A system and method for enhancing REM sleep using sensory stimulation.

Abstract

The present disclosure relates to systems and methods for automatically detecting rapid eye movement (REM) sleep and delivering sensory stimuli to extend REM duration without disturbing sleep. The sensory stimuli may be auditory or other stimuli. The systems and methods ensure timely delivery of the stimuli and automatically adjust the amount, intensity, and / or timing of the stimuli as needed. REM sleep is detected based on brain activity, cardiac activity, and / or other information. REM sleep may be detected and / or predicted by a trained neural network. The amount, timing, and / or intensity of the sensory stimuli may be determined and / or adjusted to enhance REM sleep in a subject based on one or more values ​​in one or more hidden layers of the neural network and one or more brain activity parameters and / or cardiac activity parameters.

Description

Technical Field

[0001] This disclosure relates to systems and methods for enhancing rapid eye movement (REM) sleep by delivering stimuli to a subject during a sleep session. Hearing

Background Art

[0002] Systems that monitor sleep and deliver sensory stimuli to a subject during sleep are known. Sleep monitoring and sensory stimulation systems based on electroencephalogram (EEG) sensors are known. These systems are state-based, which means that stimuli are sent in response to EEG parameters exceeding a sleep stage stimulation delivery threshold. The sleep state stimulation delivery threshold typically indicates deep sleep (e.g., NREM stages N2 or N3 sleep). These systems do not deliver stimuli that enhance REM sleep. As a result, the benefits of enhanced REM sleep are not experienced by users of typical systems. International Publication No. 2017 / 115368 (A1) describes the neural regulatory systems for the treatment of physiological disorders. Some information regarding the enhancement of REM sleep is found in Lim et al., "Selective Enhancement of Rapid Eye Movement sleep by Deep Brain Stimulation of the Human Pons", Ann. Neurol., 2009, 66:110-114.

Summary of the Invention

Problems to be Solved by the Invention

[0003] It would be advantageous to enhance rapid eye movement (REM) sleep by automatically delivering sensory stimuli to a subject during a sleep session using a closed-loop system.

Means for Solving the Problems

[0004] Thus, one or more aspects of this disclosure relate to a system configured to enhance REM sleep by delivering stimuli to a subject during a sleep session. The system includes one or more sensors, one or more sensory stimulation devices, one or more hardware processors, and / or other components. The one or more sensors are configured to generate an output signal that conveys information about the subject's sleep stage during a sleep session. The one or more sensory stimulation devices are configured to deliver stimuli to the subject during a sleep session Hearing Hearing ​​It is configured to provide stimulation. One or more hardware processors are coupled to one or more sensors and one or more sensory stimulators. One or more hardware processors are configured to provide machine-readable instructions. One or more hardware processors are configured to detect REM sleep in the subject during a sleep session based on output signals. One or more hardware processors control one or more sensory stimulators to enhance REM sleep in the subject during a sleep session. Hearing It was configured to provide stimulation.

[0005] In some embodiments, one or more sensors are configured such that information regarding the subject's sleep stage includes information regarding brain activity and / or cardiac activity in the subject. In some embodiments, one or more sensors include one or more electroencephalogram (EEG) electrodes configured to generate information regarding brain activity, one or more electrocardiogram (ECG) sensors configured to generate information regarding cardiac activity, one or more photoplethysmography (PPG) sensors configured to generate information regarding cardiac activity, and / or other sensors.

[0006] In some embodiments, one or more sensors are configured such that information about the subject's sleep stage includes information about cardiac activity. In some embodiments, one or more hardware processors are configured to detect REM sleep in the subject in response to the ratio of the low-frequency component to the high-frequency component of the cardiac activity information exceeding a ratio threshold.

[0007] In some embodiments, one or more hardware processors are further configured to detect REM sleep in a subject in response to determining that the subject remains in REM sleep for a continuous threshold period of time during a sleep session.

[0008] In some embodiments, one or more hardware processors are configured to detect REM sleep in a subject, which includes obtaining historical sleep stage information for the subject and / or a population of subjects demographically similar to the subject. The historical sleep stage information relates to the brain and / or cardiac activity of the subject and / or the population of subjects, indicating the sleep stages over time during the subject and / or the population of subjects' sleep sessions. Detecting REM sleep includes providing the historical sleep stage information to a neural network as input, thereby training the neural network based on the historical sleep stage information, and, based on the output signal, causing the trained neural network to (1) determine the duration of REM sleep in the subject during a sleep session, or (2) predict future times during a sleep session in which the subject will experience REM sleep. The trained neural network includes an input layer, an output layer, and one or more hidden layers between the input and output layers.

[0009] In some embodiments, one or more hardware processors control one or more sensory stimulators to enhance REM sleep in a subject during a sleep session, thereby enhancing REM sleep in the subject during REM sleep. Hearing Providing stimulation is configured to include determining one or more values ​​generated by one or more hidden layers of a trained neural network for each of (1) the period during which the subject is experiencing REM sleep, or (2) future time. Controlling one or more sensory stimulators is configured to provide one or more sensory stimulators to the subject for (1) the period during which the subject is experiencing REM sleep, or (2) future time. Hearing This includes providing stimulation. Controlling one or more sensory stimulation devices is provided to the subject based on one or more values ​​in one or more intermediate layers. Hearing This includes determining the amount, timing, and / or intensity of stimulation, and / or adjusting one or more sensory stimulators.

[0010] In some embodiments, one or more hardware processors are configured to determine one or more brain activity parameters and / or cardiac activity parameters of a subject based on an output signal. One or more brain activity parameters and / or cardiac activity parameters indicate the subject's sleep stage. One or more hardware processors, based on one or more values ​​of one or more intermediate layers, and one or more brain activity parameters and / or cardiac activity parameters, determine how to enhance REM sleep in the subject. Hearing It is configured to determine the amount, timing, and / or intensity of stimulation, and / or to adjust one or more sensory stimulators.

[0011] In some embodiments, one or more hardware processors are configured such that one or more values ​​from one or more hidden layers of a trained neural network include values ​​from one or more convolutional layers and one or more recurrent layers of the trained neural network. Based on one or more brain activity parameters and / or cardiac activity parameters, values ​​from one or more convolutional layers and values ​​from one or more recurrent layers, Hearing It is configured to determine the amount, timing, and / or intensity of stimulation, and / or to adjust one or more sensory stimulators.

[0012] In some embodiments, one or more sensory stimulation devices are Hearing The stimuli are configured to include audible sounds. One or more hardware processors control one or more sensory stimuli to enhance REM sleep in the subject during a sleep session. Hearing Providing a stimulus determines the tone interval, tone volume, and / or tone frequency. Hearing Determining the amount, timing, and / or intensity of the stimulus, and / or HearingThe system is configured to include adjusting the amount, timing, and / or intensity of stimulation in one or more sensory stimulators. The adjustments include modifying tone intervals, tone volume, and / or tone frequency in response to indicators that the subject is experiencing one or more micro-arousals.

[0013] In some embodiments, the stimulus is activated at a specified time to synchronize with the detection of Ponto-geniculo-occipital (PGO) waves in EEG.

[0014] Another aspect of this disclosure involves using an augmentation system to provide subjects with an augmentation system during a sleep session. Hearing It enhances REM sleep by delivering stimuli. Non-therapeutic The method relates to the above system which includes one or more sensors, one or more sensory stimulators, one or more hardware processors, and / or other components. The method includes the steps of one or more sensors generating output signals that convey information about the subject's sleep stage during a sleep session, one or more hardware processors detecting REM sleep in the subject during a sleep session based on the output signals, and one or more hardware processors controlling one or more sensory stimulators to enhance REM sleep in the subject during a sleep session to the subject during REM sleep. Hearing The step includes providing stimulation.

[0015] In some embodiments, information regarding the subject's sleep stage includes information regarding brain activity and / or cardiac activity in the subject. In some embodiments, information regarding the subject's sleep stage includes information regarding cardiac activity, and one or more hardware processors are configured to detect REM sleep in the subject in response to the ratio of the low-frequency component to the high-frequency component of the cardiac activity information exceeding a ratio threshold.

[0016] In some embodiments, the step of detecting REM sleep in a subject includes obtaining historical sleep stage information for the subject and / or a population of subjects demographically similar to the subject. The historical sleep stage information relates to the brain activity and / or cardiac activity of the subject and / or the population of subjects, indicating the sleep stages over time during the subject and / or the population of subjects' sleep sessions. The step of detecting REM sleep in a subject includes training a neural network based on the historical sleep stage information by providing the historical sleep stage information as input to the neural network, and, based on the output signal, causing the trained neural network to (1) determine the duration of REM sleep in the subject during a sleep session, or (2) predict future times during a sleep session in which the subject will experience REM sleep. The trained neural network includes an input layer, an output layer, and one or more intermediate layers between the input and output layers. To enhance REM sleep in the subject during a sleep session, one or more sensory stimulators are controlled to enhance the subject during REM sleep. Hearing The step of providing stimulation involves determining one or more values ​​generated by one or more hidden layers of a trained neural network for each of (1) the period during which the subject is experiencing REM sleep, or (2) future time, and providing one or more sensory stimulation devices to the subject during (1) the period during which the subject is experiencing REM sleep, or (2) future time. Hearing To provide stimulation, and based on one or more values ​​in one or more intermediate levels, provided to the subject. Hearing This includes determining the amount, timing, and / or intensity of stimulation, and / or adjusting one or more sensory stimulators.

[0017] In some embodiments, the method further includes one or more hardware processors determining one or more brain activity parameters and / or one or more heart activity parameters of a subject based on an output signal. The one or more brain activity parameters and / or one or more heart activity parameters indicate the sleep stage of the subject. In some embodiments, the method includes one or more hardware processors determining, based on one or more values of one or more intermediate layers and on one or more brain activity parameters and / or one or more heart activity parameters, Hearing the amount, timing, and / or intensity of a stimulus for enhancing REM sleep in the subject and / or causing one or more sensory stimulation devices to be adjusted.

[0018] In some embodiments, the stimulus is activated at a specified time so as to be synchronized with the detection of Ponto-geniculo-occipital (PGO) waves in the EEG.

[0019] These and other objects, features, and characteristics of the present disclosure, as well as the operating methods and functions of combinations of related structural elements and parts, as well as the elimination of manufacturing waste, will become more apparent by considering the following description and the appended claims in reference to the accompanying drawings, all of which form a part of this specification, and like reference numerals indicate corresponding parts in various figures. However, it should be clearly understood that the drawings are for illustrative and explanatory purposes only and are not intended to define the scope of the present disclosure.

Brief Description of the Drawings

[0020] [Figure 1] FIG. 1 is a schematic diagram of a system configured to enhance REM sleep by delivering sensory stimuli to a subject during a sleep session, according to one or more embodiments. [Figure 2] FIG. 2 is a diagram illustrating some of the operations performed by the system, according to one or more embodiments. [Figure 3]This figure illustrates an example of a deep neural network architecture that is part of the system, according to one or more embodiments. [Figure 4] This figure illustrates threshold values ​​for the ratio of the low-frequency component to the high-frequency component of the cardiac activity signal, which can be used by a system to detect REM sleep and distinguish REM sleep from NREM sleep or wakefulness, according to one or more embodiments. [Figure 5] This figure illustrates a method of enhancing REM sleep by delivering sensory stimuli to a subject during a sleep session using an augmentation system, according to one or more embodiments. [Modes for carrying out the invention]

[0021] Where used herein, the singular form with an indefinite or definite article includes its plural form unless the context explicitly indicates otherwise. Where used herein, the term “or” means “and / or” unless the context explicitly indicates otherwise. Where used herein, the statement that two or more parts or components are “joined” means that the parts are joined together or operate together directly or indirectly, i.e., through one or more intermediate parts or components, insofar as the connection occurs. Where used herein, “directly joined” means that the two elements are in direct contact with each other. Where used herein, “fixed joined” or “fixed” means that two components are joined so as to move as one while maintaining a constant orientation relative to each other.

[0022] As used herein, the term “unitary” means that the components are manufactured as a single piece or unit. That is, components that include parts manufactured separately and then joined together as a unit are not “unitary” components or “unitary” bodies. As used herein, the statement that two or more parts or components “interlock / engage” with each other means that the parts exert force on each other directly or through one or more intermediate parts or components. As used herein, the term “number” means one or one or more integers (i.e., plural).

[0023] For example, without limitation, directional phrases such as up, down, left, right, superior, inferior, front, back, and their derivatives do not limit the scope of the claims when used herein, unless explicitly stated, with respect to the orientation of the elements shown in the drawings.

[0024] Figure 1 is a schematic diagram of a system 10 configured to deliver sensory stimuli to a subject 12 during a sleep session. System 10 is configured to facilitate the delivery of sensory stimuli to subject 12 for the purpose of enhancing the restorative effect of sleep in subject 12 and / or other purposes. System 10 is configured so that sensory stimuli, including auditory and / or other stimuli, delivered during sleep enhance rapid eye movement (REM) sleep in subject 12 without causing arousal. As described herein, in some embodiments, system 10 is configured to determine the duration of REM sleep during a sleep session (based on, for example, output from a neural network and / or other information, based on cardiac activity and / or other information). In some embodiments, based on such a determination, system 10 is configured to modulate the sensory (e.g., auditory) stimuli delivered to subject 12 to enhance REM sleep without causing arousal. In some embodiments, the duration of REM sleep may be determined in real time and / or near real time during subject 12's sleep session.

[0025] REM sleep is a crucial component of sleep quality, and enhancing REM sleep correlates with improved performance accuracy (e.g., in tasks performed by subjects after a sleep session). More REM sleep, not necessarily slow-wave sleep, is associated with a reduction in cognitive decline over time in subjects. Sensory stimuli delivered during REM sleep (e.g., auditory sounds) can extend the duration of REM sleep during a sleep session. Previous attempts to enhance REM sleep in subjects required strict supervision by sleep specialists. Sleep specialists monitored polysomnography signals from subjects during sleep to detect periods of REM sleep, and then manually delivered auditory stimuli to the subjects. In addition to the drawback of requiring sleep specialist intervention, manual delivery of sensory stimuli was not precisely or consistently controlled.

[0026] System 10 is configured to automatically detect REM sleep and deliver sensory (e.g., auditory) stimuli to the subject 12 to extend the duration of REM sleep while substantially preventing sleep disturbances. System 10 ensures timely delivery of stimuli and automatically adjusts the amount and / or other characteristics of the stimuli (e.g., intensity, timing, etc.). System 10 addresses the limitations of prior art systems (e.g., requiring manual input from a sleep professional, inaccurate or inconsistent delivery of sensory stimuli by a sleep professional, etc.) by leveraging machine learning models (e.g., deep neural networks and / or any other supervised machine learning algorithms described below) for automatic real-time or near real-time closed-loop augmentation of REM sleep based on sensory output signals by delivering sensory stimuli to the subject 12 during a sleep session. System 10 uses the entire output from the machine learning model for sleep staging, as well as intermediate values ​​output from the model to adjust the sensory stimuli (e.g., the amount) provided by System 10. In some embodiments, System 10 does not use a machine learning model and instead detects REM sleep based on cardiac activity thresholds and / or other thresholds for the subject 12. It should be noted that System 10 is described herein as being configured to enhance REM sleep (for example, to increase the proportion of REM sleep). This may include, for example, characterization of the duration and / or proportion of eye movements (or phasic activity) during REM, and / or other characterizations. In some embodiments, System 10 includes one or more of the following components: a sensor 14, a sensory stimulator 16, an external resource 18, a processor 20, an electronic memory device 22, a user interface 24, and / or other components.

[0027] Sensor 14 is configured to generate an output signal that transmits information about the sleep stages of subject 12 during a sleep session. The output signal that transmits information about the sleep stages of subject 12 may include information about brain activity in subject 12, cardiac activity in subject 12, and / or other physiological activity in subject 12. As such, sensor 14 is configured to generate an output signal that transmits information about brain activity, cardiac activity, and / or other activity in subject 12. In some embodiments, sensor 14 is configured to generate an output signal that transmits information about stimuli to be provided to subject 12 during a sleep session. In some embodiments, (as described below) the information in the output signal from sensor 14 is used to control the sensory stimulator 16 to provide sensory stimuli to subject 12.

[0028] Sensor 14 may include one or more sensors that generate output signals that directly transmit information about brain activity in subject 12. For example, sensor 14 may include electroencephalography (EEG) electrodes configured to detect electrical activity along subject 12's scalp resulting from electrical currents in subject 12's brain. Sensor 14 may also include one or more sensors that generate output signals that indirectly transmit information about subject 12's brain activity. For example, one or more sensors 14 may include a heart rate sensor that generates an output based on the subject's heart rate (for example, sensor 14 may be a heart rate sensor that can be positioned on the subject's chest and / or may be configured as a bracelet on the subject's wrist and / or may be positioned on another limb of the subject's arm), the subject's movement (for example, sensor 14 may include an accelerometer that can be carried in a wearable, such as a bracelet around the subject's wrist and / or ankle, so that sleep can be analyzed using actigraphy signals), the subject's respiration, and / or other characteristics of the subject.

[0029] In some embodiments, sensor 14 may include one or more other sensors configured to generate output signals relating to stimuli provided to subject 12 (e.g., the amount, frequency, intensity, and / or other characteristics of the stimuli), relating to the subject 12's brain activity, relating to the subject 12's cardiac activity, and / or other sensors, including EEG electrodes, electrooculography (EOG) electrodes, actigraphy sensors, electrocardiogram (EKG) electrodes, respiratory sensors, pressure sensors, vital signs cameras, photoplethysmogram (PPG) sensors, functional near-infrared (fNIR) sensors, temperature sensors, microphones, and / or other sensors. Although sensor 14 is exemplified in a single location close to subject 12, this is not intended to be limiting. The sensor 14 may include sensors positioned in multiple locations, such as within (or in a position communicating with) the sensory stimulation device 16, in a position attached to the subject 12's clothing (in a removable manner), in a position worn by the subject 12 (e.g., as a headband, wristband, etc.), in a position positioned to point at the subject 12 while the subject 12 is sleeping (e.g., as a camera that transmits output signals regarding the subject 12's movements), in a position attached to the bed and / or other furniture on which the subject 12 is sleeping, and / or other positions.

[0030] In Figure 1, the sensor 14, sensory stimulator 16, processor 20, electronic memory device 22, and user interface 24 are shown as separate entities. This is not intended to be limiting. Some and / or all of the components of system 10 and / or other components can be grouped into one or more singular devices. For example, these and / or other components may be included in a headset and / or other clothing worn by the subject 12. Such a headset may include, for example, a detection electrode, a reference electrode, one or more devices associated with EEG, means for delivering auditory stimuli (e.g., wired and / or wireless audio devices and / or other devices), and one or more audio speakers. In this example, the audio speakers may be located inside and / or near the ear of the subject 12 and / or elsewhere. The reference electrode may be located behind and / or elsewhere the user's ear. In this example, the detection electrode may be configured to generate an output signal that transmits information and / or other information regarding the brain activity of the subject 12. The output signal may be sent wirelessly and / or via wire to a processor (e.g., processor 20 shown in Figure 1), a computing device which may or may not include a processor (e.g., a bedside laptop), and / or other devices. In this example, acoustic stimuli may be delivered to the subject 12 via a wireless audio device and / or speaker. In this example, the detection electrode, reference electrode, and EEG device can be represented, for example, by the sensor 14 in Figure 1. The wireless audio device and speaker can be represented, for example, by the sensory stimulation device 16 shown in Figure 1. In this example, the computing device may include the processor 20, electronic memory 22, user interface 24, and / or other components of the system 10 shown in Figure 1.

[0031] The stimulator 16 is configured to provide sensory stimulation to the subject 12. The sensory stimulator 16 provides auditory stimulation prior to the sleep session, during the sleep session, and / or at other times. (being billed) , visual (Not billed)Somatosensory perception (Not billed) ,electricity (Not billed) magnetic (Not billed) , and / or sensation (Not billed) The sensory stimulation device 16 is configured to provide stimuli to subject 12. In some embodiments, a sleep session may include any period during which subject 12 is asleep and / or attempting to sleep. A sleep session may include nighttime sleep, naps, and / or other sleep sessions. For example, the sensory stimulation device 16 may be configured to provide stimuli to subject 12 during a sleep session in order to enhance REM sleep in subject 12 and / or for other purposes.

[0032] The sensory stimulation device 16 is configured to enhance REM sleep in the subject 12 through non-invasive brain stimulation and / or other methods. The sensory stimulation device 16 is auditory (being billed) ,electricity (Not billed) magnetic (Not billed) , visual (Not billed) Somatosensory perception (Not billed) , and / or other senses (Not billed)REM sleep may be enhanced through non-invasive brain stimulation using the following stimuli. Auditory, electrical, magnetic, visual, somatosensory, and / or other sensory stimuli may include auditory stimuli, visual stimuli, somatosensory stimuli, electrical stimuli, magnetic stimuli, combinations of different types of stimuli, and / or other stimuli. Auditory, electrical, magnetic, visual, somatosensory, and / or other sensory stimuli include olfactory, auditory, visual stimuli, tactile, gustatory, somatosensory stimuli, electrical, magnetic, and / or other stimuli. Sensory stimuli may have intensity, timing, and / or other characteristics. For example, REM sleep in subject 12 can be enhanced by providing acoustic tones to subject 12. Acoustic tones may include one or more series of tones of a determined length, spaced apart from each other by tone intervals. The volume (e.g., intensity, etc.) of individual tones can be adjusted based on various factors (as described herein). The length (e.g., timing, etc.) and / or tone intervals (e.g., timing, etc.) of individual tones can also be adjusted. Pitch and tone can also be adjusted. In some embodiments, this illustrative auditory stimulus is in the form of a pink noise tone with a length of 40 milliseconds (the frequency limit of pink noise is 500 Hz to 5 kHz). The tone interval may be 10 seconds, and the volume of the stimulus may be 80 dB. This example is not intended to be limiting. Examples of the sensory stimulator 16 may include one or more of the following: a sound source, a speaker, a music player, a tone generator, a vibrator (such as a piezoelectric element) that delivers vibrational stimuli, a coil that generates a magnetic field that directly stimulates the cerebral cortex, one or more photogenerators or lamps, a fragrance dispenser, and / or other devices. In some embodiments, the sensory stimulator 16 is configured to adjust the intensity, timing, and / or other parameters of the stimulus provided to the subject 12 (as described below, for example).

[0033] External resource 18 includes information sources (e.g., databases, websites, etc.), external entities involved in system 10 (e.g., one or more external sleep monitoring devices, a healthcare provider's medical record system, etc.), and / or other resources. For example, external resource 18 may include historical sleep stage information sources for subject 12, historical sleep stage information sources for subject populations demographically similar to subject 12, and / or other information sources. Historical sleep stage information for subject 12 and / or subject populations may also relate to brain activity and / or cardiac activity of subject 12 and / or subject populations, indicating sleep stages over time during sleep sessions of subject 12 and / or subject populations. In some embodiments, historical sleep stage information for subject 12 and / or subject populations may also relate to user populations in a given geographical area; demographic information relating to sex, ethnicity, age, general health level, and / or other demographic information; physiological information about subject populations (e.g., weight, blood pressure, pulse rate, etc.), and / or other information. In some embodiments, this information may indicate whether individual users in the target population are similar to subject 12 in demographic, physiological, and / or otherwise.

[0034] In some embodiments, the external resource 18 includes components that facilitate the communication of information, one or more servers outside the system 10, a network (e.g., the Internet), an electronic storage device, equipment relating to Wi-Fi technology, equipment relating to Bluetooth® technology, a data input device, a sensor, a scanner, a computing device related to an individual user, and / or other resources. In some implementations, some or all of the functions attributed to the external resource 18 in this specification may be provided by resources included in the system 10. The external resource 18 can be configured to communicate with the processor 20, the user interface 24, the sensor 14, the electronic storage device 22, the sensory stimulator 16, and / or other components of the system 10 via wired and / or wireless connections, via a network (e.g., a local area network and / or the Internet), via cellular technology, via Wi-Fi technology, and / or other resources.

[0035] The processor 20 is configured to provide information processing capabilities in the system 10. As such, the processor 20 may include one or more of the following: a digital processor, an analog processor, a digital circuit designed to process information, an analog circuit designed to process information, a state machine, and / or other mechanisms for electronically processing information. Although the processor 20 is shown as a single entity in Figure 1, this is for illustrative purposes only. In some embodiments, the processor 20 may include multiple processing units. These processing units may be physically located within the same device (e.g., sensory stimulation device 16, user interface 24), or the processor 20 may represent the processing capabilities of multiple devices working together. In some embodiments, the processor 20 may be a computing device such as a desktop computer, laptop computer, smartphone, tablet computer, server, and / or other computing devices, and / or may be included therein. Such computing devices can run one or more electronic applications having a graphical user interface configured to facilitate user interaction with the system 10.

[0036] As shown in Figure 1, the processor 20 is configured to execute one or more computer program components. The computer program components may include, for example, software programs and / or algorithms coded and / or otherwise embedded in the processor 20. The one or more computer program components may include one or more of the information component 30, the model component 32, the control component 34, the adjustment component 36, and / or other components. The processor 20 may be configured to execute components 30, 32, 34, and / or 36 by software; hardware; firmware; a combination of some software, hardware, and / or firmware; and / or other mechanisms for configuring the processing capacity on the processor 20.

[0037] Although components 30, 32, 34, and 36 are illustrated in Figure 1 as being jointly installed within a single processing unit, it should be correctly understood that in embodiments in which the processor 20 includes multiple processing units, one or more of components 30, 32, 34, and / or 36 may be located separately from the other components. The descriptions of the functions provided by the different components 30, 32, 34, and / or 36 provided below are for illustrative purposes only and are not intended to be limiting, as any of the components 30, 32, 34, and / or 36 may provide more or less functionality than described. For example, one or more of components 30, 32, 34, and / or 36 may be removed, and some or all of their functions may be provided by the other components 30, 32, 34, and / or 36. As another example, the processor 20 may be configured to perform one or more further components that can perform some or all of the functions attributed to one of the components 30, 32, 34, and / or 36.

[0038] The information component 30 is configured to determine one or more brain activity parameters and / or cardiac activity parameters and / or other information of the subject 12. The brain activity parameters and / or cardiac activity parameters are determined based on the output signals from the sensor 14 and / or other information. The brain activity parameters and / or cardiac activity parameters indicate the depth of sleep in the subject 12. In some embodiments, the information in the output signals regarding brain activity and / or cardiac activity indicates the depth of sleep over time. In some embodiments, the information indicating the depth of sleep over time is or includes information regarding REM sleep in the subject 12.

[0039] In some embodiments, information indicating sleep depth over time may indicate other sleep stages of subject 12. For example, the sleep stages of subject 12 may relate to REM sleep, non-rapid eye movement (NREM) sleep, and / or other sleep. NREM sleep may be one or more of NREM stages N1, N2, or N3, and / or other sleep stages. In some embodiments, the sleep stages of subject 12 may be one or more of stages S1, S2, S3, or S4. In some embodiments, NREM stages 2 and / or 3 (and / or S3 and / or S4) may be slow-wave (e.g., deep) sleep. In some embodiments, information indicating sleep depth over time may be one or more further brain activity parameters and / or cardiac activity parameters and / or relating to them.

[0040] In some embodiments, information regarding brain activity and / or cardiac activity indicating sleep depth over time is and / or includes EEG information, ECG information, PPG information, and / or other information generated during and / or at other times of the subject's sleep sessions. In some embodiments, brain activity parameters and / or cardiac activity parameters may be determined based on EEG information, ECG information, PPG information, and / or other information. In some embodiments, brain activity parameters and / or cardiac activity parameters may be determined by information component 30 and / or other components of system 10. In some embodiments, brain activity parameters and / or cardiac activity parameters may be predetermined and may be part of historical sleep stage information obtained from an external resource 18 (described below). In some embodiments, one or more brain activity parameters are and / or relate to the presence of a specific sleep pattern, such as frequency, amplitude, phase, eye movements, ponto-geniculo-occipital (PGO) waves, slow waves, and / or other features of the EEG signal. In some embodiments, one or more brain activity parameters are determined based on the frequency, amplitude, and / or other features of the EEG signal. In some embodiments, one or more cardiac activity parameters are and / or relating to frequency, amplitude, phase, heart rate, heart rate variability, pulse interval, power in the low-frequency band (0.05–0.15 Hz), power in the high-frequency band (0.15–0.4 Hz), and / or other features of the ECG and / or PPG signals. In some embodiments, one or more brain activity parameters and / or cardiac activity parameters are determined based on the frequency, amplitude, and / or other features of the EEG, ECG, and / or PPG signals. In some embodiments, the determined brain activity parameters and / or cardiac activity parameters and / or EEG, ECG, and / or PPG features may and / or represent sleep stages corresponding to the REM and / or NREM sleep stages described above.

[0041] The information component 30 is configured to obtain historical sleep stage information. In some embodiments, the historical sleep stage information is for subject 12, a population of subjects demographically similar to subject 12, and / or other users. In some embodiments, the population of subjects is demographically similar to subject 12. In some embodiments, the historical sleep stage information is for subject 12. The historical sleep stage information relates to brain activity, cardiac activity, and / or other physiological information of the population of subjects and / or subject 12, indicating sleep stages over time between previous sleep sessions of the population of subjects and / or subject 12. The historical sleep stage information relates to sleep stages, and / or other brain activity parameters and / or cardiac activity parameters and / or other information of the population of subjects and / or subject 12 between corresponding sleep sessions.

[0042] In some embodiments, the information component 30 is configured to electronically obtain historical sleep stage information from external resources 18, electronic storage devices 22, and / or other sources. In some embodiments, obtaining historical sleep stage information electronically from external resources 18, electronic storage devices 22, and / or other sources includes searching one or more databases and / or servers, uploading and / or downloading information, facilitating user input (e.g., criteria used to define input for a target patient population via a user interface 24), sending and receiving emails, sending and receiving text messages, and / or sending and receiving other communications, and / or performing other retrieval operations. In some embodiments, the information component 30 is configured to aggregate information from various sources (e.g., one or more of the external resources 18 and electronic storage devices 22 mentioned above), place the information in one or more electronic databases (e.g., electronic storage devices 22 and / or other electronic databases), normalize the information based on one or more characteristics of the historical sleep stage information (e.g., length of sleep sessions, number of sleep sessions, etc.), and / or perform other operations.

[0043] Model component 32 is configured to train a machine learning model using historical sleep stage information. In some embodiments, the machine learning model is trained on historical sleep stage information by providing the historical sleep stage information as input to the machine learning model. In some embodiments, the machine learning model may be and / or include mathematical equations, algorithms, plots, charts, networks (e.g., neural networks, etc.), and / or other tools and machine learning model components. For example, the machine learning model may be and / or include one or more neural networks having an input layer, an output layer, and one or more hidden or intermediate layers, and / or any other supervised machine learning algorithms. In some embodiments, the one or more neural networks and / or any other supervised machine learning algorithms may be and / or include deep neural networks (e.g., neural networks having one or more hidden or intermediate layers between the input and output layers).

[0044] As an example, a neural network may be based on a collection of numerous neural units (or artificial neurons). A neural network can roughly mimic the way a biological brain functions (for example, through large clusters of biological neurons connected by axons). Each neural unit in a neural network may be connected to many other neural units in the neural network. Such connections can influence or suppress the activation state of the connected neural units. In some embodiments, each individual neural unit may have a summation function that combines all of its input values ​​together. In some embodiments, each connection (or the neural unit itself) may have a threshold function such that a threshold must be exceeded before a signal can be propagated to other neural units. These neural network systems may be trained by self-learning rather than being explicitly programmed, and can perform considerably better than conventional computer programs in solving problems in specific domains. In some embodiments, a neural network may include multiple layers (for example, if the signal path traverses from a forward layer to a backward layer). In some embodiments, backpropagation techniques may be utilized by the neural network, with forward stimuli being used to reset weights for "front" neural units. In some embodiments, stimuli and inhibitors to the neural network may flow more freely, with connections interacting in a more disordered and complex manner.

[0045] As described above, the trained neural network may include one or more hidden or intermediate layers. The intermediate layers of the trained neural network include one or more convolutional layers, one or more recurrent layers, and / or other layers of the trained neural network. Each intermediate layer receives information from another layer as input and produces a corresponding output. The detected and / or predicted sleep stage and / or future time of REM sleep are generated based on the information in the output signal from sensor 14 processed by the layers of the neural network.

[0046] Model component 32 is configured to cause a trained neural network and / or any other supervised machine learning algorithm to detect and / or predict REM sleep in subject 12. In some embodiments, this may include (1) determining the duration of REM sleep that subject 12 experiences during a sleep session, (2) predicting future times during a sleep session in which subject 12 will experience REM sleep, and / or other actions. In some embodiments, this includes causing a trained neural network and / or any other supervised machine learning algorithm to predict future times during a sleep session in which subject 12 is in REM sleep. The determined and / or predicted REM sleep and / or timing indicates whether subject 12 is in (or will be in) REM sleep and / or other information due to a stimulus. As a non-limiting example, the trained neural network may, based on the output signal and / or other information (e.g., using information in the output signal as input to the model), indicate predicted sleep stages for the user and / or future times and / or timing of deep sleep stages. The trained neural network is configured to indicate the sleep stages that are predicted to occur in the future for subject 12 during a sleep session. In some embodiments, the model component 32 is configured to provide the neural network with information in the output signal in a temporal set corresponding to individual periods during the sleep session. In some embodiments, the model component 32 is configured to cause the trained neural network to output, based on the information in the temporal set, the determined and / or predicted sleep stages and / or predicted REM sleep duration for subject 12 during the sleep session. (The function of the model component 32 will be discussed further below in relation to Figures 2-3).

[0047] The control component 34 is configured to control the stimulator 16 to provide stimulation to the subject 12 during sleep and / or at other times. The control component 34 is configured to cause the sensory stimulator 16 to provide sensory stimulation to the subject 12 during REM sleep in order to enhance REM sleep in the subject 12 during a sleep session. The control component 34 is configured to cause the sensory stimulator 16 to provide sensory stimulation to the subject 12 based on detected REM sleep and / or predicted REM sleep stages (e.g., output from model component 32) and / or future times when the subject 12 will be in REM sleep and / or other information. The control component 34 is configured to cause the sensory stimulator 16 to provide sensory stimulation to the subject 12 based on temporal detection and / or predicted REM sleep stages and / or future times when the subject 12 will be in REM sleep and / or other information during a sleep session. The control component 34 is configured to cause the sensory stimulator 16 to provide sensory stimulation to the subject 12 in response to the subject 12 being in or presumed to be in REM sleep for stimulation. For example, the control component 34 is configured to control one or more sensory stimulators 16 to provide sensory stimulation to the subject 12 during REM sleep in order to enhance REM sleep in the subject 12 during a sleep session, which includes: determining one or more values ​​generated by one or more hidden layers of a trained neural network for each of (1) the period during which the subject 12 is experiencing REM sleep, or (2) future time; causing one or more sensory stimulators 16 to provide sensory stimulation to the subject 12 during the period during which the subject 12 is experiencing REM sleep, or (2) future time; and determining and / or causing one or more sensory stimulators 16 to adjust (as described herein, for example) the amount, timing, and / or intensity of the sensory stimulation provided to the subject 12 based on one or more values ​​from one or more hidden layers. In some embodiments, the stimulator 16 is delivered during REM sleep (as described herein) (e.g., peripheral hearing) (being billed) magnetic (Not billed) ,electricity (Not billed) , and / or other (Not billed) ) are controlled by control component 34 to enhance REM sleep through stimulation.

[0048] In some embodiments, the control component 34 is configured to control the sensory stimulator 16 to deliver sensory stimuli to subject 12 in response to the model component 32 determining that subject 12 remains in REM sleep for a continuous threshold time during a sleep session. For example, the model component 32 and / or the control component 34 may be configured so that, upon detection (or prediction) of REM sleep, the model component 32 starts a (physical or virtual) timer configured to track the time subject 12 spends in REM sleep. The control component 34 is configured to deliver auditory stimuli in response to the duration of continuous REM sleep that subject 12 spends beyond a predetermined duration threshold. In some embodiments, a predetermined duration threshold is determined at the time of manufacturing of the system 10 and / or at other times. In some embodiments, the predetermined duration threshold is determined based on information from previous sleep sessions of subject 12 and / or subjects demographically similar to subject 12 (for example, as described above). In some embodiments, the predetermined duration threshold is adjustable via the user interface 24 and / or other adjustment mechanisms.

[0049] In some embodiments, a predetermined REM sleep duration threshold may be, for example, 1 minute and / or other durations. As a non-limiting example, the control component 34 may be configured to initiate an auditory stimulus once every minute that continuous REM sleep is detected (and / or predicted) in the subject 12. The auditory stimulus may be in the form of a 40-millisecond long pink noise tone (500 Hz to 5 kHz) with a 10-second tone interval and a volume of 80 dB. In some embodiments, the control component 34 is configured to control the sensory stimulator 16 to keep the parameters of these stimuli constant as long as a minor awakening of non-tone-related sleep is detected. However, upon detection of a sleep stage transition (e.g., from REM to another sleep stage), the control component 34 is configured to stop the stimulation.

[0050] The modulatory component 36 is configured to cause the sensory stimulator 16 to adjust the amount, timing, and / or intensity of sensory stimuli. The modulatory component 36 is configured to cause the sensory stimulator 16 to adjust the amount, timing, and / or intensity of sensory stimuli based on brain activity parameters, cardiac activity parameters, values ​​output from the intermediate layers of the trained neural network, and / or other information. For example, the sensory stimulator 16 is configured to adjust the timing and / or intensity of sensory stimuli based on brain activity parameters, cardiac activity parameters, values ​​output from the convolutional layer, values ​​output from the recursive layer, and / or other information. For example, the modulatory component 36 can be configured so that sensory stimuli are delivered at an intensity proportional to the predicted probability value (e.g., output from the intermediate layers of the neural network) of a particular sleep stage (e.g., REM). In this example, the higher the probability of REM sleep, the more likely the stimulation is to continue. If a sleep micro-awakening is detected and the sleep stage remains REM, the modulatory component 36 can be configured so that the volume decreases (e.g., by only 5 dB) in response to the detection of individual micro-awakenings.

[0051] As a non-limiting example, Figure 2 illustrates some of the operations performed by the system 10 as described above. In the example shown in Figure 2, the EEG signal or ECG signal 200 is processed and / or otherwise provided to the deep neural network 204 in a time window 201 (ECG) or 202 (EEG) (e.g., by the information component 30 and model component 32 shown in Figure 1). Based on the information in the time window 201 or 202, the deep neural network 204 detects and / or predicts future sleep stages 208 and / or times when the user is in REM sleep (206). In some embodiments, the prediction window is, for example, about tens of seconds to several minutes. Predicting future sleep stages and / or the timing of REM sleep facilitates the provision of sensory stimuli to enhance REM sleep, as it allows system 10 to refrain from stimuli (when deeper sleep stages are detected and / or predicted) or to prepare stimuli with optimized timing and intensity when REM sleep is detected and / or predicted. The architecture of the deep neural network 204 includes convolutional layers 210 (which can be thought of as filters) and recurrent layers 212 that provide memory to the network 204 (which can be implemented as long- and short-term memory elements, just as an example) and can use past predictions to refine prediction accuracy.

[0052] As shown in Figure 2, in response to the determination and / or prediction 208 of a sleep stage indicating REM sleep 214, a stimulus 216 is provided to the subject 12 (from a sensory stimulator 16 controlled by, for example, a control component 34 shown in Figure 1). In some embodiments, the system 10 is configured to detect REM sleep in the subject 12 in response to the determination 215 that the subject 12 remains in REM sleep over a continuous threshold time during a sleep session. The intensity and / or timing of the stimulus 216 is determined by the extracted signal features 200 (for example, the information component 30 shown in Figure 1), brain activity parameters and / or cardiac activity parameters 220 (constants C1, C2, ..., C n The output 222 from the convolutional layer of the deep neural network (as illustrated), and 218 which is modulated (e.g., by the modulation module 36) based on the detected and / or predicted sleep stage 208. As described above, in some embodiments, the sensory stimulus includes an audible sound. In these embodiments, the sensory stimulator 16 can modulate the amount, timing, and / or intensity of the sensory stimulus in response to brain activity parameters and / or cardiac activity parameters and / or outputs from the intermediate layer (e.g., the convolutional layer 210 and / or recurrent layer 212) indicating that the subject 12 is in and / or predicted to be in REM sleep for the stimulus.

[0053] Figure 3 illustrates an example architecture 300, which is part of a deep neural network (e.g., the deep neural network 204 shown in Figure 2) that is part of system 10 (Figures 1 and 2). Figure 3 illustrates architecture 300 of the deep neural network for windows 302, 304, and 306 of three (expanded) EEG 301. Architecture 300 includes convolutional layers 308, 310, and 312 and recurrent layers 320, 322, and 324. As described above, the convolutional layers 308, 310, and 312 can be thought of as filters, generating convolutional outputs 314, 316, and 318 that are fed to the recurrent layers 320, 322, and 324 (LSTM (Long Short-Term Memory) layers in this example). The output of architecture 300 for each window 302, 304, 306 being processed is a set of predicted probabilities for each sleep stage, called “one or more soft output(s)” 326. The “hard” prediction 328 is determined by architecture 300 (the model component 32 shown in Figure 1) by predicting the sleep stage associated with the “soft” output having the highest value 330 (as described below, for example). The terms “soft” and “hard” are not intended to be restrictive, but may be useful in describing the actions performed by the system. For example, the term “soft output” may be used because any decision is possible at this stage. In fact, the final decision may depend on post-processing of the soft output, for example. In Figure 3, “Argmax” is the operator that indicates the sleep stage associated with the highest “soft output” (e.g., the highest probability, etc.).

[0054] For example, a useful property of neural networks is that they can generate probabilities associated with predefined sleep stages (e.g., wakefulness, REM, N1, N2, N3 sleep, etc.). Model component 32 (Figure 1) is configured such that a set of probabilities constitutes a so-called soft decision vector, which can be translated into a hard decision by determining which sleep stage is associated with the highest probability value (out of a set of possible values) compared to other sleep stages. These soft decisions allow system 10 to consider different possible sleep states on a continuum, rather than being forced to determine which isolated sleep stage "bucket" a particular piece of EEG information fits into (as in systems of prior art).

[0055] Returning to Figure 1, the model component 32 is configured such that both the values ​​output from the convolutional layer and the soft-decision value outputs are vectors containing continuous values, in contrast to discrete values ​​such as sleep stages. Thus, the convolutional and recurrent (soft-decision) value outputs are available for use by the system 10 to adjust the volume of stimulation when the deep neural network detects and / or predicts the onset of REM sleep. In addition, as described herein, the stimulation settings can be adjusted using parameters determined based on raw sensor output signals (e.g., EEG, ECG, etc.) (e.g., by the information component 30 shown in Figure 1).

[0056] As described above, the modulatory component 36 is configured to cause the sensory stimulator 16 to adjust the amount, timing, and / or intensity of sensory stimuli. The modulatory component 36 is configured to cause the sensory stimulator to adjust the amount, timing, and / or intensity of sensory stimuli based on one or more brain activity parameters, cardiac activity parameters, and / or other parameters, values ​​output from the convolutional and / or recursive layers of a trained neural network, and / or other information. As an example, the tone interval of auditory stimuli provided to subject 12 can be adjusted and / or controlled (e.g., modulated) based on value outputs from a deep neural network such as convolutional and recursive layer values ​​(e.g., sleep stage (soft) prediction probability). In some embodiments, the modulatory component 36 is configured to cause one or more sensory stimulators 16 to adjust the amount, timing, and / or intensity of sensory stimuli, and the adjustment includes adjusting the tone interval, tone volume, and / or tone frequency in response to an indicator that subject 12 is experiencing one or more micro-arousals.

[0057] In some embodiments, the modulating component 36 is configured to modulate sensory stimuli based solely on brain activity parameters, cardiac activity parameters, and / or other parameters that can be determined based on the output signal from the sensor 14 (e.g., based on raw EEG signals, ECG signals, etc.). In these embodiments, the output of the deep neural network (and / or other machine learning model) continues to be used to predict sleep stages (e.g., as described above). However, the intensity and timing of the stimuli are instead modulated based on brain activity parameters, cardiac activity parameters, and / or other parameters or characteristics determined based on the sensor output signal. In some embodiments, information in or determined based on the sensor output signal can also be combined with intermediate outputs of the network, such as the output of the convolutional layer or the final output (soft stage), to modulate the intensity and timing (e.g., as described herein).

[0058] In some embodiments, the system 10 does not use a machine learning model; instead, a model component 32 detects REM sleep based on cardiac activity thresholds and / or other thresholds for the subject 12. In some embodiments, the model component 32 is configured to detect REM sleep in the subject 12 in response to the ratio of a low-frequency component of cardiac activity information (e.g., a first cardiac activity parameter) to a high-frequency component of cardiac activity information (e.g., a second cardiac activity parameter) exceeding a ratio threshold. In such embodiments, the sensor 14 may include PPG and ECG sensors embedded in a sleep wearable device configured to generate an output signal that transmits information used to automatically detect REM sleep by taking advantage of the fact that the spectral characteristics of heart rate intervals are distinctly different in REM sleep compared to NREM sleep and wakefulness. For example, r1, r2, ..., rQ may be configured to represent a sequence of heart rate interval durations in seconds. The Fourier transform of this sequence may be used by model component 32 to determine spectral heart rate variability as the ratio of power in the low-frequency (LF) band (e.g., 0.05 to 0.15 Hz) to power in the high-frequency (HF) band (e.g., 0.15 to 0.4 Hz) of the time series, defined by r1, r2, ..., rQ generated based on the output signal.

[0059] The ratios (LF / HF) 400, 402, 404, and 406 for NREM sleep stages 408 and 410, wakefulness 412, and REM 414 are shown in Figure 4. As shown in Figure 4, a threshold 416 (e.g., about 4) for the LF / HF ratio may be used by the model component 32 (Figure 1) to detect REM sleep and distinguish REM sleep from NREM sleep or wakefulness. In some embodiments, the model component 32 is configured to determine the LF / HF ratio in real time or near real time. It should be noted that heart rate variability can also be quantified as the standard deviation of the RR interval, i.e., the square root of the variance of r1,...,rQ. For example, <r>Let this be the mean value of the sequences r1,...,rQ. <r 2 > to sequence r1 2 ,...,rQ 2 Let's use the mean. Therefore, the standard deviation is: σ=sqrt( <r> × <r>- <r 2 >)

[0060] Returning to Figure 1, the electronic storage device 22 includes an electronic storage medium for electronically storing information. The electronic storage medium of the electronic storage device 22 may include either or both system storage provided integrally with the system 10 (i.e., substantially inremovable) and / or removable storage that can be detachably connected to the system 10 via, for example, a port (e.g., a USB port, a FireWire port, etc.) or a drive (e.g., a disk drive, etc.). The electronic storage device 22 may include one or more of the following: optically readable storage medium (e.g., optical disc, etc.), magnetically readable storage medium (e.g., magnetic tape, magnetic hard drive, floppy drive, etc.), charge-based storage medium (e.g., EPROM, RAM, etc.), solid-state storage medium (e.g., flash drive, etc.), cloud storage, and / or other electronically readable storage mediums. The electronic storage device 22 can store software algorithms, information determined by the processor 20, information received via the user interface 24 and / or an external computing system (e.g., external resource 18, etc.), and / or other information that enables the system 10 to function as described herein. The electronic storage device 22 may be (whole or in part) a separate component within the system 10, or the electronic storage device 22 may be (whole or in part) provided integrally with one or more other components of the system 10 (e.g., the processor 20, etc.).

[0061] The user interface 24 is configured to provide an interface between the system 10 and the subject 12 and / or other users, through which the subject 12 and / or other users can provide information to and receive information from the system 10. This allows data, signals, results, and / or commands, as well as any other communicable items, collectively referred to as “information,” to be communicated between the user (e.g., subject 12) and one or more of the sensors 14, sensory stimulators 16, external resources 18, processor 20, and / or other components of the system 10. For example, hypnograms, EEG data, REM sleep stage probabilities, and / or other information may be displayed to the subject 12 or other users via the user interface 24. As another example, the user interface 24 may be and / or be included in a computing device such as a desktop computer, laptop computer, smartphone, tablet computer, and / or other computing device. Such a computing device can run one or more electronic applications having a graphical user interface configured to provide information to and / or receive information from a user.

[0062] Examples of interface devices suitable for inclusion in the user interface 24 include keypads, buttons, switches, keyboards, knobs, levers, display screens, touchscreens, speakers, microphones, indicator lights, audible alarms, printers, haptic feedback devices, and / or other interface devices. In some embodiments, the user interface 24 includes a plurality of separate interfaces. In some embodiments, the user interface 24 includes at least one interface provided integrally with the processor 20 and / or other components of the system 10. In some embodiments, the user interface 24 is configured to communicate wirelessly with the processor 20 and / or other components of the system 10.

[0063] It should be understood that any other communication technology, whether hardwired or wireless, is also considered in this disclosure as the user interface 24. For example, this disclosure considers that the user interface 24 can be integrated with a removable storage interface provided by the electronic storage device 22. In this example, information can be loaded into the system 10 from removable storage (e.g., smart cards, flash drives, removable disks, etc.), allowing one or more users to customize the implementation of the system 10. Other illustrative input devices and technologies adapted for use with the system 10 as the user interface 24 include, but are not limited to, RS-232 ports, RF links, IR links, and modems (telephone, cable, or others). In summary, any technology for communicating information with the system 10 is considered in this disclosure as the user interface 24.

[0064] Figure 5 illustrates a method 500 for augmenting REM sleep by delivering sensory stimuli to a subject during a sleep session using an augmentation system. This system includes one or more sensors, one or more sensory stimulators, one or more hardware processors comprising machine-readable instructions, and / or other components. One or more hardware processors are configured to execute computer program components. The computer program components include information components, model components, control components, modulatory components, and / or other components. The operation of method 500 shown below is intended to be illustrative. In some embodiments, method 500 can be achieved with one or more further operations not described and / or without one or more of the operations discussed. In addition, the order in which the operations of method 500 are illustrated in Figure 5 and described below is not intended to be limiting.

[0065] In some embodiments, Method 500 may be implemented in one or more processing units, such as one or more processors 20 as described herein (e.g., digital processors, analog processors, digital circuits designed to process information, analog circuits designed to process information, state machines, and / or other mechanisms for electronically processing information). One or more processing units may include one or more devices that perform some or all of the operations of Method 500 in response to instructions electronically stored on an electronic storage medium. One or more processing units may include one or more devices configured via hardware, firmware, and / or software specifically designed for performing one or more of the operations of Method 500.

[0066] Operation 502 generates an output signal that conveys information about the subject's sleep stage during a sleep session. The output signal is generated during the user's sleep session and / or at other times. In some embodiments, operation 502 is performed by the same or similar sensor as sensor 14 (shown in Figure 1 and described herein).

[0067] In operation 504, one or more brain activity parameters and / or cardiac activity parameters are determined. Information regarding the subject's sleep stage includes information regarding brain activity and / or cardiac activity in the subject. The brain activity parameters and / or cardiac activity parameters are determined based on the output signal and / or other information. The brain activity parameters and / or cardiac activity parameters indicate the depth of sleep in the user. In some embodiments, operation 504 is performed by the same or similar processor component as the information component 30 (shown in Figure 1 and described herein).

[0068] Operation 506 provides historical sleep stage information. This historical sleep stage information is for the subject and / or a population of subjects demographically similar to the subject. The historical sleep stage information relates to the brain and / or cardiac activity of the subject and / or the population, indicating the sleep stages over time during the subject's and / or the population's sleep sessions. In some embodiments, operation 508 is performed by a processor component that is the same as or similar to the information component 30 (shown in Figure 1 and described herein).

[0069] In operation 508, the neural network is trained using historical sleep stage information. The neural network is trained based on historical sleep stage information by providing historical sleep stage information as input to the neural network. In some embodiments, training the neural network includes letting the neural network be trained. In some embodiments, operation 508 is performed by a processor component that is the same as or similar to the model component 32 (shown in Figure 1 and described herein).

[0070] In operation 510, the trained neural network is configured to (1) determine the duration of REM sleep in the subject during a sleep session, or (2) predict future times during a sleep session in which the subject will experience REM sleep. Based on the output signal and / or other information, the trained neural network is configured to indicate REM sleep in the subject and / or future times when the subject will be in REM sleep. The trained neural network includes one or more hidden layers. The one or more hidden layers of the trained neural network include one or more convolutional layers and one or more recurrent layers of the trained neural network.

[0071] In some embodiments, operation 510 includes providing the neural network with information in the output signal in a temporal set corresponding to individual periods during a sleep session. In some embodiments, operation 510 includes causing the trained neural network to output, based on the information in the temporal set, the detected and / or predicted future times of REM sleep for the subject during the sleep session. In some embodiments, operation 510 is performed by the same or similar processor component as the model component 32 (shown in Figure 1 and described herein).

[0072] In some embodiments, operations 506–510 are replaced by detecting REM sleep in the subject based on cardiac activity parameters (e.g., without using a trained neural network). In these embodiments, REM sleep in the subject is detected in response to the ratio of the low-frequency component of cardiac activity information (e.g., low-frequency cardiac activity parameters, etc.) to the high-frequency component of cardiac activity information (e.g., high-frequency cardiac activity parameters, etc.) exceeding a ratio threshold. In some embodiments, these operations are performed by the same or similar processor component as the model component 32 (shown in Figure 1 and described herein).

[0073] Operation 512 involves having one or more sensory stimulators provide sensory stimulation to the subject during REM sleep to enhance REM sleep. One or more sensory stimulators provide sensory stimulation based on the determination of REM sleep in the subject and / or the predicted timing of REM sleep during the sleep session and / or other information. One or more sensory stimulators provide sensory stimulation to the subject in response to the determination that the subject is in REM sleep and / or a future time that indicates the subject is in REM sleep. In some embodiments, operation 512 is performed by the same or similar processor component as the control component 34 (shown in Figure 1 and described herein).

[0074] In operation 514, one or more sensory stimulators adjust the amount, timing, and / or intensity of sensory stimulation based on one or more brain activity parameters and / or cardiac activity parameters, and values ​​output from one or more hidden layers of a trained neural network.

[0075] In some embodiments, the sensory stimulus includes an audible sound. Calibrating one or more sensory stimulators to adjust the timing and / or intensity of the sensory stimulus includes adjusting the tone interval and / or tone volume in response to the detection of REM sleep. In some embodiments, the stimulus is activated at a specified time to be synchronized with the detection of a PGO wave in the EEG. In some embodiments, operation 514 is performed by a processor component that is the same as or similar to the regulating component 36 (shown in Figure 1 and described herein).

[0076] In the claims, no reference number enclosed in parentheses should be construed as limiting the claims. The term “comprising” or “including” does not exclude the existence of elements or steps other than those stated in the claims. In a claim for an apparatus listing several means, some of these means may be implemented by one and the same hardware item. When an indefinite article is used when referring to a singular element, it does not exclude the existence of a plural form of such element. In any claim for an apparatus listing several means, some of these means may be implemented by one and the same hardware item. The mere fact that certain elements are described in different dependent claims does not indicate that these elements cannot be used in combination.

[0077] While the above description provides details for illustrative purposes based on what is currently considered to be the most practical and preferred embodiment, it should be understood that such details are for that purpose only, and this disclosure is not limited to the expressly disclosed embodiments, but rather is intended to cover modifications and equivalent configurations that fall within the true intent and scope of the accompanying claims. For example, it should be understood that this disclosure considers, wherever possible, one or more features of any embodiment can be combined with one or more features of any other embodiment.< / r> < / r> < / r>

Claims

1. A system configured to extend the duration of rapid eye movement (REM) sleep by delivering auditory stimuli to a subject during a sleep session, wherein the system is One or more sensors configured to generate output signals that transmit information about the subject's sleep stage during the sleep session, One or more sensory stimulation devices configured to provide the auditory stimulation to the subject during the sleep session, One or more hardware processors coupled to the one or more sensors and the one or more sensory stimulators, which receive machine-readable instructions, Based on the output signal, REM sleep is detected in the subject during the sleep session. To extend the duration of REM sleep in the subject during the sleep session, one or more sensory stimulation devices are controlled to provide the subject with auditory stimulation during the REM sleep. One or more hardware processors configured as follows, Includes, The one or more hardware processors can detect REM sleep in the subject. An operation to obtain past sleep stage information for individual subjects in the subject and / or subject group, wherein the past sleep stage information indicates the sleep stages over time during sleep sessions of individual subjects in the subject and / or subject group, obtained by processing signals indicating the brain activity and / or cardiac activity of individual subjects in the subject and / or subject group. The operation involves providing the past sleep stage information as input to a neural network that determines or predicts REM sleep in the subject, thereby training the neural network using the past sleep stage information as training data for supervised learning, Based on the output signal, the trained neural network is trained to: (1) An action to determine the duration of REM sleep experienced by the subject during the sleep session, or (2) An action that causes the subject to predict the future time during the sleep session in which he will experience REM sleep, It is configured to include, The trained neural network includes an input layer, an output layer, and one or more intermediate layers between the input layer and the output layer. The one or more hardware processors control the one or more sensory stimulators to provide auditory stimuli to the subject during REM sleep in order to extend the duration of REM sleep in the subject during the sleep session. (1) For the period during which the subject is experiencing REM sleep or (2) for the future, the operation of generating one or more predicted probability values ​​for individual sleep stages by one or more hidden layers of the trained neural network, The operation of causing one or more sensory stimulation devices to (1) provide auditory stimulation to the subject during the period in which the subject is experiencing REM sleep or (2) at a future time, Based on one or more predicted probability values ​​of individual sleep stages in the one or more intermediate layers, the length (in milliseconds), timing, and / or intensity (dB) of the auditory stimulus provided to the subject is determined, and / or adjusted by the one or more sensory stimulators. It is configured to include, The system wherein the subject group is a subject group that is demographically similar to the subject, and the subject that is demographically similar to the subject is a subject that is similar to the subject in terms of sex, ethnicity, age, and / or health level.

2. The system according to claim 1, wherein the output signals generated by the one or more sensors are configured such that the information relating to the sleep stage of the subject includes information relating to brain activity and / or cardiac activity in the subject.

3. The system according to claim 2, wherein the one or more sensors include one or more electroencephalogram (EEG) electrodes configured to generate information relating to the brain activity, one or more electrocardiogram (ECG) sensors configured to generate information relating to the cardiac activity, and / or one or more photoplethysmography (PPG) sensors configured to generate information relating to the cardiac activity.

4. The system according to claim 2, wherein the output signal generated by one or more sensors is configured such that information relating to the sleep stage of the subject includes information relating to the cardiac activity, and the one or more hardware processors are configured to detect REM sleep in the subject in response to the ratio of the low-frequency component of the cardiac activity information to the high-frequency component of the cardiac activity information exceeding a ratio threshold.

5. The system according to claim 1, wherein the one or more hardware processors are further configured to detect REM sleep in the subject in response to determining that the subject is in a REM sleep state for a continuous threshold period of time during the sleep session.

6. The one or more hardware processors described above are: Based on the output signal, one or more brain activity parameters and / or cardiac activity parameters of the subject indicating the subject's sleep stage are determined, and Based on one or more predicted probability values ​​for individual sleep stages in the one or more intermediate layers, and one or more brain activity parameters and / or cardiac activity parameters, the length (in milliseconds), timing, and / or intensity (dB) of the auditory stimulation to extend the duration of REM sleep in the subject are determined, and / or adjusted to the one or more sensory stimulators. The system according to claim 1, further configured as follows.

7. The system according to claim 6, wherein the one or more hardware processors are configured such that the one or more predicted probability values ​​of individual sleep stages from the one or more hidden layers of the trained neural network include values ​​from one or more convolutional layers and values ​​from one or more recurrent layers of the trained neural network, and determine the length (in milliseconds), timing, and / or intensity (dB) of the auditory stimulus, and / or cause the one or more sensory stimulators to adjust it based on the one or more brain activity parameters and / or cardiac activity parameters, the values ​​from the one or more convolutional layers, and the values ​​from the one or more recurrent layers.

8. The one or more sensory stimulation devices are configured such that the auditory stimulation includes audible sounds, and further, The one or more hardware processors control the one or more sensory stimulators to provide auditory stimuli to the subject during REM sleep in order to extend the duration of REM sleep in the subject during the sleep session. The length (in milliseconds), timing, and / or intensity (dB) of the auditory stimulus are determined by determining the length of the tone of the audible sound, the interval between one tone and the next, the volume of the tone, and / or the tone frequency, and / or The length, timing, and / or intensity of the auditory stimulus is adjusted by one or more sensory stimulators, the adjustment including the length of the tone, the interval, the volume of the tone, and / or the frequency of the tone in response to an indicator that the subject is experiencing one or more micro-arousals. The system according to claim 1, configured to include:

Images