the threshold of the source values ​​displayed, the transparency of the brain, the size of the displayed voxels, etc. Below box 1306, a visualization of the maximum amplitude projection 1308 is provided. The maximum amplitude projection 1308 includes an additional 3D rendering of the model brain shown with circles representing nodes overlaid, where these nodes are color-coded to qualitatively indicate values ​​calculated for the maximum amplitude projection for various nodes. Similar to source localization estimation, the nodes and / or their color coding may change over various durations (e.g., every second, throughout the seizure, or within a user-selected time range of interest) and may be provided as a video. The maximum amplitude projection 1308 may be generated in some examples as described with reference to Figures 5 and 6.

[0053] Users can select some or all of the provided data by providing input through the user interface.

[0054] Figure 14 is an exemplary user interface for displaying maximum amplitude projections shown with EEG data, according to an example of the present disclosure. Similar to Figure 13, the user interface 1400 provides a plot 1402 of EEG traces over time and a graphic drawing 1404 of the subject's scalp, the graphic drawing 1404 being overlaid with a color gradient showing the maximum deviation and / or standard deviation of electrical signals in the scalp over various time periods. A 3D rendering of a model brain is shown, overlaid with circles representing nodes, where these nodes are color-coded to qualitatively show values ​​calculated for maximum amplitude projections for various nodes. The nodes and / or their color coding may change for various time points, and therefore the 3D rendering may be a series of images over time, frame by frame and / or observed as a video. User controls allow the user to control various display functions, such as the speed of the video, the threshold of the maximum amplitude values ​​displayed, the transparency of the brain, the voxel size, and so on.

[0055] Figure 15 shows an exemplary user interface for displaying node visit frequencies shown with EEG data, according to an example of the present disclosure. User interface 1500 provides a plot 1502 of EEG traces over time 1502 and a graphic drawing 1504 of the subject's scalp, similar to Figures 13 and 14, with a color gradient overlaid on the graphic drawing 1504 showing the maximum deviation and / or standard deviation of electrical signals in the scalp over various time periods. A 3D rendering of a model brain is shown, as indicated by box 1508, overlaid with circles representing nodes, where these nodes are color-coded to qualitatively show values ​​calculated for node visit frequencies for various nodes. The nodes and / or their color coding may change for various time points, and therefore the 3D rendering may be a series of images over time, which may be provided for observation frame by frame and / or as a video. In some examples, node visit frequencies may be calculated as described with reference to Figures 7 and 8. User controls allow users to control various display functions, such as video speed, the threshold for the frequency of node visits (counts / second), brain transparency, and voxel size.

[0056] Figure 16 is an exemplary user interface for displaying node transition frequencies shown with EEG data, according to an example of the present disclosure. User interface 1600 provides a plot 1602 of EEG traces over time 1602 and a graphic drawing 1604 of the subject's scalp, similar to Figures 13–15, with a color gradient overlaid on the scalp showing the mean / maximum electrical signals over various time periods. A 3D rendering of a model brain is shown, overlaid with vectors, as indicated by box 1608, where these vectors are color-coded to qualitatively show values ​​calculated for node transition frequencies for various nodes (e.g., pairs of nodes). The vectors and / or their color coding may change for various time points, and therefore the 3D rendering may be a series of images over time, which may be provided for observation frame by frame and / or as a video. In some examples, node transition frequencies may be calculated as described with reference to Figures 9 and 10. User controls allow users to control various display functions, such as video speed, node visit polarity threshold (counts / second), brain transparency, and voxel size.

[0057] Figures 13–16 show only one graphical representation of a calculated parameter (e.g., a metric), but in other examples, the user interface may allow the user to visualize multiple parameters simultaneously (e.g., graphical representations of both maximum amplitude projection and node visit frequency may be provided on the user interface). Furthermore, in some examples, the user interface may provide a report summarizing the calculated parameters.

[0058] Figure 17 shows an exemplary user interface for displaying a seizure analysis summary in a report, according to an example of the present disclosure. User interface 1700 provides a summary of one or more parameters calculated from the EEG trace as a report 1702. Report 1702 may include a timeline 1704 of epilepsy-like activity, various numerical statistics 1706 (e.g., number of seizures, mean duration, number of spikes), one or more graphic representations 1708 of metrics (e.g., maximum amplitude projection, node visit frequency, spike group) overlaid on brain images, one or more charts 1710 of other parameters (e.g., spike distribution) and / or combinations thereof. In some examples, the user may be able to select the content of report 1702. For example, the user may provide input via one or more input devices to indicate which content to include in report 1702.

[0059] In some cases, EEG data can be received and analyzed by a computing system. The computing system can calculate various metrics and provide various outputs, such as the calculated metrics, to a display communicatively coupled to the computing system.

[0060] Figure 18 is a schematic diagram showing a system configured according to an example of the present disclosure. System 1800 includes a computing system 1806, a processor 1808, executable instructions 1810 for calculating metrics / parameters and generating display information, memory 1812, a display 1814, and a network interface 1816. System 1800 can receive EEG data 1802 (e.g., an EEG as shown in plot 100). In other examples, system 1800 may include additional components, fewer components, and / or other components. For example, in some examples, system 1800 may include an EEG system for acquiring EEG data 1802. In other embodiments, the EEG data 1802 may further or alternatively include one or more other types of neurological activity data, such as fMRI data.

[0061] The computing system 1806 may analyze the EEG data 1802 to calculate one or more metrics and / or to generate display information for providing the metrics to a display 1814, for example, as described with reference to Figures 3 to 17. In some examples, memory 1812 may be encoded by executable instructions 1810 for calculating metrics and / or generating display information for the metrics. In some examples, memory 1812 may be further encoded by executable instructions 1810 for providing a graphical user interface (GUI) that can provide text, graphics and / or other visual information on the display 1814.

[0062] The computing system 1806 may include one or more processors 1808. The processors 1808 can be implemented, for example, using one or more central processing units (CPUs), graphics processing units (GPUs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other processor circuits. In some examples, the processors 1808 can execute some or all of the executable instructions 1810. The processors 1808 can communicate with memory 1812. Memory 1812 may generally be implemented by any non-temporary computer-readable medium (e.g., read-only memory (ROM), random-access memory (RAM), flash memory, solid-state drives, etc.). Although one memory 1812 is shown, any number can be used, and these may be integrated with the processors 1808 in one computing system 1806 and / or located in another computing system to communicate with the processors 1808. For example, EEG data 1802 may be contained in one computer-readable medium of memory 1812, and executable instructions 1810 may be contained in another computer-readable medium of memory 1812. Some or all of the medium contained in memory 1812 may be accessible to the processor 1808.

[0063] In some examples, system 1800 may include a display 1814 that can communicate with computing system 1806 (for example, using wired and / or wireless connections), or the display 1814 may be integrated with computing system 1806. Display 1814 can display one or more metrics / parameters calculated by computing system, one or more graphics based on metrics / parameters, EEG data 1802, and / or one or more GUI elements. For example, images, data, and / or displays such as those shown in Figures 1-4, 6, 8, 10, and / or Figures 12-17 may be provided to display 1814. Any number of displays or a variety of displays may be present, including one or more LEDs, LCDs, plasma or other display devices.

[0064] In some examples, system 1800 may include a network interface 1816. The network interface 1816 can provide a communication interface to any network (e.g., LAN, WAN, Internet). The network interface 1816 may be implemented using wired and / or wireless interfaces (e.g., Wi-Fi, Bluetooth, HDMI, USB, etc.). The network interface 1816 can communicate data that may include EEG data 1802 and / or metrics calculated by computing system 1806.

[0065] In some examples, system 1800 may include various input devices 1818 that can be configured to receive input from a user. Examples of input devices include, but are not limited to, a mouse, keyboard, trackball, joystick, touchpad, and touchscreen. In some examples, a display 1814 may be an input device (for example, if display 1814 is a touchscreen). In some examples, the input devices 1818 and / or display 1814 may be included in a user interface 1820, which may be at least partially a graphical user interface. In some examples, a user may be able to provide input to actions performed by the computing system 1806 via one or more of the input devices 1818. For example, a user may provide input to determine which metrics are calculated and / or what metrics are provided on display 1814.

[0066] Figure 19 is a flowchart of a method according to an example of this disclosure. In some examples, the method shown in flowchart 1900 can be implemented in whole or in part by a computing system, such as computing system 1800.

[0067] In block 1902, "receiving neurological activity data" may be performed. In some embodiments, the neurological activity data may be EEG data received from an EEG device. In other embodiments, the neurological activity data may be fMRI data received from an MRI system. In other embodiments, other data types may be received from other devices. In some embodiments, the neurological activity data may include a three-dimensional (3D) source localization including multiple nodes, where each node corresponds to a part of the brain.

[0068] In block 1904, the task may be to "calculate at least one metric from neurological activity data." In some examples, at least one metric may include maximum amplitude projection, node visit frequency, node transition frequency, node transition polarity, or a combination thereof.

[0069] As shown in blocks 1906 and 1908, calculating the maximum amplitude projection of one node among multiple nodes may involve "finding the number of times this node among multiple nodes was labeled as a local maxima within the analysis time window" and "determining the maximum value of the local maxima from the number of times this node among multiple nodes was labeled as a local maxima."

[0070] As shown in blocks 1910 and 1912, calculating the node visit frequency for one node among several nodes may involve "counting the number of times this node among several nodes was labeled as a local maxima within the analysis time window" and "dividing this number by the duration of the analysis time window."

[0071] As shown in blocks 1914 and 1916, calculating node transition frequencies can involve "counting the number of local maximal transitions between a first node and a second node among several nodes within the analysis time window" and "dividing this number by the duration of the analysis time window."

[0072] As shown in blocks 1918, 1920, and 1922, calculating the node transition polarity for one node among multiple nodes may involve "counting a first number of local maxima transitions from one or more nodes among multiple nodes to this node among multiple nodes," "counting a second number of local maxima transitions from this node among multiple nodes to one or more nodes among multiple nodes," and "taking the difference between the first and second counts." In some examples, this node among multiple nodes has inward polarity when the first count is greater than the second count, and outward polarity when the second count is greater than the first count.

[0073] In block 1924, the action may be to "provide a graphic representation of at least one metric on the display." For example, the graphic representation may include graphic representations of metrics such as those shown in Figures 6, 8, 10 and / or Figures 12-17.

[0074] In some examples, the method shown in flowchart 1900 may further include preprocessing of the neurological activity, which may include comparing values ​​of the neurological activity data with thresholds, discarding or ignoring one or more of the values ​​of the neurological activity data based on this comparison, and calculating at least one local maxima from the remaining values ​​of the neurological activity data.

[0075] To summarize 4D source localization, such as 4D seizure source localization, one or more metrics can be calculated as disclosed herein. In some applications, this can reduce the time required to generate and / or interpret 4D source localization. Metrics such as maximum amplitude projection, node visit frequency, node transition frequency, and / or node transition polarity can be calculated as disclosed herein. These metrics may then be provided in various formats, such as text, charts, generated images, or a combination thereof. The metrics may be used for diagnosis, to determine and / or adjust treatment, and / or to take other measures.

[0076] From the above description, it will be understood that while specific embodiments are described herein for illustrative purposes, various modifications can be made within the scope of the technology described in the claims.

Claims

1. A system, wherein the system is At least one processor, The system comprises memory accessible to at least one processor, the memory being encoded with computer-readable instructions, and the computer-readable instructions, when executed, cause the system to calculate at least one metric comprising node visit frequency from neurological activity data. The aforementioned neurological activity data includes a three-dimensional (3D) source localization comprising multiple nodes, where each of the multiple nodes corresponds to a part of the brain. Calculating the node visit frequency for one of the aforementioned multiple nodes is: The number of times that one of the aforementioned nodes is labeled as a local maximum within the analysis time window, The number of times is divided by the duration of the analysis time window, Equipped with, system.

2. The system according to claim 1, wherein the neurological activity data includes electroencephalogram (EEG) data.

3. The system according to claim 1, further comprising a display communicatively coupled to the at least one processor, wherein the computer-readable instruction, when executed, causes the system to provide a graphic representation of the at least one metric on the display.

4. The system according to claim 3, wherein the graphic representation of the at least one metric comprises a node for three-dimensional source localization that is overlaid on an image of the brain.

5. The system according to claim 1, further comprising a display communicatively coupled to the at least one processor, wherein when a computer-readable instruction is executed, the system further causes the system to provide a report on the display, the report comprising a plurality of contents comprising one or more graphic representations of the at least one metric, and further comprising a timeline of the neurological activity data, one or more statistics, one or more additional parameters, or a combination thereof.

6. The system according to claim 5, further comprising an input device configured to receive user input, the user input indicating which of the plurality of contents is included in the report.

7. Receiving neurological activity data, From the aforementioned neurological activity data, calculate at least one metric that includes node visit frequency, Equipped with, The aforementioned neurological activity data includes a three-dimensional (3D) source localization comprising multiple nodes, where each of the multiple nodes corresponds to a part of the brain. Calculating the node visit frequency for one of the aforementioned multiple nodes is: The number of times that one of the aforementioned nodes is labeled as a local maximum within the analysis time window, The number of times is divided by the duration of the analysis time window, Equipped with, method.

8. The at least one metric further includes node transition frequency, Calculating the node transition frequency means The number of local maximal transitions between the first node and the second node of the plurality of nodes within the analysis time window, The number of times is divided by the duration of the analysis time window, Equipped with, The method according to claim 7.

9. The at least one metric further includes node transition polarity, Calculating the node transition polarity for one node of multiple nodes is, The first number of times a local maximum occurs when transitioning from one or more of the aforementioned nodes to the aforementioned node, The number of times the local maximum occurs when transitioning from one of the multiple nodes to one or more of the multiple nodes is counted. Taking the difference between the first number of times and the second number of times, Equipped with, The method according to claim 7.

10. The method according to claim 9, wherein the node among the plurality of nodes has inward polarity when the first number of times is greater than the second number of times, and the node among the plurality of nodes has outward polarity when the second number of times is greater than the first number of times.

11. The aforementioned method, By comparing the values ​​of the neurological activity data with a threshold, Based on comparing, discarding, or ignoring one or more of the values ​​of the neurological activity data, By calculating at least one local maxima from the remaining values ​​of the aforementioned neurological activity data, Preprocessing the aforementioned neurological activity data. The method according to claim 7, further comprising:

12. The method according to claim 7, further comprising providing a graphic representation of the at least one metric on a display.

13. The method according to claim 12, wherein the graphic representation of the at least one metric comprises a node of three-dimensional (3D) source localization that is overlaid on the image of the brain.

14. The method according to claim 13, wherein the size of the node is proportional to the value of the node visit frequency for the node.

15. The method according to claim 13, wherein the shading, color, or combination thereof of the node indicates a value for the node visit frequency.

16. The method according to claim 13, wherein the graphic representation further comprises a second node of the 3D source localization and an arrow between the node and the second node, the direction of the arrow indicating a local maximum transition between the node and the second node.

17. The method according to claim 16, wherein the shading, color, or combination thereof of the arrows indicates the frequency of the transitions over an analysis time window.

18. The method according to claim 7, further comprising providing a report on a display, the report comprising a plurality of contents comprising one or more graphic representations of the at least one metric, and further comprising a timeline of the neurological activity data, one or more statistics, one or more additional parameters, or a combination thereof.

Images