Systems and methods for seizure detection based on changes in electroencephalogram (EEG) non-linearities

1 . A seizure detection system comprising one or more circuits, the one or more circuits configured to:

receive an electroencephalogram (EEG) signal generated based on electrical brain activity of a patient;

determine a plurality of different types of metrics based on the EEG signal, the plurality of different types of metrics indicating non-linear features of the EEG signal, the plurality of different types of metrics including eigenvalues;

generate a user interface, the user interface comprising:

a real-time trend of the EEG signal; and

a real-time trend of the eigenvalues;

cause a user interface device to display the user interface;

determine an occurrence of a pattern of a trajectory of a metric indicating a non-linear feature of the non-linear features;

determine that a probability value of the pattern of the trajectory occurring is less than a probability level;

detect a candidate seizure responsive to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level; and

update an appearance of the metric in the user interface in response to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level.

2 . The seizure detection system of claim 1 , wherein the seizure detection system is a cloud-based system, wherein the one or more circuits are configured to receive the EEG signal from a local EEG acquisition system via a network and provide result data to the local EEG acquisition system via the network.

3 . The seizure detection system of claim 1 , wherein the seizure detection system is a local system, wherein the local system is:

integrated with a local EEG system; or

connected locally to an EEG acquisition system.

4 . The seizure detection system of claim 1 , wherein the plurality of different types of metrics comprise at least one of dimensionality, synchrony, Lyapunov exponents, entropy, global non-linearity, distance differences between recurrence trajectories, or self-similarity.

5 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to:

perform a preliminary analysis with the eigenvalues, wherein the preliminary analysis indicates whether the EEG signal is insignificant.

6 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to:

determine probabilities of occurrences of patterns of trajectories of each of the plurality of different types of metrics at a plurality of points in time;

determine whether the probabilities of the patterns of the occurrences of the trajectories of each of the plurality of different types of metrics meet a predefined level of statistical significance based on the probabilities; and

map metrics that meet the predefined level of statistical significance of the plurality of different types of metrics to a category, wherein the category is a seizure category.

7 . The seizure detection system of claim 1 , wherein each value of a plurality of values of the metric of the non-linear feature decreases with respect to a previous value of the plurality of values;

wherein the one or more circuits are configured to determine the probability value of the pattern of the trajectory occurring based on a number of the plurality of values each decreasing with respect to the previous value of the plurality of values.

8 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to:

receive user input via the user interface device; and

set the probability level to a value selected by a user via the user input.

9 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to:

retrieve a default value from a memory device of the seizure detection system; and

set the probability level to the default value retrieved from the memory device.

10 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to update the appearance of the real-time trend of the eigenvalues in the user interface by:

highlighting a section of the real-time trend of the eigenvalues in the user interface.

11 . The seizure detection system of claim 1 , wherein the one or more circuits are configured to update the appearance of the real-time trend of the eigenvalues in the user interface by:

changing a color of at least a portion of the real-time trend of the eigenvalues in the user interface.

12 . A method of seizure detection, comprising:

receiving, by a processing circuit, an electroencephalogram (EEG) signal generated based on electrical brain activity of a patient;

determining, by the processing circuit, a plurality of different types of metrics based on the EEG signal, the plurality of different types of metrics indicating non-linear features of the EEG signal, the plurality of different types of metrics including eigenvalues;

generating, by the processing circuit, a user interface, the user interface comprising:

a real-time trend of the EEG signal; and

a real-time trend of the eigenvalues;

causing, by the processing circuit, a user interface device to display the user interface;

determining, by the processing circuit, an occurrence of a pattern of a trajectory of a metric of a non-linear feature of the non-linear features;

determining, by the processing circuit, that a probability value of the pattern of the trajectory occurring is less than a probability level;

detecting, by the processing circuit, a candidate seizure responsive to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level; and

updating, by the processing circuit, an appearance of the metric in the user interface in response to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level.

13 . The method of claim 12 , wherein each value of a plurality of values of the metric of the non-linear feature decreases with respect to a previous value of the plurality of values;

wherein the method further comprises determining, by the processing circuit, the probability value of the pattern of the trajectory occurring based on a number of the plurality of values each decreasing with respect to the previous value of the plurality of values.

14 . The method of claim 12 , further comprising:

receiving, by the processing circuit, user input via the user interface device; and

setting, by the processing circuit, the probability level to a value selected by a user via the user input.

15 . The method of claim 12 , further comprising:

retrieving, by the processing circuit, a default value from a memory device of a seizure detection system; and

setting, by the processing circuit, the probability level to the default value retrieved from the memory device.

16 . The method of claim 12 , further comprising updating, by the processing circuit, the appearance of the real-time trend of the eigenvalues in the user interface by:

highlighting a section of the real-time trend of the eigenvalues in the user interface.

17 . The method of claim 12 , further comprising updating, by the processing circuit, the appearance of the real-time trend of the eigenvalues in the user interface by:

changing a color of at least a portion of the real-time trend of the eigenvalues in the user interface.

18 . One or more non-transitory storage media storing instructions thereon that, when executed by one or more processors, cause the one or more processors to perform operations, comprising:

receiving an electroencephalogram (EEG) signal generated based on electrical brain activity of a patient;

determining a plurality of different types of metrics based on the EEG signal, the plurality of different types of metrics indicating non-linear features of the EEG signal, the plurality of different types of metrics including eigenvalues;

generating a user interface, the user interface comprising:

a real-time trend of the EEG signal; and

a real-time trend of the eigenvalues;

causing a user interface device to display the user interface;

determining an occurrence of a pattern of a trajectory of a metric of a non-linear feature of the non-linear features;

determining that a probability value of the pattern of the trajectory occurring is less than a probability level;

detecting a candidate seizure responsive to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level; and

updating an appearance of the metric in the user interface in response to a determination that the probability value of the pattern of the trajectory occurring is less than the probability level.

19 . The one or more non-transitory storage media of claim 18 , wherein the instructions cause the one or more processors to perform operations, comprising:

monitoring a moving window of the eigenvalues of the EEG signal; and

determining that the eigenvalues are decreasing.

20 . The one or more non-transitory storage media of claim 18 , wherein the instructions cause the one or more processors to perform operations, comprising:

determining a dimensionality of the EEG signal by performing a phase space analysis by increasing a value of the dimensionality until a number of false neighbors reaches zero, wherein a starting value of the dimensionality is based on an age of the patient.