Identifying risk levels for seizure activity based on sleep state

Abstract

A computer-implemented method may include receiving data indicative of brain wave activity over a particular time period. The method may also include determining at least one metric associated with at least one sleep state of the subject based on the data. In addition, the method may include determining a risk level associated with seizure activity of the subject based on the at least one metric. The method may also include generating an output indicative of the risk level.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 484,570, filed on February 13, 2023, entitled “Identifying Risk Levels of Epileptic Seizure Activity Based on Sleep State,” the entire contents of which are hereby incorporated herein by reference. Technical Field

[0003] The present disclosure relates generally to analyzing physiological data and, more particularly, although not necessarily exclusively, to identifying risk levels of seizure activity based on sleep state. Background Art

[0004] An electroencephalogram (EEG) is a tool used to measure electrical activity generated by the brain. Brain activity is collected by electrodes placed on the subject's scalp. Traditional monitoring and diagnostic equipment consists of several electrodes attached to the subject, which tap brain signals and transmit them via cables to an amplifier unit. The resulting EEG signals can be used to diagnose and monitor various conditions affecting the brain, such as epilepsy.

[0005] Epilepsy is a brain disease that causes epileptic seizures. There are many types of epilepsy, which can be diagnosed based on the type of epileptic seizure suffered by the subject. Examples of the type of epilepsy can include focal epilepsy, generalized epilepsy and unknown epilepsy. The subject can experience more than one type of epileptic seizure, and epileptic seizures can vary in appearance by, for example, occurring with different intensities, durations, symptoms, etc. The treatment of epilepsy can include drug therapy, but the subject can usually develop resistance to drug therapy. In addition, drugs may increase the possibility of sleep deprivation or substance abuse, which may increase the possibility of epileptic seizures. Therefore, epilepsy is a complex disease that is difficult to diagnose and treat.

[0006] It is further difficult to predict when a person with epileptic seizures will experience a seizure. For some subjects, subtle changes in heart rate, skin conductance, or breathing patterns can be indicators for predicting seizure activity when a seizure is about to begin. In addition, conventional systems have applied machine learning techniques to EEG data to extract features that can be used for seizure prediction. However, the complexity of seizure activity can make it difficult to extract features that can be used to accurately predict seizures for a population of subjects (e.g., particularly for subjects outside the population used in training the machine learning technique). Therefore, there is a need for a comprehensive and accurate method for predicting seizures. In addition, there is a further need to predict that a seizure will occur earlier than when it is about to begin so that preparation, intervention, or treatment can be performed. Summary of the Invention

[0007] Aspects of the present disclosure relate to identifying a risk level for seizure activity based on sleep states. One aspect relates to a computer-implemented method. The method includes receiving data indicating brainwave activity over a specific time period; determining, based on the data, at least one metric associated with at least one sleep state of a subject; determining, based on the at least one metric, a risk level associated with seizure activity in the subject; and generating an output indicating the risk level.

[0008] In some embodiments, the at least one metric is an amount of time in a particular sleep state. In some embodiments, the particular sleep state is a rapid eye movement (REM) state of sleep.

[0009] In some embodiments, the data is received from an RF transmitter-receiver associated with a multi-electrode device. In some embodiments, the specific time period is a first time period. In some embodiments, the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period. In some embodiments, the second time period occurs after the first time period.

[0010] In some embodiments, the output is a first output. In some embodiments, the method further comprises identifying a treatment recommendation based on the risk level, and generating a second output indicative of the treatment recommendation, the treatment recommendation being useful in reducing the risk level.

[0011] In some embodiments, the at least one metric is a first metric. In some embodiments, determining the risk level based on the at least one metric further comprises determining a second metric for healthy people based on historical data, identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric, and determining the risk level based on the statistically significant difference.

[0012] In some embodiments, generating an output indicative of a risk level further comprises providing an output in a first color indicating a high risk level, providing an output in a second color indicating a medium risk level, and providing an output in a third color indicating a low risk level.

[0013] One aspect relates to a system. The system includes one or more data processors and a non-transitory computer-readable storage medium. The computer-readable storage medium contains instructions that, when executed on the one or more data processors, cause the one or more data processors to receive data indicating brainwave activity over a specific time period, determine at least one metric associated with at least one sleep state of a subject based on the data, determine a risk level associated with seizure activity in the subject based on the at least one metric, and generate an output indicative of the risk level.

[0014] In some embodiments, the at least one metric is an amount of time in a particular sleep state. In some embodiments, the particular sleep state is a rapid eye movement (REM) state of sleep.

[0015] In some embodiments, the data is received from an RF transmitter-receiver associated with a multi-electrode device. In some embodiments, the specific time period is a first time period. In some embodiments, the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period. In some embodiments, the second time period occurs after the first time period.

[0016] In some embodiments, the output is a first output. In some embodiments, the instructions, when executed on the one or more data processors, cause the one or more data processors to identify a treatment recommendation based on the risk level and generate a second output indicative of the treatment recommendation, the treatment recommendation being usable to reduce the risk level.

[0017] In some embodiments, the at least one metric is a first metric. In some embodiments, determining the risk level based on the at least one metric further comprises determining a second metric for healthy people based on historical data, identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric, and determining the risk level based on the statistically significant difference.

[0018] In some embodiments, generating an output indicative of a risk level further comprises providing an output in a first color indicating a high risk level, providing an output in a second color indicating a medium risk level, and providing an output in a third color indicating a low risk level.

[0019] One aspect relates to a computer program product tangibly embodied in a non-transitory machine-readable storage medium. The computer program product includes instructions that cause one or more data processors to receive data indicating brainwave activity over a specific time period, determine at least one metric associated with at least one sleep state of a subject based on the data, determine a risk level associated with seizure activity in the subject based on the at least one metric, and generate an output indicative of the risk level.

[0020] In some embodiments, the at least one metric is an amount of time in a particular sleep state. In some embodiments, the particular sleep state is a rapid eye movement (REM) state of sleep.

[0021] In some embodiments, the data is received from an RF transmitter-receiver associated with a multi-electrode device. In some embodiments, the specific time period is a first time period. In some embodiments, the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period. In some embodiments, the second time period occurs after the first time period.

[0022] In some embodiments, the output is a first output. In some embodiments, the instructions cause the instructions to further cause one or more data processors to identify a treatment recommendation based on the risk level and generate a second output indicative of the treatment recommendation, the treatment recommendation being usable to reduce the risk level.

[0023] In some embodiments, the at least one metric is a first metric. In some embodiments, determining the risk level based on the at least one metric further comprises determining a second metric for healthy people based on historical data, identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric, and determining the risk level based on the statistically significant difference.

[0024] In some embodiments, generating an output indicative of a risk level further comprises providing an output in a first color indicating a high risk level, providing an output in a second color indicating a medium risk level, and providing an output in a third color indicating a low risk level. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a block diagram of an example of a system for acquiring physiological data according to one example of the present disclosure.

[0026] Figure 2 is a block diagram of an example of a system for identifying risk levels of seizure activity based on sleep state according to one example of the present disclosure.

[0027] Figure 3 is a block diagram of an example of a computing system for identifying risk levels of seizure activity based on sleep state according to one example of the present disclosure.

[0028] Figure 4 is a flow chart of a process for identifying risk levels of seizure activity based on sleep state according to one example of the present disclosure. DETAILED DESCRIPTION

[0029] Certain aspects and examples of the present disclosure relate to a system and method for predicting the risk level of epileptic seizure activity within a specific time frame based on analysis of physiological data. The risk level can be the likelihood that a subject will exhibit epileptic seizure activity (e.g., within a predefined time period). Epileptic seizure activity can include neural activity consistent with an epileptic seizure (e.g., as collected via one or more electroencephalogram (EEG) electrodes) and / or clinical seizures consistent with an epileptic seizure. The risk level can be determined using various environmental, physiological, or other suitable factors. The physiological data used to predict the risk level can include electroencephalogram (EEG) data, electrocardiogram (ECG) data, electromyogram (EMG) data, electrooculogram (EOG) data, or other suitable physiological data. The physiological data can be obtained via a physiological data acquisition component that includes at least a single physiological data channel having at least one reference electrode and at least one active electrode in close proximity. The component can be worn by the subject. For example, the component can include a patch configured to be positioned on (e.g., adhered to) the user's forehead. In addition, the patch can have an adhesive film to which the electrodes can be attached to collect the physiological data.

[0030] In one example of the present disclosure, the risk level of epileptic seizure activity can be identified based on EEG data indicative of sleep state. A sleep state can be any distinguishable state of sleep or wakefulness that represents behavioral, physical, or signal characteristics. In some cases, the EEG data is processed to infer a category indicating a prediction as to whether the subject is awake or asleep for each of a plurality of time intervals, and potentially, if the subject is estimated to be asleep, a specific type or stage of sleep. The inference can be made based on converting the time domain electrical signal into a frequency domain intensity or power value for each of the plurality of time intervals. Features can be defined as cumulative or maximum intensity or power values ​​within various frequency bands. The sleep state can then be inferred based on the absolute or relative values ​​of one or more features.

[0031] For example, a wakeful sleep state can be detected or defined by processing EEG data to detect signals within one or more specific frequency bands (e.g., a frequency band extending between about 13 Hz and about 60 Hz and an amplitude of at least about 30 microvolts (μV) (i.e., beta waves)). The frequency and amplitude can be determined by transforming the time-domain electrical signal into the frequency domain via a mathematical transformation (e.g., a Fourier transform) or other suitable technique. In some examples, additional sleep states can be characterized by stage one, stage two, stage three, and rapid eye movement (REM). The frequency band used to detect stage one sleep from EEG data can be defined as corresponding to a specific type of wave and / or sleep stage. For example, the frequency band can be defined as extending between 3 and 8 Hz. The detection algorithm can be configured in the time domain or frequency domain to detect features that support a prediction of whether the subject is in a given sleep stage (e.g., stage one sleep). For illustration, if the amplitude in the 3 to 8 Hz frequency band is between 50 and 100 μV (i.e., theta waves), it can be inferred that the subject is in stage one sleep. Other features of the sleep state, such as sleep spindles and K complexes, can be distinguished by a detection algorithm to predict the sleep state. For example, a high-frequency band (e.g., a band of approximately 15 Hz) that lasts less than 2 seconds in the time domain can be detected as a sleep spindle. Similarly, a low-frequency band (e.g., a band extending between 1 and 4 Hz and having an amplitude between 100 and 200 μV) (i.e., a delta wave) that lasts about 1 second in the time domain can be detected as a K complex. Therefore, if one or more portions of the EEG data are detected as sleep spindles, and after or in the vicinity of one or more portions of the EEG data detected as K complexes, it can be inferred that the subject is in stage two sleep. In another example, a frequency band extending between 1 and 4 Hz can be detected that lasts significantly longer than 2 seconds (e.g., for 20 minutes), and from this, it can be predicted that the subject is in stage three sleep. Stage three sleep can also be referred to as slow waves or delta sleep. In addition, for a frequency band extending between about 13 and about 60 Hertz (Hz) and for an amplitude of at least about 30 μV (i.e., beta waves), it can be predicted that the subject is in REM sleep. However, beta waves can also be detected during a wakeful sleep state. Therefore, additional physiological data, physical or biological indicators or other suitable data can be obtained and identified within the detection algorithm to distinguish between REM sleep and a wakeful sleep state. For example, EMG data can be obtained, and the detection algorithm can detect phase events (e.g., rapid eye movements and limb twitching) or tonic phenomena (e.g., loss of tension in anti-gravity muscles), both of which can indicate REM sleep. The detection of phase events or tonic phenomena can be compared or combined with EEG data to distinguish a REM sleep state from a wakeful sleep state or another sleep state.

[0032] In some examples, sleep states can be characterized by REM sleep and non-REM sleep. For example, stage 1 sleep, stage 2 sleep, and stage 3 sleep can be combined into non-REM sleep. Thus, the detection algorithm can detect non-REM sleep by inferring EEG data outside the 13 to 60 Hz range and with an amplitude above or below approximately 30 μV as non-REM sleep, and detect REM sleep by inferring data within the 13 to 60 Hz range and with an amplitude of approximately 30 μV as REM sleep.

[0033] EEG signals are typically examined in time in sequential increments called epochs. For example, when EEG signals are used to analyze sleep, sleep can be segmented into one or more segments for analysis. The segments can be divided into different sections using a scanning window, where the scanning window defines different sections of the time series increment. The code can move the scanning window (incrementally or via shifting) via a sliding or shifting window, where the segments of the sliding window have overlapping or non-overlapping time series. For example, the segments can optionally span the entire time series. In some examples, each segment can be classified as corresponding to a predicted sleep state represented. In some cases, before classification, the segments are normalized or doubly normalized based on (for example) frequency information, amplitude information, power, intensity, or other suitable features of the EEG data that can be associated with the sleep state. U.S. patent application Ser. No. 11 / 431,425, filed May 9, 2006, discloses exemplary techniques for normalizing biodata, which is incorporated herein by reference for all purposes.

[0034] In some examples, a given segment can be classified as REM (or alternatively, non-REM) sleep based on whether the power (or normalized or doubly normalized power) in a given frequency band (e.g., in an absolute sense or relative to the power in one or more other frequency bands) exceeds a threshold. The threshold can be an absolute threshold, a threshold defined based on data from a population of subjects diagnosed with a given condition, a threshold defined based on data from a population of subjects experiencing a given condition, or a threshold defined based on empirical data associated with the subject.

[0035] Any of the same or similar techniques can be used to predict different types of sleep stages. For example, the intensity within one or more bands of a normalized or doubly normalized segment can be used to predict whether a subject is in stage 1, 2, 3, or REM sleep, whether spindles are present in the sleep data, whether K complexes are present in the sleep data, etc.

[0036] Any set of segments initially classified as one sleep state can be further subdivided into multiple sub-classified sleep states according to increasing levels of classification detail. For example, a set of segments classified as non-REM can be further divided into stage one, stage two, stage three, or a combination thereof.

[0037] In some embodiments, artificial intelligence techniques can be used to predict that a subject has a given sleep disorder, predict the severity of a sleep disorder, or predict the efficacy of a treatment for a given sleep disorder. Artificial intelligence techniques can include embedded signal processing (e.g., which can include applying one or more signal transformations) and using one or more models or rules to generate segment-specific, night-specific, or subject-specific predictions. For example, EEG signals can be collected across a sleep period (e.g., a night). The EEG signal can be divided into segments corresponding to absolute or relative time increments of the entire time period (e.g., 1 minute, 5 minute, or 10 minute time intervals), and a spectrum can be generated for each segment so that the power or intensity of each of the frequency bands can be identified for each time increment. Alternatively, a spectrogram can be generated for the time period, wherein the spectrogram identifies the power or intensity values ​​of each frequency band for each of the multiple time increments in the time period (e.g., 1 minute, 5 minute, or 10 minute time increments). A set of features can be defined such that the feature indicates or corresponds to: (1) the power, intensity, or other suitable property of a frequency band in a spectrogram or spectrum; and / or (2) the intensity of a spectrum or spectrogram, a derivative of a spectrum (e.g., across time), or a bi-derivative of a spectrum or spectrogram (e.g., across time, across frequency, or across both time and frequency). Features can, for example, predict the likelihood of a particular sleep stage or state (e.g., REM or non-REM). In some instances, features are defined based on other segment-specific features. For example, subject-specific and / or day- or night-specific features can indicate the percentage of time or sleep that is predicted to be in a REM state.

[0038] Artificial intelligence rules can be defined to predict REM deprivation and / or seizure propensity based on the features. For example, clustering techniques, support vector machine (SVM) techniques, principal component techniques, independent component techniques, logistic regression techniques, etc. can be used to predict whether the subject is in REM sleep (relative to non-REM sleep or wakefulness) for each time segment. In some cases, for each segment, a probability (likelihood) that the subject is in REM sleep is generated, which can then be compared with a predefined or learned threshold to predict whether the subject is or has been in REM sleep.

[0039] Rules can be defined to predict whether a subject is (or was) sleep deprived based on REM sleep predictions. For example, if less than a threshold percentage (e.g., 10%, 15%, 20%, 25%, 30%, or 35%) of segments are predicted to be REM sleep, the rule can indicate that the subject is (or was) sleep deprived. In another example, a rule can indicate that a subject is or was sleep deprived by identifying that the length of REM sleep predicted for one or more segments by the detection algorithm is less than a predetermined threshold for healthy or normal sleep or a learned threshold for the subject. An additional rule can indicate that for subsequent sleep cycles (i.e., a series of sleep cycles over a night's sleep), if the length or percentage of REM sleep does not increase by at least a threshold amount (e.g., 1 minute, 5 minutes, etc. or 5%, 10%, etc.), it can be inferred that the subject is or was sleep deprived.

[0040] To identify risk levels, metrics can be extracted from EEG data of sleep states, and the metrics can be associated with the impact on seizure activity. For example, the metrics can be the mean, variance, skewness, etc. of the frequency band, amplitude, or other suitable features of the EEG data. In other examples, the metrics can be statistics derived from a segment or a group of segments, such as the percentage of sleep predicted to be REM sleep. Metrics can also be extracted from EMG data, EOG data, ECG data, or other suitable physiological data obtained during sleep. For example, metrics can include heart rate, oxygen levels in the blood, eye or leg movements, etc.

[0041] Additionally, in some examples, characteristics of the subject can be further used to correlate the metric with the effect on seizure activity. For example, the metric can be the ratio of REM to non-REM sleep, and the metric can be correlated with the effect on seizure activity based on the subject's age. Additional characteristics of the subject can include medications or medication dosages prescribed to the subject, sleep disorders previously diagnosed for the subject, the subject's gender, etc.

[0042] In a particular example, sleep deprivation can be detected based on predefined rules, and the metric extracted from the EEG data can be associated with the rule. For example, the predefined rule can include detecting sleep deprivation if the percentage of REM sleep associated with a group of segments is less than a threshold percentage. Thus, the metric can be the percentage of REM sleep. A classification algorithm (e.g., K-nearest neighbor, decision tree, support vector machine, etc.) or another suitable algorithm can then be implemented to predict the risk level based on the percentage of REM sleep. Additional metrics can be extracted from the group of segments and provided to the classification algorithm. Additional metrics can be other suitable indicators of sleep deprivation (e.g., predicting the length of time the subject is in a sleep state). The classification algorithm can also receive characteristics of the subject (e.g., age). The classification algorithm can then indicate the risk level of epileptic seizure activity for the subject based on the percentage of REM sleep, the additional metrics, and / or the characteristics of the subject. For example, the classification algorithm can indicate the risk level by classifying the risk level as high, medium, or low.

[0043] Moreover, a typical sleep cycle for a healthy subject detected and analyzed via EEG data can be determined to last approximately ninety to one hundred and ten minutes. For a typical sleep cycle, the segments of each sleep state in chronological order can indicate that a typical sleep cycle occurs as stage one, stage two, stage three, stage two, and then REM. In addition, for EEG data collected during a typical sleep night for a healthy subject, the EEG data can include four to five sleep cycles, wherein the time period associated with REM sleep can increase for each subsequent sleep cycle. Furthermore, in a typical sleep cycle, approximately five percent of the EEG data obtained can be associated with stage one, approximately forty-five percent can be associated with stage two, approximately twenty-five percent can be associated with stage three, and approximately twenty-five percent can be associated with REM. Therefore, approximately seventy-five percent of the EEG data for a typical sleep cycle can be associated with non-REM sleep, and approximately twenty-five percent can be associated with REM sleep.

[0044] Therefore, because abnormal sleep activity (e.g., sleep deprivation) is a common trigger for epileptic seizure activity, and typical sleep activity can be well defined, EEG data associated with sleep states can provide a comprehensive mechanism for predicting the risk level of epileptic seizure activity. Sleep states can be predicted based on frequency bands, amplitudes, or other suitable features of the EEG data. Artificial intelligence techniques or rules can then be implemented to predict sleep deprivation for other suitable abnormal sleep activities. In addition, metrics can be derived from data associated with predictions of sleep states and / or sleep deprivation and then classified (i.e., by a classification algorithm) to predict the risk level of epileptic seizure activity for the subject. In addition, abnormal sleep activity may not immediately trigger epileptic seizure activity, so the risk level of epileptic seizure activity for a certain time frame after the abnormal sleep activity can be predicted based on the EEG data. Therefore, the examples of the present disclosure can provide an accurate and comprehensive method for predicting the risk level of epileptic seizure activity, and can further predict the risk level of a specific time frame to enable preparation or prevention of epileptic seizure activity.

[0045] Illustrative examples are provided to introduce the reader to the general subject matter discussed herein and are not intended to limit the scope of the disclosed concepts. The following section describes various additional features and examples with reference to the accompanying drawings, in which like numerals indicate like elements and directional descriptions are used to describe illustrative aspects, but, like the illustrative aspects, should not be used to limit the present disclosure.

[0046] Figure 1 is a block diagram of an example of a system for acquiring physiological data according to an example of the present disclosure. The system 100 may include a multi-electrode device 104, which may have one or more active electrodes 106a for collecting active signals and one or more reference electrodes 106b that may collect corresponding reference signals. In addition, the multi-electrode device 104 may include a ground electrode 106c. The electrodes 106a-c may be fixed in position within the device (e.g., patch 102) or removable (e.g., strapped to the device). The system 100 may also include a processing subsystem 116, a storage subsystem 118, a (radio frequency) RF transmitter-receiver 114, a connector interface 112, a power subsystem 108, and an environmental sensor 120, each of which may be communicatively coupled to the multi-electrode device 104 or be part of the multi-electrode device 104.

[0047] The processing subsystem 116 can be implemented as one or more integrated circuits, for example, one or more single-core or multi-core microprocessors or microcontrollers, examples of which are known in the art. The processing subsystem 116 can control the operation of the multi-electrode device 104 by executing various programs in response to program code, and can maintain multiple concurrently executing programs or processes. For example, the processing subsystem 116 can execute code that can control the collection, analysis, application, and / or transmission of physiological data (e.g., electroencephalogram (EEG) data, electromyogram (EMG) data, etc.). Some or all of the program code can be stored in the processing subsystem 116, or the program code can be stored in a storage medium such as the storage subsystem 118. In addition, the processing subsystem 116 can amplify, filter, or a combination thereof the signals detected by the electrodes 106a-c of the multi-electrode device 104, and can further store the signals and recording details (e.g., recording time or user identifier). In some examples, the processing subsystem 116 can analyze physiological data or signals to detect physiological correspondences. For example, the recorded signals can reveal frequency properties corresponding to sleep stages.

[0048] In addition, the storage subsystem 118 can be implemented using, for example, magnetic storage media, flash memory, other semiconductor memories (e.g., DRAM, SRAM), or any other non-transitory storage media, or a combination of media, and can include volatile and / or non-volatile media. In some examples, the storage subsystem 118 can store physiological data, information about the subject (e.g., identification information or medical history information), or analysis variables (e.g., frequency, amplitude, etc.) obtained from the physiological data. The storage subsystem 118 can also store one or more programs that can be executed by the processing subsystem 116. The one or more programs can initiate or otherwise control the collection, analysis, or transmission of physiological data.

[0049] The RF transmitter-receiver 114 may enable the multi-electrode device 104 to communicate wirelessly with various interface devices, such as phones, tablets, laptops, etc. The RF transmitter-receiver 114 may include a combination of hardware components, including, for example, driver circuits, antennas, modulators, demodulators, encoders, decoders, other suitable analog and / or digital signal processing circuits, and may also include software components. Various wireless communication protocols may be implemented via the RF transmitter-receiver 114 using the software components and associated hardware. The RF transceiver component of the RF transmitter-receiver 114 may include an antenna and supporting circuitry to enable data communication over a wireless medium, such as Wi-Fi, or other suitable medium for wireless data communication.

[0050] The connector interface 112 can enable the multi-electrode device 104 to communicate with various interface devices via a wired communication path, for example, using a universal serial bus (USB), a universal asynchronous receiver / transmitter (UART), or other protocols for wired data communication. In some examples, the connector interface 112 can provide a power port for allowing the multi-electrode device 104 to receive power. The connector interface 112 can also provide a connection to send or receive physiological data. For example, physiological data can be transmitted to or from another device, such as another multi-electrode device, in an analog or digital format.

[0051] Environmental sensors 120 may include various electronic, mechanical, electromechanical, optical, or other devices that provide information related to external conditions surrounding multi-electrode device 104 or information about the subject. Any type and combination of environmental sensors 120 may be used. For example, an accelerometer may be used to estimate whether the user is sleeping or attempting to sleep or otherwise estimate activity status. In another example, an electro-oculogram sensor may be used to detect eye movement to help identify rapid eye movement (REM) sleep stages.

[0052] Additionally, power subsystem 108 may provide power and power management capabilities for multi-electrode device 104. For example, power subsystem 108 may include battery 110 and associated circuitry to distribute power from battery 440 to other components of system 100 that may require power.

[0053] It will be appreciated that the system 100 is illustrative and that variations and modifications are possible. In an example, the processing subsystem 116 may execute code from the storage subsystem 118 for analyzing sleep states based on EEG data and predicting a risk level for seizure activity based on the analysis. Accordingly, the system 100 may also include a user interface to enable a user to interact directly with the device, for example, to receive a risk level. The risk level may be the likelihood that a subject will experience seizure activity within a particular time frame. The risk level may be displayed at the user interface in a color indicating whether the risk level is, for example, high, medium, or low, or the risk level may be displayed as a percentage or in another suitable format. Further, while the system 100 is described with reference to particular blocks, it will be understood that these blocks are defined for ease of description and are not intended to imply a particular physical arrangement of component parts.

[0054] Figure 2is a block diagram of an example of a system 200 for identifying a risk level 202 of epileptic seizure activity based on a sleep state 216 according to an example of the present disclosure. The risk level 202 can be the likelihood that a subject will experience epileptic seizure activity (e.g., within a predetermined time period). The system 200 can include a computing device 201, which can be communicatively coupled with a display device 220 and a multi-electrode device 204 to identify the risk level 202 based on physiological data indicative of the sleep state 216 and provide the risk level 202 to the subject, a physician, or another suitable user. The physiological data can include electroencephalogram (EEG) data 232, electromyogram (EMG) data, electrocardiogram (ECG) data, electrooculogram (EOG) data, or other suitable physiological data. The computing device 201 can communicate with the display device 220 and the multi-electrode device 204 via a network 230, such as a local area network (LAN) or the Internet.

[0055] In some examples, computing device 201 may receive physiological data from multi-electrode device 204 or other suitable devices or sensors. In one example, multi-electrode device 204 may correspond to Figure 1 The computing device 201 may also include a multi-electrode device 104. Thus, the computing device 201 may receive physiological data, such as EEG data 232, from the RF transmitter-receiver 114 associated with the multi-electrode device 204. The EEG data 232 may indicate brainwave activity 238 of the subject. The multi-electrode device 204 may also be associated with environmental sensors such as accelerometers, ECG, EOG, EMG, etc., for collecting additional physiological data. In addition, the computing device 201 may receive the EEG data 232 within a first time period 222a, such as within a sleep cycle, within a series of sleep cycles (e.g., four consecutive sleep cycles), etc. In addition to recording the EEG data 232, the multi-electrode device 204 may also filter, amplify, transmit, or otherwise perform operations on the EEG data 232 or additional physiological data to improve the acquisition of the EEG data 232 or additional physiological data at the computing device 201, thereby improving the efficiency and accuracy of the identification of the risk level 202.

[0056] In addition, the computing device 201 can analyze the EEG data 232 or additional physiological data to detect or otherwise distinguish between sleep states 216. For example, the computing device 201 can collect EEG data 232 for a first time period 222a (i.e., an 8-hour sleep period) via the multi-electrode device 204 and segment the EEG data 232 for the first time period 222a into segments, for example, using a scanning window. The computing device 201 can analyze the segments to predict a sleep state associated with each of the segments. These segments can be initially segmented into a number of shorter time segments (e.g., 1-minute, 3-minute, or 5-minute time segments) and then reorganized based on the predicted sleep state.

[0057] To analyze the segments, the computing device 201 may further perform a fast Fourier transform (FFT) or other suitable mathematical transform, algorithm, or technique to transform the EEG data 232 of each segment into the frequency domain. Once the EEG data 232 is transformed into the frequency domain, additional features such as peak amplitude, peak frequency, median frequency, etc. may be extracted. In some cases, the segments may be further normalized or doubly normalized based on, for example, frequency information, amplitude information, power, intensity, or other suitable features of the EEG data. In addition, the computing device 201 may perform a short-time Fourier transform (STFT) or other suitable technique to identify time-frequency domain features from the EEG data 232, which may depend on both time-domain and frequency-domain characteristics. For example, the computing device 201 may perform an STFT to extract time-frequency domain features such as average power, entropy, fractional energy, etc.

[0058] After analysis, frequency information, amplitude information, time-frequency domain features, etc. associated with each of the segments can be associated with a sleep state to predict the sleep state of each of the segments. For example, the sleep state 216 can be characterized or defined as a wakeful sleep state, stage one, stage two, stage three, and rapid eye movement (REM), or the sleep state 216 can be characterized or defined as REM and non-REM. The computing device 201 can detect the sleep state 216 based on time domain features, frequency domain features, time-frequency domain features, other suitable features, or a combination thereof identified by the computing device 201 based on the EEG data 232. Features of the EEG data 232 can indicate characteristics or changes in brainwave activity 228, which can correspond to the sleep state 216.

[0059] In some examples, the detection algorithm can be configured by the computing device 201 in the time domain or the frequency domain to detect characteristic signatures that support sleep state prediction. For example, brainwave activity 228 for a stage three sleep state may include delta waves, which may exhibit low frequencies (i.e., a frequency band extending between 1 Hz and 4 Hz) and high amplitudes (i.e., an amplitude extending between 100 μV and 200 μV) in the frequency domain of the EEG data 232. Therefore, if the first segment includes a frequency band of 1 to 4 Hz and an amplitude between 100 and 200 μV, the first sleep state of the first segment may be predicted to be stage three.

[0060] In other examples, a classification algorithm (i.e., K-nearest neighbor, SVM, or other suitable classification algorithm) can be configured to predict the sleep state of each of the segments by classifying the segments based on a feature signature that supports sleep state prediction. The classification algorithm can be trained to predict sleep states by inputting segments labeled with sleep states, or the classification algorithm can classify sleep states based on predefined rules. Thus, for example, the classification algorithm can classify a second segment having an amplitude between 50 and 100 μV (i.e., theta waves) in the 3 Hz to 8 Hz frequency band as stage 1 sleep.

[0061] After predicting the sleep state for each of the segments, computing device 201 may further determine metrics 206a-b associated with sleep state 216. In some examples, metrics 206a-b may be determined based on frequency patterns, amplitudes, or other suitable features that can be extracted by computing device 201 from EEG data 232 and correspond to sleep state 216. For example, metrics 206a-b may be the power, intensity, or other suitable properties of a frequency band. Metrics 206a-b may be averaged, summed, or otherwise determined for one or more segments associated with a particular sleep state. Furthermore, in some examples, metrics 206a-b may be statistical values ​​derived based on the predicted sleep state. For example, metrics 206a-b may be the ratio of REM sleep to non-REM sleep, the percentage of time spent in a particular sleep state, and so on. In a specific example, first metric 206a may be the cumulative amount of time over first time period 222a that the subject is predicted to be in a REM sleep state. First metric 206a may be determined by summing the time periods associated with each segment in which the subject is predicted to be in REM sleep. The first metric 206a may be an absolute amount of time (eg, twenty minutes) or a relative amount of time (eg, eight percent across the first time period 222a).

[0062] In another example, the first metric 206a can be an estimated amount of time in REM sleep for a sleep cycle or for a series of sleep cycles. Thus, the segments of the first time period 222a can be further grouped or organized by sleep cycle, and for each sleep cycle, a time period during which the subject is predicted to be in a REM sleep state can be determined. The first metric 206a can further be the mean, median, mode, etc. of REM sleep across the series of sleep cycles.

[0063] Additionally, second metric 206b may be the amount of time spent asleep (e.g., the accumulation of stage 1, stage 2, stage 3, and REM sleep). In some examples, metrics 206a-b may further include the amount of time spent in stage 1, stage 2, stage 3, or a combination thereof. Additionally, metrics 206a-b may include characteristics of the subject's sleep cycle (e.g., the length of the sleep cycle or the percentage of time spent in each sleep state during the sleep cycle) or other suitable metrics that can be calculated, estimated, or otherwise identified based on EEG data 232 or other suitable physiological data.

[0064] Computing device 201 may further determine risk level 202 based on metrics 206a-b. In some examples, computing device 201 may implement artificial intelligence techniques that can be used to predict risk level 202 based on metrics 206a-b. The artificial intelligence techniques may include using one or more models or rules to generate subject-specific predictions based on metrics 206a-b derived from one or more segments of first time period 222a. Rules may be defined to predict REM deprivation and / or seizure propensity based on metrics 206a-b. For example, rules may be defined to predict whether a subject is (or was) sleep deprived. For example, if less than a threshold percentage (e.g., 10%, 15%, 20%, 25%, 30%, or 35%) of the segments are predicted to be REM sleep, the rule may indicate that the subject is (or was) sleep deprived. In another example, the rule may indicate that the subject is or was sleep deprived by identifying that the length of REM sleep predicted by the detection algorithm for one or more segments is less than a predefined threshold for healthy or normal sleep or a threshold learned by the subject. Artificial intelligence techniques can further make subject-specific predictions based on other characteristics of the subject (i.e., age, previous diagnosis of sleep- or epilepsy-related conditions, prescribed medications, etc.).

[0065] In a specific example, the first metric 206a can be a relative amount of time that the subject is predicted to be in REM sleep during the first time period 222a. The threshold amount of time for REM sleep can be predefined based on age groups. The threshold amount of time for REM sleep can be a gradient, such that each of the multiple threshold amounts of REM sleep corresponds to a different risk level. For example, for the age group of thirty to fifty years old, the threshold amounts of REM sleep can be relative amounts of 25 percent, 20 percent, and 15 percent. Thus, to illustrate a risk level prediction, if the subject is thirty-five years old and the predicted relative amount of time that the subject will be in REM sleep during the first time period 222a is fourteen percent, the computing device 201 can predict a high risk level for the subject.

[0066] Additionally, or alternatively, computing device 201 can determine risk level 202 based on historical data 236. For example, historical data 236 can be previous EEG data associated with the subject's sleep (e.g., when the subject was taking different medications) or EEG data collected for a healthy population, a population of subjects with a history of epileptic seizures, a population of subjects diagnosed with epilepsy, or another suitable population for which physiological data can be obtained and for which metrics 208a-b can be determined. Metrics 208a-b can be the same metrics as metrics 206a-b or otherwise reflect the same features of EEG data 232. Metrics 208a-b can be marked with seizure occurrences for a predefined time period or otherwise include an indication of seizure activity. Thus, metrics 208a-b, historical data 236, or a combination thereof can be used to train artificial intelligence technology or to generate predefined rules for artificial intelligence technology.

[0067] In response to determining the risk level 202, computing device 201 can generate a first output 224a indicating the risk level 202. First output 224a can be output for display on display device 220. First output 224a can indicate a high risk level, a medium risk level, or a low risk level. In some examples, a likelihood associated with the risk level can also be displayed. For example, a low risk level can be associated with a likelihood less than 33%, a medium risk level can be associated with a likelihood greater than 33% and less than 66%, and a high risk level can be associated with a likelihood greater than 66%. The likelihood can be the likelihood 210 of seizure activity within a second time period 222b. For example, second time period 222b can be an eight-hour period occurring after first time period 222a. Additionally, in some examples, first output 224a can provide a high risk level in a first color, a medium risk level in a second color, and a low risk level in a third color. In other examples, first output 224a can include a percentage of the likelihood 210 of seizure activity or otherwise indicate the risk level 202 on display device 220.

[0068] Additionally, or alternatively, computing device 201 may generate second output 224b, which may include treatment recommendation 226. Computing device 201 may determine treatment recommendation 226 based on EEG data 232 for the subject, additional physiological data, risk level 202, electronic health records, other suitable data, or a combination thereof. Treatment recommendation 226 may be a medication, a medication dosage, or another suitable treatment recommendation that may reduce risk level 202 for the subject.

[0069] Figure 33 is a block diagram of an example of a computing system 300 for identifying a risk level of seizure activity based on sleep state according to one example of the present disclosure. Computing system 300 includes a processor 303 communicatively coupled to a memory device 305. In some examples, processor 303 and memory device 305 can be part of the same computing device, such as server 310. In other examples, processor 303 and memory device 305 can be separate (e.g., remote) from each other.

[0070] Processor 303 may include one processor or multiple processors. Non-limiting examples of processor 303 include a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or a microprocessor. Processor 303 may execute instructions 307 stored in memory device 305 to perform operations. Instructions 307 may include processor-specific instructions generated by a compiler or interpreter from code written in any suitable computer programming language (such as C, C++, C#, Java, or Python).

[0071] Memory device 305 may include one or more memories. Memory device 305 may be volatile or non-volatile. Non-volatile memory includes any type of memory that retains stored information when power is removed. Examples of memory device 305 include electrically erasable programmable read-only memory (EEPROM) or flash memory. At least some of memory devices 305 may include non-transitory computer-readable media from which processor 303 can read instructions 307. Non-transitory computer-readable media may include electronic, optical, magnetic, or other storage devices capable of providing computer-readable instructions or other program code to processor 303. Examples of non-transitory computer-readable media may include magnetic disks, memory chips, ROM, random access memory (RAM), ASICs, configured processors, and optical storage.

[0072] Processor 303 may execute instructions 307 to perform operations. For example, processor 303 may receive data 308 indicating brainwave activity within a time period 312. Processor 303 may also determine a metric 314 associated with a sleep state 316 of the subject based on data 308. Additionally, processor 303 may determine a risk level 318 associated with a seizure activity 320 of the subject based on metric 314. Processor 303 may further generate an output 322 for display at display device 304. Output 322 may indicate risk level 318.

[0073] Figure 4 is a flow chart of a process for identifying a risk level 202 of seizure activity based on a sleep state 216 according to one example of the present disclosure. In some examples, the processor 303 may implement Figure 4Some or all of the steps shown. Other examples may include Figure 4 More steps, fewer steps, different steps, or a different order of steps may be shown. Figure 4 The steps below will refer to the above Figure 2 and Figure 3 The components in question are discussed.

[0074] At block 402, the processor 303 can receive data indicating brainwave activity 228 over a specific time period. The data can be electroencephalogram (EEG) data obtained via the multi-electrode device 204. The data can be received by the processor 303 from an RF transmitter-receiver associated with the multi-electrode device 204. In some examples, the specific time period can be a first time period 222a, which is a predefined time period (e.g., an eight-hour time period) or can be defined by multiple sleep cycles (e.g., four sleep cycles). Additionally, the processor 303 can temporally divide the data into segments. The segments can be any suitable length of time (e.g., 3 minutes, 5 minutes, 1 hour, etc.).

[0075] At block 404, processor 303 can determine at least one metric 206a-b associated with at least one sleep state of the subject based on the data. Sleep state 216 can be characterized or defined by a wakeful sleep state, stage one, stage two, stage three, and rapid eye movement (REM), or sleep state 216 can be characterized or defined by REM and non-REM. Sleep state 216 can be predicted based on frequency patterns, amplitudes, or other suitable characteristics of EEG data 232 via a detection algorithm, a classification algorithm, or other suitable algorithm executed by processor 303. In some examples, a sleep state can be predicted for each of the segments of first time period 222a. The segments can then be reorganized based on their associated sleep states.

[0076] After predicting the sleep state 216, the processor 203 can derive metrics 206a-b from the EEG data 216 associated with the sleep state 216. For example, the first metric 206a can be the amount of time in a particular sleep state (e.g., REM). The amount of time in the particular sleep state can be accumulated over a predefined time period. In some examples, the amount of time in the particular sleep state can be the average amount of time in the particular sleep state for each sleep cycle based on a series of sleep cycles. In addition, the metrics 206a-b can include other suitable statistics derived from the predicted sleep state or from the EEG data 232. For example, the metrics 206a-b can be the ratio of REM to non-REM sleep, the amount of time spent sleeping (e.g., in stage 1, stage 2, stage 3, and REM), the average frequency of brain waves during a particular sleep stage, the average amplitude of brain waves during a particular sleep stage, etc.

[0077] At block 406, processor 303 can determine a risk level 202 associated with the subject's seizure activity based on at least one metric 206a-b. In some examples, processor 303 can implement artificial intelligence techniques to predict risk level 202 based on metrics 206a-b. Artificial intelligence techniques can include using one or more models or rules to generate subject-specific predictions based on metrics 206a-b derived from one or more segments of first time period 222a. Rules can be defined to predict REM deprivation, other suitable abnormal sleep patterns, and / or seizure propensity based on metrics 206a-b. For example, a rule can be defined to predict whether a subject is (or was) sleep deprived. For example, if less than a threshold percentage (e.g., 10%, 15%, 20%, 25%, 30%, or 35%) of the segments are predicted to be REM sleep, the rule can indicate that the subject is (or was) sleep deprived. The rule can further define thresholds based on the subject's age or other suitable characteristics. The rule can also include a gradient (i.e., multiple different levels of thresholds) that can correspond to risk levels. Thus, the rules may enable the processor to predict the risk level 202 based on the first metric 206a.

[0078] In some examples, historical data 236 may be EEG data of healthy people, people with a history of epileptic seizure activity, etc., or historical data 236 may be EEG data obtained for a previous subject. Thus, processor 303 may further use historical data 236 to train artificial intelligence technology or generate predefined rules, thresholds, etc. for determining risk level 202.

[0079] At box 408, the processor 303 can generate an output 322 indicating the risk level 202. The output 322 can indicate a high risk level, a medium risk level, or a low risk level. In some examples, a low risk level can be indicated by an output of a first color, a medium risk level can be indicated by an output of a second color, and a high risk level can be indicated by an output of a third color. The output 322 can also include metrics 206a-b associated with the predicted sleep state, a confidence value for the risk level, a likelihood value (i.e., a percentage probability of epileptic seizure activity), or other suitable indicators of the risk level 202. Furthermore, in some examples, the processor 303 can generate a second output indicating a second time period. Additionally or alternatively, the processor 303 can identify a treatment recommendation based on the risk level 202 and can include the treatment recommendation in the output 322. The treatment recommendation can be used to reduce the risk level 202 of the subject.

[0080] The foregoing description of certain examples, including illustrated examples, is presented for purposes of illustration and description only and is not intended to be exhaustive or to limit the disclosure to the precise form disclosed. Various modifications, adaptations, and uses thereof will be apparent to those skilled in the art without departing from the scope of the disclosure.

Claims

1. A computer-implemented method comprising: receiving data indicative of brainwave activity over a specific time period; determining, based on the data, at least one metric associated with at least one sleep state of the subject; determining a risk level associated with seizure activity in the subject based on the at least one metric; as well as An output is generated that is indicative of the risk level.

2. The computer-implemented method of claim 1 , wherein: The at least one metric is the amount of time in a particular sleep state.

3. The computer-implemented method of claim 2, wherein: The specific sleep state is the rapid eye movement (REM) state of sleep.

4. The computer-implemented method of any one of claims 1 to 3, wherein the data is received from an RF transmitter-receiver associated with a multi-electrode device.

5. The computer-implemented method of any one of claims 1 to 4, wherein the specific time period is a first time period, and wherein the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period, wherein the second time period occurs after the first time period.

6. The computer-implemented method according to any one of claims 1 to 5, wherein: The output is a first output and further includes: Identifying recommended actions based on the risk level; and A second output is generated that is indicative of the treatment recommendation, the treatment recommendation being usable to reduce the risk level.

7. The computer-implemented method according to any one of claims 1 to 6, wherein: The at least one metric is a first metric, and determining the risk level based on the at least one metric further comprises: Determine a second measure of healthy populations based on historical data; identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric; and Based on the statistically significant difference, the risk level is determined.

8. The computer-implemented method of any one of claims 1 to 7, wherein: Generating an output indicative of the risk level further comprises: providing an output indicating a high risk level in a first color; providing an output indicating a medium risk level in a second color; and Provides output in a third color indicating a low risk level.

9. A system comprising: one or more data processors; as well as A non-transitory computer-readable storage medium containing instructions that, when executed on the one or more data processors, cause the one or more data processors to: receiving data indicative of brainwave activity over a specific time period; determining, based on the data, at least one metric associated with at least one sleep state of the subject; determining a risk level associated with seizure activity in the subject based on the at least one metric; as well as An output is generated that is indicative of the risk level.

10. The system according to claim 9, wherein: The at least one metric is the amount of time in a particular sleep state.

11. The system of claim 10, wherein the specific sleep state is a rapid eye movement (REM) state of sleep.

12. The system of any one of claims 9 to 11, wherein the data is received from an RF transmitter-receiver associated with a multi-electrode device.

13. The system of any one of claims 9 to 12, wherein the specific time period is a first time period, and wherein the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period, wherein the second time period occurs after the first time period.

14. The system according to any one of claims 9 to 13, wherein: The output is a first output and further includes: Identifying recommended actions based on the risk level; and A second output is generated that is indicative of the treatment recommendation, the treatment recommendation being usable to reduce the risk level.

15. The system according to any one of claims 9 to 14, wherein: The at least one metric is a first metric, and determining the risk level based on the at least one metric further comprises: Determine a second measure of healthy populations based on historical data; identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric; and Based on the statistically significant difference, the risk level is determined.

16. The system according to any one of claims 9 to 15, wherein: Generating an output indicative of the risk level further comprises: providing an output indicating a high risk level in a first color; providing an output indicating a medium risk level in a second color; and Provides output in a third color indicating a low risk level.

17. A computer program product tangibly embodied in a non-transitory machine-readable storage medium, comprising instructions configured to cause one or more data processors to: receiving data indicative of brainwave activity over a specific time period; determining, based on the data, at least one metric associated with at least one sleep state of the subject; determining a risk level associated with seizure activity in the subject based on the at least one metric; as well as An output is generated that is indicative of the risk level.

18. The computer program according to claim 17, wherein The at least one metric is the amount of time in a particular sleep state.

19. The computer program of claim 18, wherein the specific sleep state is a rapid eye movement (REM) state of sleep.

20. A computer program according to any one of claims 17 to 19, wherein the data is received from an RF transmitter-receiver associated with a multi-electrode device.

21. The computer program of any one of claims 17 to 20, wherein the specific time period is a first time period, and wherein the risk level is a prediction of the likelihood that the subject will experience seizure activity in a second time period, wherein the second time period occurs after the first time period.

22. The computer program according to any one of claims 17 to 21, wherein the output is a first output, and further comprising: Identify treatment recommendations based on the stated risk level; as well as A second output is generated that is indicative of the treatment recommendation, the treatment recommendation being usable to reduce the risk level.

23. The computer program of any one of claims 17 to 22, wherein the at least one metric is a first metric, and determining the risk level based on the at least one metric further comprises: Determine a second measure of healthy populations based on historical data; identifying a statistically significant difference between the first metric and the second metric based on a comparison of the first metric and the second metric; as well as Based on the statistically significant difference, the risk level is determined.

24. A computer program according to any one of claims 17 to 23, wherein Generating an output indicative of the risk level further comprises: providing an output indicating a high risk level in a first color; providing an output indicating a medium risk level in a second color; and Provides output in a third color indicating a low risk level.

Images