Methods of identifying sleep and waking patterns and uses
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11696724-B2
Publication Date
2023-07-11
Expiration Date
Abstract
Traditional analysis of sleep patterns requires several channel of data. This analysis can be useful for customized analysis including assessing sleep quality, detecting pathological conditions, determining the effect of medication on sleep states and identifying biomarkers, and drug dosages or reactions.
Core Innovation
The invention provides a method and apparatus that access signal data indicative of brain activity of a subject using a noninvasive sensor comprising at least one electrode collected over a time period including one or more epochs. The approach generates spectrogram data in which, for each frequency band of a set of frequency bands, power is estimated at the frequency band for each epoch, and identifies a spectral signature of a particular frequency band whose fluctuation represents a relative amount of the power across the one or more epochs.
The spectral signature is estimated to be consistent with an indication that one or more physical movements of the subject have occurred during the one or more epochs. The method then generates an output based on the estimation that predicts an attribute of a sequence of predicted sleep stages over the time period, and determines that the output is consistent with a classification of a pathological condition by comparing the output to one or more reference attributes indicative of sleep stages associated with other subjects with known classifications.
In the provided embodiments, the pathological condition includes epilepsy, Alzheimer's disease, Parkinson's disease, depression, brain trauma, insomnia, restless leg syndrome, or sleep apnea. The framework further supports sleep-state related signatures derived from spectrogram analysis, including preferred-frequency space and fragmentation measures, and includes clustering of sleep states into wakefulness and sleep stages such as SWS and REM, including intermediate sleep.
Claims Coverage
Two independent claims are present: a method claim and a corresponding apparatus claim. Both cover the same core workflow components, with inventive features centered on noninvasive spectrogram-based spectral signatures, movement-consistent fluctuation estimation, sleep-stage sequence prediction, and pathological-condition classification using comparison to reference attributes.
Noninvasive spectrogram power per frequency band per epoch
Access signal data indicative of brain activity of a subject collected by a noninvasive sensor comprising at least one electrode over a time period that includes one or more epochs, and generating based on the signal data spectrogram data comprising, for each frequency band of a set of frequency bands, a power estimated at the frequency band for each epoch.
Spectral signature fluctuation consistent with physical movement
Identifying, based on the spectrogram data, a spectral signature of a particular frequency band of the set of frequency bands, wherein the spectral signature identifies a fluctuation of a relative amount of the power estimated across the one or more epochs, and estimating that the spectral signature is consistent with an indication that one or more physical movements of the subject have occurred during the one or more epochs.
Predicting an attribute of a sequence of predicted sleep stages
Generating, based on the estimation, an output that predicts an attribute of a sequence of predicted sleep stages over the time period.
Pathological-condition classification by comparison to reference sleep-stage attributes
Determining that the output is consistent with a classification of pathological condition of the subject by comparing the output to one or more reference attributes indicative of sleep stages associated with other subjects with known classifications, wherein the pathological condition includes one of epilepsy, Alzheimer's disease, Parkinson's disease, depression, brain trauma, insomnia, restless leg syndrome, or sleep apnea.
Noninvasive sensor and computing device configured to perform the method
Providing an apparatus comprising a noninvasive sensor comprising at least one electrode and a computing device configured to access signal data over a time period including one or more epochs, generate spectrogram data with power estimated per frequency band per epoch, identify a spectral signature identifying fluctuation of relative power across epochs, estimate that the spectral signature is consistent with an indication that one or more physical movements occurred, generate an output predicting an attribute of a sequence of predicted sleep stages, and determine that the output is consistent with a classification of pathological condition by comparing to reference attributes indicative of sleep stages.
Claim coverage centers on the same technical idea in both the method and apparatus: derive spectrogram-based spectral signatures whose relative power fluctuation is linked to physical movement, use the resulting estimation to predict attributes of a sequence of predicted sleep stages, and determine consistency with pathological-condition classifications by comparing to reference sleep-stage attributes from other subjects with known classifications.
Stated Advantages
Supports classification of pathological conditions based on predicted sleep-stage attributes compared to reference attributes indicative of sleep stages associated with other subjects.
Provides an output that predicts an attribute of a sequence of predicted sleep stages over the time period using spectrogram-derived spectral signatures.
Documented Applications
Biomarkers/diagnosis of pathological conditions including epilepsy, Alzheimer's disease, Parkinson's disease, depression, brain trauma, insomnia, restless leg syndrome, and sleep apnea.
Assessing medication/drug effects and drug-induced changes, including detection of disappearance of Stage II spindle activity after drug administration.
Interested in licensing this patent?