Systems and methods for designating REM and wake states

Abstract

The present disclosure provides systems and methods for analyzing whether a sleep epoch is a REM sleep epoch or a wake epoch. According to aspects of the present disclosure, a computer-implemented method includes accessing cardiopulmonary coupling data spanning a sleep period of a person, identifying an epoch of the sleep period corresponding to infrasonic coupling in the cardiopulmonary coupling data, accessing high-frequency coupling data and / or low-frequency coupling data in the cardiopulmonary coupling data corresponding to the epoch, and designating the epoch as a REM sleep epoch or a wake epoch based on the high-frequency coupling data and / or low-frequency coupling data corresponding to the epoch, wherein the epoch is designated based on the cardiopulmonary coupling data without using non-cardiopulmonary coupled physiological data.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of and priority to U.S. Provisional Patent Application No. 62 / 903,833, filed September 21, 2019, which is incorporated herein by reference in its entirety.

[0002] The present disclosure relates to sleep analysis, and more particularly to the analysis of cardiopulmonary coupling (CPC) data, or CPC data and physiological data, during a person's sleep periods to designate REM and wake states during the sleep periods. [Background technology]

[0003] Cardiopulmonary coupling is a technique for assessing sleep quality by performing a quantitative analysis between two physiological signals: a series of NN intervals from heart rate variability combined with a corresponding direct or derived respiratory signal, to determine the coherent cross-power of these two signals. Cardiopulmonary coupling is described, inter alia, in U.S. Patent Nos. 7,324,845, 7,734,334, 8,403,848, and 8,401,626, all of which are incorporated herein by reference in their entireties.

[0004] Cardiopulmonary coupling is characterized in terms of coupling frequency. High-frequency coupling represents stable sleep, a biomarker of integrated, stable N-REM sleep, and is associated with periods of stable breathing, high vagal tone, a generally non-periodic alternating pattern on the electroencephalogram (EEG), high relative delta power, a physiological drop in blood pressure (healthy), and / or a stable arousal threshold. In high-frequency coupling (HFC), the coupling frequency is above 0.1 Hz.

[0005] Low-frequency coupling represents unstable sleep, a biomarker of unsettled N-REM sleep, which has characteristics opposite to those of stable sleep. Unsettled sleep is associated with EEG activity called periodic alternating patterns (CAP), periods of fluctuating breathing patterns (variation in tidal volume), periodic heart rate fluctuations (CVHR), non-declining blood pressure, and / or variable arousal thresholds. Fragmented REM sleep has low-frequency coupling characteristics. In low-frequency coupling (LFC), the coupling frequency is in the range [0.01, 0.1] Hz. Low-frequency coupling can be further classified as elevated low-frequency coupling broadband or elevated low-frequency coupling narrowband.

[0006] Very low frequency coupling represents REM sleep and wakefulness. The frequency range below 0.01 Hz is defined as very low frequency coupling (vLFC). The physiology of REM and wakefulness is closely related to the electrooculogram, which is used as the primary tool for distinguishing the two states from the perspective of polysomnography (PSG). REM and wakefulness have very similar appearances in cardiopulmonary coupling (CPC), manifesting as very low frequency coupling (vLFC). Summary of the Invention [Problem to be solved by the invention]

[0007] There is interest in further developing and improving sleep analysis techniques to designate various sleep states based on cardiopulmonary coupling data. [Means for solving the problem]

[0008] The present disclosure relates to sleep analysis, and more particularly to the analysis of cardiopulmonary coupling (CPC) data, or CPC data and physiological data, during a person's sleep periods to designate REM and wake states during the sleep periods.

[0009] According to aspects of the present disclosure, a computer-implemented method includes accessing cardiopulmonary coupling data spanning a sleep period of a person; identifying an epoch of the sleep period comprising very low frequency coupling in the cardiopulmonary coupling data; accessing at least one of high frequency coupling data or low frequency coupling data in the cardiopulmonary coupling data corresponding to the epoch; and designating the epoch as a REM sleep epoch or a wake epoch based on the at least one of the high frequency coupling data or the low frequency coupling data corresponding to the epoch, wherein the epoch is designated based on the cardiopulmonary coupling data without using non-cardiopulmonary coupled physiological data.

[0010] In various embodiments of the method, an epoch exhibits a predominance of low frequency binding, and the method further comprises comparing the power of the very low frequency binding during said epoch to a threshold.

[0011] In various embodiments of the method, the threshold is based on at least one of the person, a condition of the person, or a population that includes the person.

[0012] In various embodiments of the method, designating the epoch includes designating the epoch as a REM sleep epoch based on (i) a dominance of low frequency binding and (ii) the power of infrasonic binding during the epoch exceeding a threshold.

[0013] In various embodiments of the method, the epoch exhibits a predominance of very low frequency coupling.

[0014] In various embodiments of the method, the dominance of infrasonic coupling for the epoch is based on the dominance of a predetermined upper limit of infrasonic coupling range in the cardiopulmonary coupling data corresponding to the epoch.

[0015] In various embodiments of the method, designating the epoch includes designating the epoch as a REM sleep epoch based on (i) a predominance of infrasonic coupling within a predetermined upper range of an infrasonic coupling range, and (ii) the power of at least one of the low-frequency coupling or the high-frequency coupling exceeds a threshold in at least one of the high-frequency coupling data or the low-frequency coupling data corresponding to the epoch.

[0016] In various embodiments of the method, designating the epoch includes designating the epoch as a REM sleep epoch based on (i) a predominance of very low frequency binding in the epoch, and (ii) the presence of a narrow band of elevated low frequency binding in low frequency binding data corresponding to the epoch.

[0017] According to aspects of the present disclosure, a system includes one or more processors and at least one memory storing instructions that, when executed by the one or more processors, cause the system to access cardiopulmonary coupling data spanning a sleep period of a person, identify epochs of the sleep period that include infrasonic coupling in the cardiopulmonary coupling data, access at least one of high frequency coupling data or low frequency coupling data in the cardiopulmonary coupling data corresponding to the epochs, and designate the epochs as REM sleep epochs or wake epochs based on the at least one of the high frequency coupling data or low frequency coupling data corresponding to the epochs, wherein the epochs are designated based on the cardiopulmonary coupling data without use of non-cardiopulmonary coupled physiological data.

[0018] In various embodiments of the system, an epoch exhibits a dominance of low frequency coupling, and the instructions, when executed by the one or more processors, cause the system to compare the power of the very low frequency coupling during said epoch with a threshold.

[0019] In various embodiments of the system, the threshold is based on at least one of the person, a condition of the person, or a population that includes the person.

[0020] In various embodiments of the system, in designating the epoch, the instructions, when executed by one or more processors, cause the system to designate the epoch as a REM sleep epoch based on (i) a dominance of low frequency binding and (ii) the power of infrasonic binding during the epoch exceeding a threshold.

[0021] In various embodiments of the system, the epochs exhibit a predominance of very low frequency coupling.

[0022] In various embodiments of the system, the dominance of infrasonic coupling for the epoch is based on the dominance of a predetermined upper limit of infrasonic coupling range in the cardiopulmonary coupling data corresponding to the epoch.

[0023] In various embodiments of the system, in designating the epoch, the instructions, when executed by one or more processors, cause the system to designate the epoch as a REM sleep epoch based on (i) a predominance of infrasonic coupling within a predetermined upper range of an infrasonic coupling range, and (ii) the power of at least one of the low-frequency coupling or the high-frequency coupling exceeds a threshold in at least one of the high-frequency coupling data or the low-frequency coupling data corresponding to the epoch.

[0024] In various embodiments of the system, in designating the epoch, the instructions, when executed by one or more processors, cause the system to designate the epoch as a REM sleep epoch based on (i) a predominance of very low frequency coupling in the epoch, and (ii) the presence of an elevated low frequency coupling narrow band in low frequency coupling data corresponding to the epoch.

[0025] According to aspects of the present disclosure, a computer-implemented method includes accessing cardiopulmonary coupling data spanning a sleep period for a person, classifying epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data, accessing actigraphy data for the person corresponding to the vLFC epochs, and designating the vLFC epochs as REM epochs based on a predetermined percentage of actigraphy measurements in the actigraphy data corresponding to the vLFC epochs indicating movement below the motor threshold.

[0026] In various embodiments of the method, the predetermined percentage is 95% of the actigraphy measurements in the actigraphy data corresponding to the vLFC epoch, and the motor threshold is 0.01 G / s.

[0027] In various embodiments of the method, the method includes varying at least one of the predetermined percentage or motor threshold for different actigraphy sensors.

[0028] According to aspects of the present disclosure, a system includes one or more processors and at least one memory storing instructions that, when executed by the one or more processors, cause the system to access cardiopulmonary coupling data spanning a sleep period for a person, classify epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data, access actigraphy data for the person corresponding to the vLFC epochs, and designate the vLFC epochs as REM epochs based on a predetermined percentage of actigraphy measurements in the actigraphy data corresponding to the vLFC epochs indicating movement below a motor threshold.

[0029] In various embodiments of the system, the predetermined percentage is 95% of the actigraphy measurement and the motor threshold is 0.01 G / s in the actigraphy data corresponding to said vLFC epoch.

[0030] In various embodiments of the system, the instructions, when executed by the one or more processors, further cause the system to vary at least one of the predetermined percentages or motor thresholds for various actigraphy sensors.

[0031] According to aspects of the present disclosure, a computer-implemented method includes accessing cardiopulmonary coupling data over a period of sleep for a person; classifying epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data; accessing pseudo-actigraphy data for the person corresponding to the vLFC epochs, the pseudo-actigraphy data being based on physiological measurements of the person and not on actigraphy measurements; and designating the vLFC epochs as REM epochs or wake epochs based on the pseudo-actigraphy data corresponding to the vLFC epochs.

[0032] In various embodiments of the method, the method includes generating pseudo actigraphy data corresponding to the vLFC epoch based on signal quality of the physiological measure.

[0033] In various embodiments of the method, generating pseudo actigraphy data includes generating data corresponding to greater movement when signal quality is lower and generating data corresponding to less movement when signal quality is higher.

[0034] In various embodiments of the method, the physiological measurements include at least one of ECG measurements or plethysmography measurements of the person.

[0035] In various embodiments of the method, generating pseudo-actigraphy data includes processing physiological measurements to detect peaks during vLFC epochs, and generating data corresponding to smaller movements when the count of detected peaks is below a predetermined threshold and the shape of the detected peaks matches the shape of expected peaks, and generating data indicating larger movements when the count of detected peaks is above a predetermined threshold and the shape of the detected peaks differs from the shape of expected peaks.

[0036] In various embodiments of the method, the physiological measurement comprises an oxygen saturation measurement.

[0037] According to aspects of the present disclosure, a system includes one or more processors and at least one memory storing instructions that, when executed by the one or more processors, cause the system to access cardiopulmonary coupling data over a period of sleep for a person, classify epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data, access pseudoactigraphy data for the person corresponding to the vLFC epochs, the pseudoactigraphy data being based on physiological measurements and not actigraphy measurements for the person, and designate the vLFC epochs as REM epochs or wake epochs based on the pseudoactigraphy data corresponding to the vLFC epochs.

[0038] In various embodiments of the system, the instructions, when executed by one or more processors, cause the system to generate pseudo-actigraphy data corresponding to vLFC epochs based on signal quality of the physiological measurements.

[0039] In various embodiments of the system, when generating simulated actigraphy data, the instructions, when executed by one or more processors, cause the system to generate data corresponding to greater movement when signal quality is lower and data corresponding to less movement when signal quality is higher.

[0040] In various embodiments of the system, the physiological measurements include at least one of ECG measurements or plethysmography measurements of the person.

[0041] In various embodiments of the system, when generating simulated actigraphy data, the instructions, when executed by one or more processors, cause the system to process physiological measurements to detect peaks during the vLFC epochs, generate data corresponding to smaller movements when the count of detected peaks is below a predetermined threshold and the shape of the detected peaks matches the shape of expected peaks, and generate data indicative of larger movements when the count of detected peaks is above a predetermined threshold and the shape of the detected peaks differs from the shape of expected peaks.

[0042] In various embodiments of the system, the physiological measurements include oxygen saturation measurements.

[0043] According to aspects of the present disclosure, a computer-implemented method includes accessing cardiopulmonary coupling data over a period of sleep for a person; classifying epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data; accessing physiological data for the person corresponding to the vLFC epochs, the physiological data including physiological measures and not including actigraphy measures; and designating the vLFC epochs as REM epochs based on the physiological data corresponding to the vLFC epochs indicative of sleep disordered breathing.

[0044] In various embodiments of the method, the physiological measurements include oxygen saturation measurements, and the method includes processing the oxygen saturation measurements to identify sleep disordered breathing events during the vLFC epoch.

[0045] According to aspects of the present disclosure, a system includes one or more processors and at least one memory storing instructions that, when executed by the one or more processors, cause the system to access cardiopulmonary coupling data over a period of sleep for a person, classify epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data, access physiological data for the person that includes physiological measures and does not include actigraphy measures and corresponds to the vLFC epochs, and designate the vLFC epochs as REM epochs based on the physiological data corresponding to the vLFC epochs indicative of sleep-disordered breathing.

[0046] In various embodiments of the system, the physiological measurements include oxygen saturation measurements, and the instructions, when executed by the one or more processors, cause the system to process the oxygen saturation measurements to identify sleep disordered breathing events during the vLFC epoch.

[0047] The objects and features of the disclosed systems and methods will become apparent to those skilled in the art upon reading the description of various embodiments thereof in conjunction with the accompanying drawings. [Brief explanation of the drawings]

[0048] [Figure 1] FIG. 1 is a diagram of an exemplary measurement system according to aspects of the present disclosure. [Figure 2] FIG. 10 is a flow diagram of example operations for designating REM / wake states based on coupled cardiopulmonary data, according to aspects of the present disclosure. [Figure 3] FIG. 10 is a diagram of an example epoch with non-zero vLFC power during LFC dominance, according to an embodiment of the present disclosure. [Figure 4] FIG. 10 is a diagram of an example epoch with eLFCNB during vLFC dominance, according to an aspect of the present disclosure. [Figure 5] FIG. 10 is a flow diagram of example operations for designating REM / wake states based on CPC and actigraphy data, according to aspects of the present disclosure. [Figure 6]FIG. 10 is a diagram of an exemplary epoch having vLFC dominance and corresponding to an actigraphy signal indicating movement, according to aspects of the present disclosure. [Figure 7] FIG. 10 is a flow diagram of example operations for designating REM / wake states based on CPC data and pseudo actigraphy data, according to aspects of the present disclosure. [Figure 8] 1 is a graph of an exemplary signal quality measurement and actigraphy signal, according to aspects of the present disclosure. [Figure 9] FIG. 1 is a block diagram of an exemplary computing system according to aspects of the present disclosure. [Figure 10] FIG. 10 is a flow diagram of example operations for designating REM / wake states based on CPC data and various physiological signals, according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0049] The present disclosure relates to the analysis of cardiopulmonary coupled (CPC) data, or CPC data and physiological data, during a person's sleep period to designate REM and wake states during the sleep period. During REM sleep, the subject remains largely motionless or in a state of "skeletal muscle paralysis," with the primary mechanical movement being ocular. Because REM manifests as vLFC and should be absent of any significant movement, one approach according to the present disclosure identifies REM states based on vLFC without sufficient actigraphy and identifies wake states based on vLFC with sufficient actigraphy. Other aspects of the present disclosure do not use actigraphy data to designate REM or wake states. For example, pseudo-actigraphy data can be used, as described in more detail later herein. Other aspects of the present disclosure use only cardiopulmonary coupled data to designate REM or wake states without using non-CPC physiological data, as described later herein.

[0050] Referring now to FIG. 1 , a diagram of an exemplary measurement system 100 according to aspects of the present disclosure is shown. The measurement system 100 can be attached to a sleeping person to obtain physiological measurements that can be used to calculate cardiopulmonary coupling (“CPC”), such as electrocardiogram measurements or other physiological measurements. The measurement system 100 also obtains various measurements, such as ECG measurements, plethysmography measurements, oxygen saturation measurements, and / or actigraphy measurements, the uses of which are described later in this specification. FIG. 1 illustrates various exemplary sensors that can be placed on various parts of a person's body, including parts not shown in FIG. 1 . For example, various sensors can be placed on the person's torso, head, and / or limbs, among other locations. Various sensors for detecting physiological signals will be understood by those skilled in the art. For example, in various embodiments, the sensors can be sensors that contact the person's body or can be touchless sensors that do not directly contact the person (e.g., sensors based on ballistocardiography). The physiological measurements can be recorded on a storage medium, such as a disk drive, flash drive, solid-state drive, or other storage medium. In various embodiments, various physiological measurements can be recorded in parallel. In various embodiments, each recorded data can be tagged with or associated with a timestamp. By tagging or associating the recorded data with a timestamp, the various recorded measurements can be correlated in time. It is contemplated that other methods of correlating the recorded measurements in time can be used.

[0051] One aspect of the present disclosure relates to a system and method for designating REM or wakefulness states based on cardiopulmonary coupling spectral analysis without using non-CPC data. As described above, very low frequency coupling (vLFC) represents REM sleep or wakefulness states. Figure 2 illustrates an example operation of analyzing CPC data from a person's sleep period to distinguish between REM sleep and wakefulness without using non-CPC physiological data. In the description herein, a sleep period can be divided into segments, sometimes referred to herein as "epochs." In various embodiments, different epochs can have the same duration, or different epochs can have different durations.

[0052] Referring to FIG. 2 , in block 210, the operations include accessing cardiopulmonary coupling data spanning a sleep period of a person. The CPC data can be accessed, for example, from the measurement system of FIG. 1 or another system. In block 220, the operations include identifying an epoch of the sleep period that includes very low frequency coupling in the cardiopulmonary coupling data. As described in more detail below, an epoch that includes very low frequency coupling (vLFC) may be vLFC dominant (i.e., the highest high frequency coupling power is in the vLFC band) or may not be vLFC dominant. In block 230, the operations include accessing high frequency coupling data and / or low frequency coupling data in the cardiopulmonary coupling data corresponding to the epoch. In block 240, the operations include designating the epoch as a REM sleep epoch or a wake epoch based on the high frequency coupling data and / or low frequency coupling data corresponding to the epoch, without using non-cardiopulmonary coupling physiological data. The operations of FIG. 2 can be performed by a computing system such as the computing system of FIG. 9 , which will be described later in this specification. An embodiment of the operation of FIG. 2 will now be described.

[0053] The following discussion includes vLFC coupling, but is for epochs in which the dominant CPC state is classified as low-frequency coupling (LFC), i.e., the highest frequency coupling power is in the LFC band. During such epochs, vLFC power is non-zero and less than LFC power.

[0054] According to aspects of the present disclosure, for epochs with non-zero vLFC power and LFC dominance, such epochs are characterized as fragmented REM rather than unstable N-REM. During fragmented REM epochs, and in the absence of dominance of vLFC frequency bands within the upper range, such epochs can be designated as REM sleep states based on dynamic thresholds applied to vLFC bands, which may vary from person to person. An example of such an epoch is shown in FIG. 3, where epoch 310 is shown to have non-zero vLFC power and LFC dominance.

[0055] In such epochs with non-zero vLFC power and LFC dominance, REM classification may not be based on a fixed threshold, in that the designation is based on non-zero vLFC power rather than vLFC dominance. Rather, a dynamic threshold allows for more accurate indication of REM sleep and wake states. For example, a particular fixed threshold may be appropriate for a person with a healthy sleep state, but that particular threshold may not accurately designate REM sleep and wake states for a person with an unhealthy sleep state, whose worsening conditions may affect the vLFC band. Therefore, dynamic thresholds appropriate for various conditions, individuals, or populations, among others, may be used to designate REM sleep and wake states in epochs with non-zero vLFC power and LFC dominance (e.g., FIG. 3 ). In various embodiments, the dynamic threshold may be based on a particular population average. For example, if an epoch is LFC-dominant and the vLFC power is above the average vLFC power of a particular population, the epoch may be classified as REM. Other types of dynamic thresholds are contemplated within the scope of this disclosure.

[0056] The following description relates to epochs in which the predominant CPC state is classified as vLFC. According to aspects of the present disclosure, the designation of a vLFC epoch as a REM or wake state is based on an analysis of the CPC frequency band after the predominant CPC state is classified as vLFC.

[0057] In various embodiments, vLFC-dominant epochs can be designated as REM states when there is activation (e.g., non-zero power) in the LFC and / or HFC bands and when the dominant CPC frequency is within the upper range of the vLFC band (e.g., close to but not exceeding 0.01 Hz). Figure 4 shows examples of such epochs 410, 420.

[0058] In various embodiments, vLFC-dominant epochs are characterized by elevated low-frequency narrow-band (eLFC) NB ) is also present, it can be designated as a REM state. Low-frequency coupling is also called elevated low-frequency coupling broadband (eLFC). BB ) or fragmented, elevated low frequency coupled narrow band (eLFC) NB ) or periodic, or without elevated low-frequency coupling. NB eLFC is a marker of periodicity and is associated with periodic breathing, Cheyne-Stokes respiration, and central apnea. BB may be caused by other disorders such as pain or other disturbances during sleep that cause fragmentation, whereas eLFC NB can be caused by periodic limb movements.

[0059] Figure 4 shows the eLFC NB 4 shows examples of vLFC-dominant epochs 410, 420 where REM is also present. During testing, such epochs were designated as REM states based on polysomnography data and also accurately designated based on CPC data alone. As used herein, elevated very low frequency narrow band (eVLFC) NB ), or "periodic REM" sleep, is a sleep pattern characterized by elevated low-frequency narrow-band (eLFC) activity that predominates in the vLFC frequency band.NB ) is used to identify the occurrence of eVLFC. NB It will be appreciated that REM, or periodic REM, can serve as new CPC states. Therefore, a method for designating REM sleep and wake states is available for eVLFC. NB as a REM sleep state. NB If no REM power is present, dominance within the upper range of the vLFC band (e.g., CPC frequency power above 0.05 but below total power in the vLFC band) can be used to designate an epoch as a REM sleep state, as previously described.

[0060] 2-4 and the above description illustrate embodiments in which epochs can be designated as REM sleep states by analyzing only CPC data without using non-CPC data. The above-described embodiments and the embodiments in FIGS. 2-4 are exemplary and are not intended to limit the scope of the present disclosure.

[0061] Another aspect of the present disclosure relates to systems and methods for designating REM or wakefulness states based on analyzing cardiopulmonary data and various physiological signals such as actigraphy, oxygen saturation, and / or pseudo-actigraphy physiological signals such as ECG and plethysmography. The use of CPC data and actigraphy data is described in connection with Figures 5 and 6. The use of CPC data and pseudo-actigraphy physiological data and / or oxygen saturation data is described in connection with Figures 7 and 10.

[0062] According to aspects of the present disclosure, FIG. 5 illustrates a flow diagram of exemplary operations for designating a REM or wakefulness state based on analyzing cardiopulmonary data and actigraphy data. The operations apply a threshold to movement measurements to designate an epoch as a REM or wakefulness state. In various embodiments, to collect movement information, the recording device (e.g., hardware) includes an accelerometer sensor. An accelerometer is a sensor device that measures the acceleration (rate of change of velocity) of an object, which in this disclosure is a person. According to aspects of the present disclosure, the raw actigraphy signal can be processed to generate a quantity that reports acceleration in a particular unit of measurement. A common unit of measurement is m / s. 2 or G-force. Actigraphy data can be acquired and stored, for example, using the system of FIG.

[0063] Continuing with reference to FIG. 5, at block 510, the operations include accessing cardiopulmonary coupling data spanning a sleep period for a person. At block 520, the operations include classifying an epoch of the sleep period as a very low frequency coupling (vLFC) epoch based on the cardiopulmonary coupling data. The classification may be based on vLFC dominance in the epoch. At block 530, the operations include accessing actigraphy data for the person corresponding to the vLFC epoch. The actigraphy data may be accessed from a storage or computing system, as described in connection with FIG. 9. The actigraphy data corresponding to a vLFC epoch may be identified, for example, based on a timestamp. At block 540, the operations include designating the vLFC epoch as a REM epoch based on a predetermined percentage of actigraphy measurements in the actigraphy data corresponding to the vLFC epoch indicating movement below the motor threshold. An example is provided below.

[0064] In various embodiments, a threshold of 0.01 G / s may be used to designate REM and wake states, such that accelerations below 0.01 G / s are treated as indicative of REM sleep and accelerations equal to or greater than 0.01 G / s are treated as indicative of a wake state. The particular value of the threshold is exemplary, and other values ​​may be used. In various embodiments, the number of acceleration samples above the threshold is compared to the total number of samples in an epoch to generate a measure for designating the epoch as a REM sleep or wake state. In various embodiments, if 95% of the acceleration samples in an epoch are below the threshold, the epoch may be designated as a REM state. Otherwise, the epoch is designated as a wake state. The percentage threshold is exemplary, and other values ​​may be used. In various embodiments, the length of the epoch or period analyzed may be varied to increase the certainty of deriving a measure of fragmentation, or lack thereof.

[0065] In various embodiments, the acceleration threshold may need to be modified based on the accelerometer's hardware and firmware specifications (e.g., dynamic range, sampling rate, etc.) For example, a new accelerometer sensor may need to be analyzed and compared to a reference device to set the acceleration threshold for designating REM / wake.

[0066] Figure 6 shows an example of CPC and physiological data with REM / wake states indicated based on the behavior in Figure 5. Prolonged periods of wakefulness and all REM periods scored by PSG are shown in purple boxes for comparison with the PSG reference.

[0067] According to aspects of the present disclosure, FIG. 7 illustrates a flow diagram of an example operation for designating a REM or wakefulness state based on analyzing cardiopulmonary data and pseudo actigraphy data. As used herein, the term "pseudo actigraphy signal" or data refers to a non-actigraphy physiological signal having certain characteristics indicative of actigraphy. Pseudo actigraphy signals may include, for example, ECG signals, plethysmography signals, and oxygen saturation signals, among others. In various embodiments, pseudo actigraphy signals may be physiological signals that exhibit increased signal quality when a person is in a REM state and decreased signal quality when a person is in a wakeful state. Signal quality may be degraded by, for example, changes affecting signal strength, signal availability, or signal presence, among others. According to some aspects, intermittent degradation of signal quality may be correlated with motion artifacts and utilized as pseudo actigraphy. In various embodiments, intermittent degradation of signal quality that is vLFC dominant can be designated as a wake state, while near-original or original signal quality that is vLFC dominant can be designated as a REM state. Thus, the disclosed systems and methods can analyze simulated actigraphy signals to designate epochs as REM or wake states.

[0068] Continuing with reference to FIG. 7, at block 710, the operations include accessing cardiopulmonary coupling data spanning a sleep period for a person. At block 720, the operations include classifying an epoch of the sleep period as a very low frequency coupling (vLFC) epoch based on the cardiopulmonary coupling data. For example, the vLFC epoch may exhibit vLFC dominance. At block 730, the operations include accessing pseudo actigraphy data for the person corresponding to the vLFC epoch. The pseudo actigraphy data is based on physiological measurements of the person, not on actigraphy measurements. As previously mentioned, the pseudo actigraphy data may include, among other things, an ECG signal, a plethysmography signal, and an oxygen saturation signal. At block 740, the operations include designating the vLFC epoch as a REM epoch or a wake epoch based on the pseudo actigraphy data corresponding to the vLFC epoch. The operations of FIG. 7 can be implemented in a computing system, such as the computing system of FIG. 9, which will be described later in this specification. Various embodiments of the operations of FIG. 7 are described below.

[0069] According to aspects of the present disclosure, with respect to ECG and plethysmography signals, signal quality can be quantified by evaluating feature extraction performance of the ECG and plethysmography signals. For ECG, such features include, but are not limited to, the R-peak, P-wave, ST-segment, and / or QRS complex, among others. For plethysmography, such features include, but are not limited to, the systolic peak, diastolic peak, and / or dicrotic notch, among others. Signal degradation and / or lack of detected features results in poor signal quality results.

[0070] In various embodiments, certain features may be rejected. During periods of weak signal, the detector may fail to detect a feature in the signal, resulting in a "missing feature." In various embodiments, all marked features may be compared to a preset template, and correlations with the template may be calculated. During periods of motion artifact, the signal may be distorted, causing a low correlation between the marked feature and the template, resulting in the feature being rejected as a "rejected feature."

[0071] In various embodiments, signal quality can be quantified by comparing the amount of features detected to the expected number of features detected over a given time segment. The length of the time segment can vary depending on the desired granularity. In various embodiments, signal quality can be expressed as the percentage of expected features over a selected time segment that are marked features.

[0072] In various embodiments, when the detected feature is a signal peak, signal quality can be quantified as data corresponding to less movement when the count of detected peaks is below a predetermined threshold and the shape of the detected peaks matches the shape of the expected peaks, and as data showing more movement when the count of detected peaks is above a predetermined threshold and the shape of the detected peaks differs from the shape of the expected peaks.

[0073] According to aspects of the present disclosure, with respect to the oxygen saturation signal, the signal quality of the oxygen saturation signal is based on evaluating the value and rate of change. The oxygen saturation reports blood oxygen saturation within the range of [0%, 100%]. If the sensor is completely disconnected from the person (e.g., due to movement), the oxygen saturation value is expected to drop to 0%. During these periods, the signal quality will be zero (0). Intermittent disconnection from the subject, such as due to movement, can cause the sensor to loose contact, but not completely disconnect. In these situations, the oxygen saturation value will rapidly drop at a rate that is not possible in real human physiology.

[0074] According to aspects of the present disclosure, the oxygen desaturation rate is assessed and compared to a threshold, such as a threshold change of 3% / second (i.e., 0.03 / s) or another value. In various embodiments, during periods when the threshold is exceeded, the signal quality value can be set to 0; otherwise, the signal quality value can be set to 1. Other methods of scoring signal quality are contemplated within the scope of the present disclosure. The 3% threshold is exemplary, and other values ​​can be used. In various embodiments, the threshold can be modified to vary sensitivity, and care can be taken to avoid setting the threshold such that false negative desaturations outweigh true desaturations.

[0075] In various embodiments, signal quality can be quantified by comparing the amount of features detected to the expected number of features detected over a given time segment. The length of the time segment can vary depending on the desired granularity. In various embodiments, signal quality can be expressed as the percentage of expected features over a selected time segment that are marked features.

[0076] Accordingly, various examples of pseudo actigraphy signals are described, including ECG, polysomnography, and oxygen saturation. These examples are provided for illustrative purposes and are not intended to limit the scope of the present disclosure. Other physiological signals can be used as pseudo actigraphy signals and are contemplated as being within the scope of the present disclosure. The embodiments described herein for determining signal quality of a physiological signal are exemplary, and other methods of determining signal quality as a pseudo actigraphy signal are contemplated as being within the scope of the present disclosure.

[0077] 8 is a graph of exemplary signal quality for an exemplary actigraphy signal and a simulated actigraphy signal for a person, with the signal quality score over each time segment on the left y-axis, actigraphy (G / s) on the right y-axis, and sample number on the x-axis. As shown in FIG. 8, the signal quality of the actigraphy signal and the simulated actigraphy signal are inversely correlated, as the graph shows how the number of high signal quality features decreases with increasing actigraphy.

[0078] According to embodiments of the present disclosure, and with continued reference to FIG. 7 , when oxygen saturation data is available, the disclosed systems and methods can analyze the oxygen saturation data to designate epochs as REM or awake states. According to embodiments of the present disclosure, blood oxygen saturation (e.g., SO2, SaO2, SpO2) measurements can be analyzed to detect periods of oxygen desaturation events, which are often associated with sleep-disordered breathing. The presence of such events during periods of vLFC dominance indicates REM sleep. Various techniques and approaches for identifying oxygen desaturation events and / or sleep-disordered breathing events can be used, such as the techniques described in International Application Publication No. WO2020061014A1, the entire contents of which are incorporated herein by reference. Other techniques and approaches for identifying oxygen desaturation events and / or sleep-disordered breathing events using oxygen saturation data are contemplated within the scope of the present disclosure.

[0079] While each of the aforementioned techniques (e.g., Figures 2, 5, and 7) can be used independently, they can also be used in combination to increase certainty, accuracy, and / or versatility based on the signals available to assist in designating REM / wake states. These techniques can also aid in the diagnosis of sleep disorders specific to REM sleep states. For example, "REM apnea" is considered a subcategory of sleep-disordered breathing in which apnea / hypopnea events occur during REM sleep. To this end, the presence of an oxygen saturation signal may improve the accuracy of disease classification. Furthermore, the complete absence of periods classified as REM through the use of actigraphy or by analysis of the actigraphy signal during periods indicative of REM by CPC spectral analysis may indicate the presence of REM behavior disorder (RBD), in which REM periods are associated with mechanical movements (including sleepwalking).

[0080] FIG. 10 is a flowchart illustrating exemplary operations for determining which signals to use for REM / wake classification. The operations of FIG. 10 can be implemented in a computing system, such as the computing system of FIG. 9, and are described below. At block 1010, the operations include reading a data file to access available signals. At block 1020, the operations include determining whether an ECG signal and / or a plethysmography signal is available. If not, the operations can end at block 1022. If such signals are present, at block 1030, the operations include performing CPC processing to generate an array of CPC epochs, each epoch being classified into one of three CPC states based on the dominant frequency band: HFC, LFC, or vLFC. This process can be referred to as "base labeling." Additionally, each epoch can be classified as no eLFC, eLFC, or vLFC. BB , or eLFC NB This process may be referred to as "extended labeling." In block 1040, for each epoch classified as vLFC, the extended label is added to eLFC, as described previously herein. NBIf the epoch is classified as REM, then the epoch may be designated as a REM state. If a REM / wake state is not indicated in block 1030, operation continues to block 140.

[0081] At block 1040, the operations determine whether actigraphy data is present. If present, the operations include designating the epoch as a REM state if the motion artifact is below a predetermined threshold for a sufficient number of samples, as described above in connection with FIG. 5. Otherwise, the epoch is designated as a wake state. If actigraphy data is not available, the operations continue to block 1050.

[0082] At block 1050, the operations include using the pseudo-actigraphy signal to designate a REM / wake state, as described in connection with Figure 7. For example, if the sum of detected excess and missed NN intervals is below a predetermined threshold and there are no artifacts in the oxygen saturation signal quality (if present), the epoch can be designated as a REM state.

[0083] At block 1060, the operations include determining whether an oxygen saturation signal is present. If present, the operations determine whether a desaturation event is present and artifact-free. If a desaturation event is present and artifact-free, the operations can designate the epoch as a REM state.

[0084] Thus, an array of sleep stage classifications 1070 is generated based on the above operations. The foregoing operations are exemplary, and variations are contemplated within the scope of this disclosure. For example, in various embodiments, the presence of a desaturation event as determined by block 1060 may override the determination of blocks 1040 and / or 1050, or blocks 1040 and / or 1050 may designate the epoch as an "unknown" state.

[0085] The operations of FIG. 10 are exemplary, and other methods of using a combination of physiological and CPC data to determine REM / wake states are contemplated within the scope of this disclosure.

[0086] Aspects and embodiments of the present disclosure may be implemented in one or more computing systems capable of performing the functions described herein. Referring to Figure 9, an example computing system 900 for implementing the present disclosure is shown. Various embodiments of the present disclosure described herein may be performed by computing system 900. However, it will be apparent to one skilled in the art how to implement the present disclosure using other computer systems and / or computer architectures.

[0087] Computing system 900 includes one or more processors, such as processor 904. Processor 904 is connected to a communications infrastructure 906 (e.g., a communications bus, crossover bar, or network).

[0088] The computing system 900 may include a display 930 that receives graphics, text, and other data from the communications infrastructure 906 (or from a frame buffer, not shown) for display. In various embodiments, the display 930 may present the various measurements and metrics described herein, including CPC data, actigraphy data, oxygen saturation data, ECG data, and / or plethysmography data, among others. In various embodiments, the display 930 may present graphs and numerical representations. The presentations and reports may include some or all of the various metrics disclosed herein above.

[0089] Computing system 900 also includes main memory 908, preferably random access memory (RAM), and may also include secondary memory 910. Secondary memory 910 may include, for example, a hard disk drive 912 and / or a removable storage drive 914, representing a floppy disk drive, magnetic tape drive, optical disk drive, etc. Removable storage drive 914 reads from and / or writes to a removable storage unit 918, in well-known fashion. Removable storage unit 918 represents a floppy disk, magnetic tape, optical disk, etc. that is read from and written to by removable storage drive 914. As will be appreciated, removable storage 918 includes a computer-usable storage medium having stored thereon computer software (e.g., programs or other instructions) and / or data.

[0090] In various embodiments, secondary memory 910 may include other similar devices for allowing computer software and / or data to be loaded into computing system 900. Such devices may include, for example, removable storage 922 and interfaces 920. Such examples may include program cartridges and cartridge interfaces (such as those found in legacy devices), removable memory chips (such as EPROMs or PROMs) and associated sockets, and other removable storage devices 922 and interfaces 920 that allow software and data to be transferred from the removable storage devices 922 to computing system 900.

[0091] Computing system 900 may also include a communications interface 924. Communications interface 924 allows software and data to be transferred between computing system 900 and external devices. Examples of communications interface 924 include a modem, a network interface (such as an Ethernet or WiFi card), a communications port, a PCMCIA or SD or other slot, and a card, among other components. The software and data transferred via communications interface 924 are in the form of signals 928, which may be electronic, electromagnetic, optical, or other signals that can be received by communications interface 924. These signals 928 are provided to communications interface 924 via communications path (i.e., channel) 926. Communications path 926 carries signals 928 and may be implemented using wire or cable, optical fiber, telephone line, cellular phone link, RF link, free-space optics, and / or other communications channels.

[0092] As used herein, the terms "computer program medium" and "computer usable medium" are used generally to refer to media such as removable storage 918, removable storage 922, a hard disk inserted in hard disk drive 912, and signal 928. These computer program products are devices for providing software to computing system 900. The present disclosure includes such computer program products.

[0093] Computer programs (also called computer control logic or computer-readable program code) are stored in main memory 908 and / or secondary memory 910. Computer programs may also be received via communications interface 924. Such computer programs, when executed, enable computing system 900 to implement the present disclosure as described herein. In particular, when executed, the computer programs enable processor 904 to perform the processes and operations of the present disclosure, such as, for example, the various steps of methods 200, 300, 400, 500, and 600 described above. Such computer programs therefore represent controllers of computing system 900.

[0094] In embodiments in which the present disclosure is implemented using software, the software may be stored on a computer program product and loaded into computing system 900 using removable storage drive 914, hard drive 912, interface 920, or communication interface 924. The control logic (software), when executed by processor 904, causes processor 904 to perform the functions of the present disclosure as described herein. Thus, the techniques of the present disclosure may be provided as a software medical device (SaMD) or as non-medical software. In various embodiments, the software may include a cloud-based application.

[0095] The embodiments disclosed herein are examples of the present disclosure and may be embodied in various forms. For example, although certain embodiments herein are described as separate embodiments, each of the embodiments herein may be combined with one or more of the other embodiments herein. The specific structural and functional details disclosed herein are not limiting and should be construed as a representative basis for teaching those skilled in the art to employ the present disclosure in a variety of ways in substantially any appropriately detailed structure.

[0096] The phrases "in one embodiment," "in an embodiment," "various embodiments," "in some embodiments," or "in other embodiments" may each refer to one or more of the same or different embodiments according to the disclosure. A phrase of the form "A or B" means "(A), (B), or (A and B)." A phrase of the form "at least one of A, B, or C" means "(A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C)."

[0097] Any of the methods, programs, algorithms, or codes described herein may be converted into or expressed in a programming language or computer program. As used herein, the terms "programming language" and "computer program" each include any language used to specify instructions to a computer, including, but not limited to, the following languages ​​and their derivatives: Assembler, Basic, batch files, BCPL, C, C+, C++, Delphi, Fortran, Java, JavaScript, machine code, operating system command languages, Pascal, Perl, PL1, Python, scripting languages, Visual Basic, metalanguages ​​for specifying programs, and all first-, second-, third-, fourth-, fifth-, or later-generation computer languages. Databases and other data schemas, and any other metalanguages, are also included. No distinction is made between languages ​​that are interpreted, compiled, or that use both compiled and interpreted approaches. No distinction is made between compiled and source versions of a program. Thus, when a programming language can exist in more than one state (such as source, compiled, object, or linked), program is a reference to any and all such states.

[0098] The systems described herein may also utilize one or more controllers to receive various information and transform the received information to generate output. A controller may include any type of computing device, computational circuit, or any type of processor or processing circuit capable of executing a sequence of instructions stored in a memory. A controller may include multiple processors and / or multi-core central processing units (CPUs), and may include any type of processor, such as a microprocessor, digital signal processor, microcontroller, programmable logic device (PLD), field programmable gate array (FPGA), etc. A controller may also include memory for storing data and / or instructions that, when executed by the one or more processors, cause the one or more processors to perform one or more methods and / or algorithms.

[0099] It should be understood that the foregoing description merely illustrates the present disclosure. Various alternatives and modifications may be devised by those skilled in the art without departing from the present disclosure. Accordingly, the present disclosure is intended to embrace all such alternatives, modifications, and variations. The embodiments described with reference to the accompanying drawings are presented only to illustrate particular examples of the present disclosure. Other elements, steps, methods, and techniques that differ only slightly from those described above are also intended to be within the scope of the present disclosure.

Claims

1. 1. A computer-implemented method comprising: (A) accessing cardiopulmonary coupling data over a period of sleep for a person; (B) classifying a subset of epochs within the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data; (C) accessing pseudo-actigraphy data that is inversely correlated with the actigraphy data, the pseudo-actigraphy data being obtained by quantifying signal quality based on physiological measurements of the person corresponding to the vLFC epochs; (D) designating the vLFC epoch as a REM epoch or a wake epoch based on the pseudo actigraphy data corresponding to the vLFC epoch; Including, Step C of quantifying the signal quality of the physiological measurement comprises: (C1) processing the physiological measurements to detect peaks in the vLFC epochs; (C2-1) generating data indicative of higher signal quality when the count of detected peaks is below a predetermined threshold and when the shape of the detected peaks matches the shape of an expected peak; (C2-2) generating data indicative of lower signal quality when the count of the detected peaks exceeds a predetermined threshold and when the shape of the detected peaks differs from the shape of an expected peak; Including, the physiological measurements include at least one of ECG measurements or plethysmography measurements of the person; The pseudo actigraphy data is configured to be inversely correlated with the actigraphy data as follows: (a) If the signal quality is higher, it will produce a lower value; (b) producing a higher value when the signal quality is lower; (c) identifying a REM sleep epoch or a wake epoch based on the anti-correlation. characterized in that A computer-implemented method.

2. 1. A system comprising: one or more processors; at least one memory for storing instructions; the instructions, when executed by one or more processors, cause the system to: (A) accessing cardiopulmonary coupling data over a period of sleep for a person; (B) classifying a portion of the epochs of the sleep period as very low frequency coupling (vLFC) epochs based on the cardiopulmonary coupling data; (C) accessing pseudo-actigraphy data that is inversely correlated with the actigraphy data, wherein the pseudo-actigraphy data is obtained by quantifying signal quality based on physiological measurements of the person corresponding to the vLFC epochs; (D) designating the vLFC epoch as a REM epoch or a wake epoch based on the pseudo-actigraphy data corresponding to the vLFC epoch; wherein the process C of quantifying the signal quality of the physiological measurement comprises: (C1) processing the physiological measurements to detect peaks in the vLFC epochs; (C2-1) generating data indicative of higher signal quality when the count of detected peaks is below a predetermined threshold and when the shape of the detected peaks matches the shape of an expected peak; (C2-2) generating data indicating lower signal quality when the count of the detected peaks exceeds a predetermined threshold and when the shape of the detected peaks differs from the shape of an expected peak; Processing includes: the physiological measurements include at least one of ECG measurements or plethysmography measurements of the person; The pseudo actigraphy data is configured to be inversely correlated with the actigraphy data as follows: (a) If the signal quality is higher, it will produce a lower value; (b) producing a higher value when the signal quality is lower; (c) identifying a REM sleep epoch or a wake epoch based on the anti-correlation. A system characterized by:

Images