Systems and methods for monitoring slow wave responses to sensory stimuli during sleep sessions

Abstract

The present disclosure relates to automatically monitoring a subject's slow wave response to sensory stimuli during a sleep session. Upon detecting deep NREM sleep, sensory stimuli can be delivered to the subject. The sensory stimuli can be auditory, tactile, visual, or other stimuli. The system delivers stimuli to the subject in the form of blocks of stimuli separated from each other by an intra-block interval. The blocks are separated from each other by an inter-block stimulus. The system compares the subject's stimulated slow wave activity to the subject's unstimulated slow wave activity. The system can update the stimulation parameters based on the comparison and deliver a subsequent block of stimuli. Once the comparison indicates that the stimulated slow wave activity is significantly different from the unstimulated slow wave activity, the system can apply a continuous fixed sensory stimulus to the user according to the most recent stimulation parameters.

Description

background Technical Field

[0002] This disclosure relates to systems and methods for delivering sensory stimuli to an object during a sleep session. Background Technology

[0004] Systems for monitoring sleep and delivering sensory stimuli to subjects during sleep are known. Sleep monitoring and sensory stimulation systems based on electroencephalography (EEG) sensors are known. However, conventional systems require multiple sleep periods to determine whether a subject is responding to sensory stimuli and to adjust stimulation parameters. Summary of the Invention

[0005] Advantageously, the effect of sensory stimuli delivered to a subject can be determined during a single sleep period (e.g., determining whether the subject will respond to the stimulus). The system can adjust stimulus parameters to alter the effect of the sensory stimulus.

[0006] Therefore, one or more aspects of this disclosure relate to a system configured to deliver sensory stimuli to a subject during a sleep period. The system includes: one or more sensors, one or more sensory stimulators, one or more processors, and / or other components. The one or more sensors are configured to generate output signals that convey information related to the subject's brain activity during the sleep period. The one or more sensory stimulators are configured to provide sensory stimulation to the subject during sleep. The one or more processors are coupled to the one or more sensors and the one or more sensory stimulators. The one or more processors are configured by machine-readable instructions. The one or more processors are configured to control the one or more sensory stimulators based on stimulation parameters.

[0007] In some embodiments, one or more sensors include one or more electroencephalogram (EEG) electrodes configured to generate information related to brain activity. In some embodiments, one or more processors are also configured to detect deep non-rapid eye movement (NREM) sleep (also known as N3 sleep or S4 in older sleep stage nomenclature) in the subject. In some embodiments, one or more processors are configured to determine a threshold amount of time during which the subject has been in deep NREM sleep during the sleep period.

[0008] In some embodiments, detecting deep NREM sleep includes training a neural network based on information related to the subject's brain activity captured by EEG electrodes. In some embodiments, based on the output signal, the trained neural network can determine the period during which the subject is experiencing deep NREM sleep. The trained neural network includes an input layer, an output layer, and one or more intermediate layers between the input and output layers.

[0009] In some embodiments, one or more processors are configured such that once deep NREM sleep is detected, the processors apply stimulation to the object in the form of blocks of repetitive stimuli. In some embodiments, the repetitive stimuli may be repetitive vibrations, repetitive light pulses, and / or other repetitive stimuli. In some embodiments, the blocks are separated from each other by inter-block intervals, and the repetitive stimuli are separated from each other by intra-block intervals. In some embodiments, the inter-block intervals are longer than the intra-block intervals. In some embodiments, the inter-block intervals may have a certain length (e.g., more than 3 seconds or some other length). In some embodiments, the intra-block intervals may have a specific length shorter than the length of the inter-block intervals (e.g., 0.1-2 seconds or some other length). These embodiments are not limiting, and the lengths can vary.

[0010] In some embodiments, one or more processors are configured to detect unstimulated slow-wave activity in a subject during a sleep period. In some embodiments, unstimulated slow-wave activity includes slow-wave activity in the subject during inter-block intervals. One or more processors are configured to detect stimulated slow-wave activity in a subject during a sleep period. In some embodiments, stimulated slow-wave activity includes slow-wave activity in the subject during blocks of repetitive stimuli (i.e., repetitive vibrations and / or repetitive pulses). One or more processors may compare stimulated slow-wave activity with unstimulated slow-wave activity. One or more processors may update stimulation parameters of the stimulus based on this comparison.

[0011] In some embodiments, one or more processors are configured to control a sensory stimulator based on updated stimulation parameters. In some embodiments, the one or more processors may cause the sensory stimulator to deliver subsequent stimulus blocks to the object according to the updated parameters. The one or more processors may then detect and compare unstimulated slow-wave activity and stimulated slow-wave activity for the subsequent blocks based on the updated stimulation parameters. In some embodiments, the one or more processors may repeat these steps until the stimulated slow-wave activity is significantly higher than the unstimulated slow-wave activity. To determine whether the stimulated slow-wave activity is significantly higher than the unstimulated slow-wave activity, the one or more processors may compare the difference between the unstimulated and stimulated slow-wave activities with a threshold. In some embodiments, the threshold is determined based on the minimum difference that indicates the effectiveness of the stimulus.

[0012] In some embodiments, once the difference between stimulated and unstimulated slow-wave activity breaches a threshold, one or more processors may enable the stimulator to deliver continuous stimulation to the subject based on the most recently updated stimulation parameters. In some embodiments, the threshold may represent the minimum difference between stimulated and unstimulated slow-wave activity used to indicate the effectiveness of the stimulation.

[0013] In some embodiments, one or more sensory stimulators are configured such that the sensory stimulation includes audible tones. In some embodiments, one or more sensory stimulators are configured such that the sensory stimulation includes tactile vibrations. In some embodiments, one or more sensory stimulators are configured such that the sensory stimulation includes light pulses. One or more processors are configured to update stimulation parameters based on a comparison of stimulated slow-wave activity with unstimulated slow-wave activity, including: changing the duration of each stimulus, the duration of inter-block intervals, the duration of intra-block intervals, the number of stimuli, the intensity of the stimulus, and / or the stimulus frequency; and / or causing one or more sensory stimulators to adjust the stimulation parameters.

[0014] These and other objects, features, and characteristics of this disclosure, as well as the methods of operation and function of the related elements of the structure, and the economy of combination and manufacture of the components, will become more apparent by considering the following description and appended claims with reference to the accompanying drawings, all of which form part of this specification, wherein the same reference numerals denote corresponding components in the various drawings. However, it should be clearly understood that the drawings are for illustrative and descriptive purposes only and are not intended to be a definition of limitation of this disclosure. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a system configured to deliver sensory stimuli to an object during a sleep period, according to one or more embodiments.

[0016] Figure 2 Several operations performed by the system according to one or more embodiments are illustrated.

[0017] Figure 3 An exemplary architecture of a deep neural network as part of a system, according to one or more embodiments, is shown.

[0018] Figure 4 The delivery of block stimuli to an object during a sleep period is illustrated according to one or more embodiments.

[0019] Figure 5 Examples are shown of the percentage difference between stimulated and unstimulated slow-wave activity in a subject during a sleep period relative to the onset of stimulation, according to one or more embodiments.

[0020] Figure 6 Examples of the correlation between the effect of block stimulation and the effect of continuous fixed stimulation according to one or more embodiments are shown.

[0021] Figure 7 A method for delivering sensory stimuli to an object during a sleep period is illustrated according to one or more embodiments. Specific Implementation

[0022] As used herein, the singular forms “a,” “an,” and “the” include plural references unless the context clearly indicates otherwise. As used herein, the term “or” means “and / or” unless the context clearly indicates otherwise. As used herein, the expression “coupled” means that the parts or components are directly or indirectly (i.e., through one or more intermediate parts or components) connected or operated together, provided that a linking occurs. As used herein, “direct coupling” means that two elements are in direct contact with each other. As used herein, “fixedly coupled” or “fixed” means that two components are coupled to move as a single component while maintaining a constant orientation relative to each other.

[0023] Unless expressly stated herein, directional phrases used herein, such as, but not limited to, top, bottom, left, right, up, down, front, back, and their derivatives, refer to the orientation of the elements shown in the figures and do not limit the claims.

[0024] Figure 1 This is a schematic diagram of a system 10 configured to deliver sensory stimuli to a subject 12 during sleep periods. System 10 is configured to facilitate the delivery of sensory stimuli to the subject 12 to determine whether the subject 12 responds to the sensory stimuli, update stimulus parameters, and / or for other purposes. System 10 is configured to deliver sensory stimuli, including auditory, tactile, light, and / or other stimuli, during sleep. In some embodiments, stimuli are delivered to the subject only if a processor in system 10 (described below) has determined that the subject 12 is in deep NREM sleep. In some embodiments, system 10 delivers stimuli to the subject 12 in blocks of repetitive stimuli (e.g., repetitive vibrations and / or repetitive light pulses). As described herein, one or more processors may compare stimulated slow-wave activity in the subject 12 (i.e., during block stimulation) with unstimulated slow-wave activity in the subject 12 (i.e., between blocks of stimulation or before blocks of stimulation). This comparison indicates the effect of the stimulus on the subject 12. One or more processors may update stimulus parameters based on the comparison. In some embodiments, system 10 is configured to repeat these steps until the difference between stimulated slow-wave activity and unstimulated slow-wave activity exceeds a threshold. Once the difference exceeds the threshold, one or more processors can control a sensory stimulator to deliver continuous stimulation to the subject 12 based on the most recently updated stimulation parameters.

[0025] Adjusting the stimulation parameters is important to ensure that stimulation during sleep is effective for subject 12. The use of block stimulation reduces the necessary adjustment cycle of this process from several sleep periods to a single sleep period. This allows the stimulation process to improve the subject's sleep more quickly and effectively. System 10 also utilizes machine learning models (e.g., deep neural networks and / or any other supervised machine learning algorithms described below) to automatically, in real-time or near real-time, closed-loop sensor output signals to determine the subject's sleep stages during sleep periods. Figure 1 As shown, system 10 includes one or more of the following components: sensor 14, sensory stimulator 16, external resource 18, processor 20, electronic storage 22, object interface 24, and / or other components. These components are further described below.

[0026] Sensor 14 is configured to generate an output signal that conveys information related to the sleep stage of subject 12 during a sleep period. The output signal conveying information related to the sleep stage of subject 12 may include information related to brain activity in subject 12. Thus, sensor 14 is configured to generate an output signal that conveys information related to brain activity. In some embodiments, sensor 14 is configured to generate an output signal that conveys information related to stimuli provided to subject 12 during a sleep period. In some embodiments, information from the output signal of sensor 14 is used to control sensory stimulator 16 to provide sensory stimulation to subject 12 (as described below).

[0027] Sensor 14 may include one or more sensors that generate output signals that directly convey information related to brain activity in subject 12. For example, sensor 14 may include electroencephalography (EEG) electrodes configured to detect electrical activity along the scalp of subject 12 caused by electrical flow in the brain of subject 12. Sensor 14 may also include one or more sensors that generate output signals that indirectly convey information related to brain activity in subject 12. For example, one or more sensors 14 may include heart rate sensors that generate output based on: the heart rate of subject 12 (e.g., sensor 14 may be a heart rate sensor that can be located on the chest of subject 12, and / or a wristband configured to be on the wrist of subject 12, and / or a wristband located on another limb of subject 12), the movement of subject 12 (e.g., sensor 14 may include an accelerometer that can be carried on a wearable device, such as a wristband around the wrist and / or ankle of subject 12, so that activity recorder signals can be used to analyze sleep), the breathing of subject 12, and / or other characteristics of subject 12.

[0028] In some embodiments, sensor 14 may include one or more of EEG electrodes, a respiration sensor, a pressure sensor, a vital signs camera, a functional near-infrared sensor (fNIR), a temperature sensor, a microphone, and / or other sensors configured to generate output signals (e.g., their quantity, frequency, intensity, and / or other characteristics) in relation to stimuli provided to subject 12, brain activity of subject 12, and / or other sensors. Although sensor 14 is shown at a single location near subject 12, this is not intended to be limiting. Sensor 14 may include sensors disposed in multiple locations, such as within (or in communication with) sensory stimulator 16, coupled to (in a removable manner) clothing of subject 12, worn by subject 12 (e.g., as a headband, wristband, etc.), configured to point at subject 12 while subject 12 is sleeping (e.g., a camera transmitting output signals related to the movement of subject 12), coupled to the bed and / or other furniture where subject 12 is sleeping, and / or disposed in other locations.

[0029] exist Figure 1 In this design, sensor 14, sensory stimulator 16, processor 20, electronic storage 22, and object interface 24 are shown as separate entities. This is not limiting. Some and / or all components and / or other components of system 10 may be grouped into one or more individual devices. For example, these and / or other components may be included in headset 201 and / or other clothing worn by object 12. Other clothing may include hats, vests, wristbands, and / or other garments. Headset 201 and / or other clothing may include, for example, sensing electrodes, reference electrodes, one or more devices associated with EEG, means for delivering auditory stimulation (e.g., wired and / or wireless audio devices and / or other devices), and one or more audio speakers. In some embodiments, headset 201 and / or other clothing may include means for delivering visual, somatosensory, electrical, magnetic, and / or other stimuli to the object. In this example, the audio speakers may be located in and / or near and / or elsewhere in the ear of object 12. Reference electrodes may be located behind the ear of object 12, and / or elsewhere. In this example, the sensing electrodes can be configured to generate an output signal that conveys information and / or other information related to the brain activity of subject 12. The output signal can be transmitted wirelessly and / or via wires to a processor (e.g., Figure 1 The processor 20 shown may include or exclude a computing device (e.g., a bedside laptop computer) and / or other devices. In this example, acoustic stimulation can be delivered to the object 12 via a wireless audio device and / or speaker. In this example, the sensing electrodes, reference electrodes, and EEG device may be, for example, provided by... Figure 1 Sensor 14 is shown in the image. Wireless audio devices and speakers can be, for example, made by... Figure 1 The sensory stimulator 16 shown is illustrated. In this example, the computing device may include a processor 20, electronic storage 22, an object interface 24, and / or Figure 1 Other components of system 10 shown.

[0030] Stimulator 16 is configured to provide sensory stimulation to subject 12. Sensory stimulator 16 is configured to provide auditory, visual, somatosensory, electrical, magnetic, and / or sensory stimulation to subject 12 before, during, and / or at other times. In some embodiments, a sleep period may include any time period during which subject 12 is sleeping and / or attempting to sleep. Sleep periods may include nighttime sleep, sudden sleep, and / or other sleep periods. For example, sensory stimulator 16 may be configured to provide stimulation to subject 12 during a sleep period to enhance EEG signals during deep NREM sleep of subject 12 and / or for other purposes.

[0031] Sensory stimulator 16 is configured to influence deep NREM sleep in subject 12 through non-invasive brain stimulation and / or other methods. Sensory stimulator 16 can be configured to influence deep NREM sleep through non-invasive brain stimulation using auditory, electronic, magnetic, visual, somatosensory, and / or other sensory stimuli. Auditory, electronic, magnetic, visual, somatosensory, and / or other sensory stimuli can include auditory stimulation, visual stimulation, somatosensory stimulation, electrical stimulation, magnetic stimulation, combinations of different types of stimulation, and / or other stimuli. Auditory, electronic, magnetic, visual, somatosensory, and / or other sensory stimuli include odor, sound, visual stimulation, touch, taste, somatosensory stimulation, tactile, electronic, magnetic, and / or other stimuli. Sensory stimuli can have intensity, timing, and / or other characteristics. For example, sound tones can be provided to subject 12 to influence deep NREM sleep in subject 12. Sound tones can include one or more tones of a defined length separated from each other by tonal intervals. The volume (e.g., intensity) of an individual tone can be modulated based on various factors (as described herein). The length of individual pitches (e.g., timing) and / or the interval between pitches (i.e., the interval within a block) can also be adjusted. Pitch and tone can also be adjusted. In some embodiments, stimuli can be delivered to the object in blocks. In the example of auditory stimulation, each block stimulus has 15 pitches. In this example, each pitch has the form of a 50-millisecond long pitch (e.g., pink noise pitch with a frequency limit of 500Hz to 5kHz). In some embodiments, the duration of each individual stimulus can fall within the range of 10-100 milliseconds (or another duration range). The interval between blocks can be 15 seconds, and the interval within a block (i.e., the interval between pitches) can be 1 second. In some embodiments, the default volume of the stimulus can be 20dB. This example is not limiting, and stimulus parameters can vary.

[0032] Examples of sensory stimulator 16 may include: a sound generator, a speaker, a music player, a tone generator, a vibrator (e.g., a piezoelectric element) for delivering vibrational stimulation, a coil that generates a magnetic field to directly stimulate the cerebral cortex, one or more light generators or lamps, a fragrance diffuser, and / or other devices. In some embodiments, sensory stimulator 16 is configured to adjust the intensity, timing, and / or other parameters of the stimulation provided to subject 12 (e.g., as described below).

[0033] External resources 18 include: information sources (e.g., databases, websites, etc.), external entities participating in system 10 (e.g., one or more external sleep monitoring devices, a healthcare provider's medical record system, etc.), and / or other resources. In some embodiments, external resources 18 include components that facilitate information communication, one or more servers outside system 10, networks (e.g., the Internet), electronic storage, devices associated with Wi-Fi technology, and others. Technology-related equipment, data input devices, sensors, scanners, computing devices associated with individual objects, and / or other resources. In some implementations, some or all of the functions attributable here to external resource 18 may be provided by resources included in system 10. External resource 18 may be configured to communicate with processor 20, object interface 24, sensor 14, electronic storage 22, sensory stimulator 16, and / or other components of system 10 via wired and / or wireless connections, via networks (e.g., local area networks and / or the Internet), via cellular technology, via Wi-Fi technology, and / or via other resources.

[0034] Processor 20 is configured to provide information processing capabilities within system 10. Thus, processor 20 may include one or more of the following: a digital processor, an analog processor, digital circuitry designed to process information, analog circuitry designed to process information, a state machine, and / or other mechanisms for electronically processing information. Although processor 20 is... Figure 1 The processor 20 is shown as a single entity, but 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 stimulator 16, object interface 24, etc.), or the processor 20 may represent the processing capabilities of multiple devices operating in concert. In some embodiments, the processor 20 may be and / or be included in a computing device such as a desktop computer, laptop computer, smartphone, tablet computer, server, and / or other computing device. Such a computing device may run one or more electronic applications with a graphical object interface configured to facilitate interaction between objects and system 10.

[0035] like Figure 1As shown, 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 encoded and / or otherwise embedded in processor 20. One or more computer program components may include one or more of information component 30, model component 32, control component 34, modulation component 36, and / or other components. Processor 20 may be configured to execute components 30, 32, 34, and / or 36 via software, hardware, firmware, and certain combinations of software, hardware, and / or firmware, and / or other mechanisms for configuring processing capabilities on processor 20.

[0036] It should be understood that, although Figure 1 Components 30, 32, 34, and 36 are shown as co-located within a single processing unit, but in embodiments where processor 20 includes multiple processing units, one or more of components 30, 32, 34, and / or 36 may be located remotely from other components. The description below of the functionality provided by the different components 30, 32, 34, and / or 36 is for illustrative purposes and is not intended to be limiting, as any of 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 functionality may be provided by other components 30, 32, 34, and / or 36. As another example, processor 20 may be configured to execute one or more additional components that may perform some or all of the functionality attributed below to one of components 30, 32, 34, and / or 36.

[0037] Information component 30 is configured to determine one or more brain activity parameters and / or other information of subject 12. Brain activity parameters are determined based on output signals and / or other information from sensor 14. Brain activity parameters indicate the sleep depth of subject 12. In some embodiments, information related to brain activity in the output signal indicates sleep depth over time. In some embodiments, the information indicating sleep depth over time is information related to deep NREM sleep in subject 12, or includes information related to deep NREM sleep in subject 12.

[0038] In some embodiments, information indicating sleep depth over time may indicate other sleep stages of subject 12. For example, the sleep stage of subject 12 may be associated with deep NREM sleep, rapid eye movement (REM) sleep, and / or other sleep stages. Deep NREM sleep may be stage N3, and / or other deep sleep stages. In some embodiments, the sleep stage 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 is one or more additional brain activity parameters and / or is associated with one or more additional brain activity parameters.

[0039] In some embodiments, information relating to brain activity indicating sleep depth over time includes and / or includes EEG information and / or other information generated during and / or at other times during the sleep period of subject 12. In some embodiments, brain activity parameters may be determined based on EEG information and / or other information. In some embodiments, brain activity parameters may be determined by information component 30 and / or other components of system 10. In some embodiments, brain activity parameters may be part of previously determined historical sleep stage information obtained from external resource 18 (described below). In some embodiments, one or more brain activity parameters are and / or relate to frequency, amplitude, phase, the presence of specific sleep patterns such as eye movements, pons-geniculate-occipital (PGO) waves, slow waves, and / or other characteristics of the EEG signal. In some embodiments, one or more brain activity parameters are determined based on the frequency, amplitude, and / or other characteristics of the EEG signal. In some embodiments, the determined brain activity parameters and / or characteristics of the EEG may include and / or indicate a sleep stage corresponding to the aforementioned deep NREM sleep stage.

[0040] Information component 30 is configured to obtain historical sleep stage information. In some embodiments, the historical sleep stage information is specific to subject 12 and / or other subjects. The historical sleep stage information relates to brain activity and / or other physiological characteristics of the subject group and / or subject 12, indicating the sleep stage over time during previous sleep periods of the subject group and / or subject 12. The historical sleep stage information relates to other brain parameters and / or other information of the subject group and / or subject 12 during the sleep stage and / or corresponding sleep period.

[0041] In some embodiments, the information component 30 is configured to electronically obtain historical sleep stage information from external resources 18, electronic storage 22, and / or other information sources. In some embodiments, electronically obtaining historical sleep stage information from external resources 18, electronic storage 22, and / or other information sources includes querying one or more databases and / or servers; uploading and / or downloading information; facilitating object input, sending and / or receiving emails, sending and / or receiving text messages, and / or sending and / or receiving other communications, and / or other acquisition 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 described above, electronic storage 22, etc.), arrange the information in one or more electronic databases (e.g., electronic storage 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 periods, number of sleep periods, etc.), and / or perform other operations.

[0042] Model component 32 is configured to cause a trained neural network and / or any other supervised machine learning algorithm to detect deep NREM sleep in subject 12. In some embodiments, this may include and / or include determining the period during which subject 12 is experiencing deep NREM sleep during a sleep period and / or other operations. The determined deep NREM sleep and / or timing indicates whether subject 12 is in deep NREM sleep for stimulation and / or other information. As a non-limiting example, the trained neural network may be made to indicate the determination of the subject's deep sleep stage and / or the timing of the deep sleep stage based on the output signal (e.g., using information in the output signal as input to the model) and / or other information. In some embodiments, model component 32 is configured to provide the information in the output signal to the neural network in the form of a time set corresponding to individual cycles during the sleep period. In some embodiments, model component 32 is configured to cause the trained neural network to output the sleep stage for the determined deep NREM sleep of subject 12 during the sleep period based on the time set of information. (The following is relative to...) Figure 2-3 The functionality of model component 32 will be discussed further.

[0043] For example, neural networks can be based on large collections of neural units (or artificial neurons). Neural networks can loosely mimic the way a biological brain works (e.g., through large clusters of biological neurons connected by axons). Each neural unit in a neural network can be connected to many other neural units in the network. These connections can strengthen or inhibit their influence on the activation state of the connected neural units. In some embodiments, each individual neural unit can have a summation function that combines the values ​​of all its inputs. In some embodiments, each connection (or the neural unit itself) can have a threshold function such that a signal must exceed a threshold before being allowed to propagate to other neural units. These neural network systems can be self-learning and trained rather than explicitly programmed, and can perform significantly better in certain domains of problem-solving compared to conventional computer programs. In some embodiments, a neural network can include multiple layers (e.g., where signal paths traverse from previous layers to subsequent layers). In some embodiments, a neural network can utilize backpropagation techniques, where positive stimulation is used to reset weights on “previous” neural units. In some embodiments, stimulation and inhibition in a neural network can be more free-flowing, where connections interact in more chaotic and complex ways.

[0044] A trained neural network may include one or more intermediate layers or hidden layers. The intermediate layers of a trained neural network include one or more convolutional layers, one or more recurrent layers, and / or other layers. Each intermediate layer receives information from another layer as input and produces a corresponding output. The detected deep NREM sleep stage is generated based on information processed by the layers of the neural network from the output signal of sensor 14.

[0045] Control component 34 is configured to control stimulator 16 to provide stimulation to subject 12 during sleep and / or at other times. Control component 34 is configured to cause sensory stimulator 16 to provide sensory stimulation to subject 12 during deep NREM sleep to influence deep NREM sleep in subject 12 during sleep periods. Control component 34 is configured to cause sensory stimulator 16 to provide sensory stimulation to subject 12 based on detected deep NREM sleep stages (e.g., output from model component 32) and / or other information. Control component 34 is configured to cause sensory stimulator 16 to provide sensory stimulation to subject 12 based on detected deep NREM sleep stages and / or other information over time during sleep periods. Control component 34 is configured to cause sensory stimulator 16 to provide sensory stimulation to subject 12 in response to subject 12 being in or potentially in deep NREM sleep for stimulation. For example, control component 34 is configured to control one or more sensory stimulators 16 to provide sensory stimulation to subject 12 during deep NREM sleep to influence deep NREM sleep in subject 12 during a sleep period, including: determining a period of deep NREM sleep in subject 12; causing one or more sensory stimulators 16 to provide sensory stimulation to subject 12 during the period of deep NREM sleep in subject 12; and / or causing one or more sensory stimulators 16 to modulate (e.g., as described herein) the amount, timing, and / or intensity of the sensory stimulation provided to subject 12 based on one or more values ​​of one or more intermediate layers. In some embodiments, stimulators 16 are controlled by control component 34 to influence deep NREM sleep (as described herein) through stimulation delivered during deep NREM sleep (e.g., ambient auditory, magnetic, electronic, and / or other stimuli).

[0046] In some embodiments, control component 34 is configured to control sensory stimulator 16 to deliver sensory stimulation to subject 12 in response to model component 32 determining that subject 12 has maintained a continuous threshold amount of time in deep NREM sleep during a sleep period. For example, model component 32 and / or control component 34 may be configured such that when deep NREM sleep is detected, model component 32 starts a (physical or virtual) timer configured to track the time spent by subject 12 in deep NREM sleep. Control component 34 is configured to deliver auditory stimulation in response to subject 12 spending more than a predefined continuous time threshold in continuous deep NREM sleep. In some embodiments, the predefined duration threshold is determined at the time of manufacture of system 10 and / or at other times. In some embodiments, the predefined duration threshold is determined based on demographic information from previous sleep periods of subject 12 and / or similar subjects (e.g., as described above). In some embodiments, the predefined duration threshold may be adjusted via object interface 24 and / or other adjustment mechanisms.

[0047] In some embodiments, the predefined threshold for the duration of deep NREM sleep can be, for example, one minute and / or other durations. As a non-limiting example, control component 34 can be configured to initiate auditory stimulation once one minute of sustained deep NREM sleep is detected in object 12. In some embodiments, once stimulation begins, control component 34 is configured to control stimulation parameters. Upon detection of a sleep stage transition (e.g., from deep NREM sleep to another sleep stage), control component 34 is configured to stop stimulation.

[0048] Modulation component 36 is configured to cause sensory stimulator 16 to modulate the amount, timing, and / or intensity of sensory stimulation. Modulation component 36 is configured to cause sensory stimulator 16 to modulate the amount, timing, and / or intensity of sensory stimulation based on brain activity parameters, values ​​from the output of intermediate layers of a trained neural network, and / or other information. For example, sensory stimulator 16 may modulate the timing and / or intensity of sensory stimulation based on brain activity parameters, values ​​from the output of convolutional layers, values ​​from the output of recurrent layers, and / or other information. For example, modulation component 36 may be configured to deliver sensory stimulation at an intensity proportional to the predicted probability value (e.g., from the output of intermediate layers of the neural network) of a particular sleep stage (e.g., deep NREM). In this example, the higher the probability of deep NREM sleep, the greater the likelihood that stimulation will continue. If sleep micro-awakening is detected and the sleep stage remains in deep NREM, modulation component 36 may be configured to reduce the intensity of stimulation (e.g., reduce it by 5 dB) in response to individual micro-awakening detection.

[0049] As a non-restrictive example, Figure 2Several operations performed by system 10 and as described above are shown. Figure 2 In the example shown in process 200, the EEG signal 202 is processed and / or additionally (e.g., by...) within time window 204. Figure 1 The information component 30 and model component 32 shown are provided to the deep neural network 206. The deep neural network 206 detects sleep stages (e.g., N3, N2, N1, REM, and wakefulness). Determination 208 indicates whether the object is in deep NREM (N3) sleep. If the object is not in deep NREM sleep, the deep neural network 206 continues to process EEG signals in real time 202. The deep neural network 206 can determine the object's sleep stage, such as regarding... Figure 3 As described. Additionally or optionally, the deep neural network 206 can use methods described in the following publications to determine the sleep stage of an object: Bresch E, The entire contents of “Recurrent Deep Neural Networks for Real-Time Sleep Stage Classification From Single Channel EEG” by U and Garcia-Molina G (2018) published in Frontiers in Computational Neuroscience are incorporated herein by reference.

[0050] like Figure 2 As shown, in response to the sleep stage determination 208 indicating deep NREM sleep, determination 210 indicates whether the system is calibrated. Calibration may include stimulation parameters specifying the amount, timing, and / or intensity of the optimal sensory stimulus for the subject. In response to determination 210 indicating the system is not calibrated, block stimulation 214 is applied to the subject during the sleep period. Stimulation parameters for block stimulation may include default amount, timing, and intensity, and / or user-specified amount, timing, and intensity. Block stimulation 214 may be repeated until the enhancement of slow-wave activity exceeds a threshold 218. In some embodiments, threshold 218 may represent the minimum enhancement of slow-wave activity used to indicate the effectiveness of the stimulus. In some embodiments, the enhancement of slow-wave activity may be measured as the difference (e.g., percentage difference) between unstimulated and stimulated slow-wave activity in the subject during the sleep period. Each time the difference between stimulated and unstimulated slow-wave activity does not exceed threshold 218, settings 216 are adjusted. If the subject is no longer in deep NREM sleep 212, the process returns to the sleep stage determination (staging) process of neural network 206. If the subject is still in deep NREM sleep 212, block stimulation 214 is applied again. Once the difference exceeds the threshold 218, the system has been calibrated.

[0051] Once the system is calibrated, if the subject is still in deep NREM sleep 220, the system delivers continuous fixed-interval stimuli 222 to the subject. The parameters of the stimuli (e.g., magnitude, timing, and intensity) are those delivered in block stimuli 214 that break the threshold 218. Continuous fixed-interval stimuli 222 are delivered to the subject for the remainder of the sleep period and subsequent sleep periods. The system can continue to extract information 224 about the subject's sleep, such as alpha and beta power, slow-wave activity, and sleep depth. This information can be used to adjust or terminate sensory stimulation during each sleep period.

[0052] Figure 3 This shows the system 10 ( Figure 1 and 2 A part of a deep neural network (e.g., Figure 2 The example architecture 300 of the deep neural network 206 shown is shown. Figure 3 A deep neural network architecture 300 for three (expanded) EEG windows 304, 306, and 308 is shown. In some embodiments, windows 304, 306, and 308 may be windows of EEG signals 302 for a predefined time period (e.g., six seconds). Architecture 300 includes convolutional layers 310, 312, and 314 and recurrent layers 322, 324, and 326. As described above, convolutional layers 310, 312, and 314 can be considered as filters and produce convolutional outputs 316, 318, and 320, which are fed into recurrent layers 322, 324, and 326 (LSTM (Long Short-Term Memory) layers in this example). The outputs of the processed architecture 300 for each window 304, 306, and 308 are a set of predicted probabilities for the respective sleep stage, referred to as “soft outputs” 328. “Hard” predictions 330 are generated by architecture 300 (…). Figure 1 The model component 32 shown is determined by predicting 332 in relation to the sleep stage associated with the “soft” output that has the highest value (e.g., as described below). The terms “soft” and “hard” are not restrictive, but rather helpful in describing the operations performed by the system. For example, the term “soft output” can be used because any decision is possible at this stage. In fact, the final decision may depend on, for example, post-processing of the soft output. Figure 3 In this context, "Argmax" is an operator that indicates the sleep stage associated with the highest "soft output" (e.g., the highest probability).

[0053] 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). Model component 32 ( Figure 1The system is configured such that the probability set constitutes a so-called soft decision vector, which can be transformed into a hard decision by determining which sleep stage is associated with the highest probability value (in a continuum of possible values) relative to other sleep stages. These soft decisions allow the system 10 to continuously consider different possible sleep states, rather than being forced to decide which discrete sleep stage “bucket” a particular EEG information fits into (as in prior art systems).

[0054] Back Figure 1 Model component 32 is configured such that the values ​​from the convolutional layer output and the soft-determination value output are vectors comprising continuous values, rather than discrete values ​​such as those for sleep stages. Therefore, when the deep neural network detects the occurrence of deep NREM sleep, the convolutional and recursive (soft-determination) value outputs can be used by system 10 to modulate the volume of the stimulus. Furthermore, as described herein, parameters determined based on the raw sensor output signal (e.g., EEG signal) (e.g., by...) Figure 1 The information component 30 shown can be used to modulate the stimulus settings.

[0055] As described above, modulation component 36 is configured to cause sensory stimulator 16 to modulate the amount, timing, and / or intensity of sensory stimulation. Modulation component 36 is configured to cause sensory stimulators to modulate the amount, timing, and / or intensity of sensory stimulation based on one or more brain activity and / or other parameters, values ​​and / or other information from the outputs of convolutional and / or recurrent layers of a trained neural network. As an example, the inter-block or intra-block spacing of auditory stimulation provided to subject 12 may be adjusted and / or otherwise controlled (e.g., modulated) based on value outputs from a deep neural network, such as convolutional layer value outputs and recurrent layer value outputs (e.g., sleep stage (soft) prediction probabilities). In some embodiments, modulation component 36 is configured to cause one or more sensory stimulators 16 to modulate the amount, timing, and / or intensity of sensory stimulation, wherein the modulation includes adjusting the inter-block spacing, intra-block spacing, stimulation intensity, and / or stimulation frequency in response to indication that subject 12 is experiencing one or more micro-awakenings.

[0056] In some embodiments, modulation component 36 is configured to modulate sensory stimuli solely based on brain activity and / or other parameters, which may be determined based on output signals from sensor 14 (e.g., based on raw EEG signals). In these embodiments, the output of a deep neural network (and / or other machine learning model) continues to be used to detect sleep stages (e.g., as described above). However, stimulus intensity and timing are modulated based on brain activity and / or other parameters or characteristics determined based on sensor output signals. In some embodiments, information in the sensor output signals or information determined based on sensor output signals may also be combined with intermediate outputs of the network (e.g., outputs of convolutional layers) or the final output (soft stages) to modulate intensity and timing (e.g., as described herein).

[0057] Figure 4 Figure 400 illustrates the block stimulation directed at objects (e.g., during sleep periods) using a block stimulus. Figure 1 The delivery of 12) is shown. As shown in Table 400, EEG data (EEG data 414) indicates deep NREM (N3) sleep. Therefore, one or more processors (e.g., such as Figure 1 As shown in 20), block stimuli are applied to the subject during deep NREM sleep (e.g., according to...). Figure 2 The process shown is block 401). In some embodiments, the delivery of the first stimulus block may be synchronized with the rising state of EEG data 414. In some embodiments, the rising state of the slow wave includes a time period (e.g., a 300-millisecond interval) starting from a zero crossover (e.g., a second zero crossover). In some embodiments, the stimulus may be an auditory vibration, a tactile vibration, a light pulse, and / or other forms of stimulation. The block stimulus may be delivered to the object according to the following (e.g., Figure 1 Object 12) shown: Stimulation parameters, such as vibration duration, pulse duration, vibration frequency, pulse frequency, intra-block interval between vibrations, intra-block interval between pulses, inter-block interval, and / or other parameters. As mentioned herein, stimulation parameters may include any of the foregoing parameters and / or other parameters.

[0058] like Figure 4 As shown, the block stimulus comprises a block 401 of 15 stimuli 404. In other embodiments, the number of stimuli per block can vary. Each stimulus 404 within block 401 is delivered with constant intensity, duration, and frequency, and at constant intra-block intervals 406. Figure 4 As shown, each stimulus within block 402 is separated from the others by an intra-block interval 406 of 1 second. In some embodiments, the duration of the intra-block interval 406 can be varied. In embodiments where the stimulus is in the form of an auditory vibration, the pitch can be randomized within the range of 500-2000 Hz. In some embodiments, the pitch can be randomized within a wider or narrower range. Figure 4 As shown, the inter-block interval 412 can be 15 seconds. In some embodiments, the inter-block interval 412 can have other durations. In some embodiments, the inter-block interval 412 can have the same duration as the stimulus block, a shorter duration, or a longer duration. In embodiments where the stimulus is in the form of auditory vibration, the initial intensity can be 20 dB. In some embodiments, the object can set the initial intensity.

[0059] In some embodiments, one or more processors (e.g., such as...) Figure 1 As shown in 20), the first piece 401 of the stimulus 404 can be delivered to the object (e.g., as shown in 20). Figure 1(as shown in 12). One or more processors can then process the EEG data 414 to determine the effect of the block stimulus on the object's slow-wave activity. In some embodiments, one or more processors filter the stimulated slow-wave activity 410 by a frequency band (e.g., 0.5-4 Hz), square the filtered data, and / or calculate a moving average (and / or other aggregations of the data) over a time period (e.g., four seconds). The one or more processors can then compare the results to the average (and / or other aggregations) of the unstimulated slow-wave activity 408 over a time period (e.g., two seconds) prior to the occurrence of block 401. In some embodiments, the method for calculating the effect of the stimulus on the object can vary. Based on the comparison of the processed stimulated slow-wave activity data 410 and the processed unstimulated slow-wave activity data 408, one or more processors can determine differences. This comparison may include differences in slow-wave activity levels, percentage differences, and / or any other comparison.

[0060] Then, one or more processors can compare the difference between unstimulated slow-wave activity 408 and stimulated slow-wave activity 410 with a threshold. In some embodiments, the threshold may represent the minimum difference between unstimulated slow-wave activity 408 and stimulated slow-wave activity 410 used to indicate the effectiveness of stimulation. In some embodiments, the threshold may be a 40% difference between unstimulated slow-wave activity 408 and stimulated slow-wave activity 410. In some embodiments, the threshold may vary.

[0061] Figure 5 A graph 500 illustrates an example of the percentage difference between stimulated and unstimulated slow-wave activity in an object during an exemplary sleep period, relative to the start of stimulation. The percentage difference in slow-wave activity 506 is zero 504 before the first stimulus 502 begins. Once the stimulus begins (i.e., at time zero), the percentage difference 506 increases. Figure 5 The maximum percentage difference reached across the 15 stimulus blocks shown is approximately 12%. In this example, if the threshold were 40%, this percentage difference would not exceed the threshold.

[0062] Back Figure 4 In some embodiments, if the difference between stimulated slow-wave activity levels and unstimulated slow-wave activity levels does not exceed a threshold, then one or more processors (e.g., such as...) Figure 1As shown in 20), stimulus parameters (e.g., duration, intensity, frequency, inter-block interval, and / or intra-block interval) can be updated. In embodiments where the stimulus is in the form of an auditory vibration, one or more processors can increase the volume by a given amount (e.g., 3 dB). In some embodiments, one or more processors can increase the intensity of a light pulse or tactile vibration. In some embodiments, one or more processors can increase the duration of each stimulus 404, the frequency of stimulus 404, the duration of each intra-block interval 406, and / or the duration of each inter-block interval 412. In some embodiments, one or more processors can then deliver subsequent blocks 402 to the object based on the updated stimulus parameters.

[0063] In some embodiments, one or more processors may repeatedly apply stimulus blocks to a user, perform a comparison between stimulated slow-wave activity and unstimulated slow-wave activity, compare the difference to a threshold, and update the stimulus parameters until the difference exceeds the threshold. Figure 4 The block stimulation shown takes approximately 5 minutes to perform ten rounds. If the first deep NREM sleep period is shortened (e.g., due to micro-awakening, transition to another sleep stage, or transition to an awake state), stimulation can be directed to the subject (e.g., such as...). Figure 1 As shown in 12), fewer blocks of stimulation are delivered. This time requirement is significantly shorter than in the previous system, where multiple adjustments to the stimulation settings would require multiple sleep periods.

[0064] In some embodiments, one or more processors may apply continuous stimulation to the subject during subsequent sleep periods based on the final stimulation parameters of the block stimulation. In some embodiments, one or more processors may continue to monitor the subject's slow-wave activity during subsequent sleep periods.

[0065] Figure 6 An example of the correlation between the effects of block stimulation and continuous fixed stimulation is shown. The horizontal axis of graph 600 shows the percentage difference (enhancement) between unstimulated and stimulated slow-wave activity during block stimulation. The vertical axis of graph 600 shows the enhancement caused by continuous fixed stimulation (e.g., as in intelligent sleep therapy systems). Each data point represents a subject, and the position of each data point represents the enhancement of slow-wave activity in the subject due to block stimulation (i.e., the horizontal axis) and the enhancement of slow-wave activity due to continuous stimulation (i.e., the vertical axis). The position of the data points indicates that the enhancement of slow-wave activity due to block stimulation is significantly correlated with the enhancement of slow-wave activity associated with continuous fixed stimulation.

[0066] like Figure 6 As shown, the data points indicate that a threshold percentage enhancement due to block stimuli (e.g., threshold 602) must be satisfied for the object to respond to continuous fixed stimuli. Figure 6 As shown, any data point showing enhanced slow-wave activity for block stimuli below the threshold 602 implies that the corresponding subject is not a responder to continuous fixed stimuli (i.e., the data point on the vertical axis falls below zero). For example, data points 604, 606, 608, 610, and 612 failed to exceed the threshold 602 for enhancement due to block stimuli. Therefore, data points 604, 606, 608, 610, and 612 all have values ​​below zero for enhancement due to continuous fixed stimuli. Data points 614, 616, 618, 620, and 622 all exceed the threshold 602 for enhancement due to block stimuli. Therefore, data points 614, 616, 618, 620, and 622 all have values ​​greater than zero for enhancement due to continuous fixed stimuli. Furthermore, for data points 614, 616, 618, 620, and 622, enhanced slow-wave activity due to block stimuli is significantly correlated with enhanced slow-wave activity associated with continuous fixed stimuli. Figure 600 indicates that when the applied block stimulus causes the enhancement to exceed the threshold, the effect of the block stimulus is similar to that of a continuous fixed stimulus applied with the same parameters.

[0067] Back Figure 1 Electronic storage 22 includes an electronic storage medium that electronically stores information. The electronic storage medium of electronic storage device 22 may include one or both of system storage that is integrally provided with system 10 (i.e., substantially non-removable) and / or removable storage that is removably connected to system 10 via, for example, a port (e.g., USB port, FireWire port, etc.) or a drive (e.g., disk drive, etc.). Electronic storage 22 may include one or more optically readable storage media (e.g., optical disc, etc.), magnetically readable storage media (e.g., magnetic tape, magnetic hard disk drive, floppy disk drive, etc.), charge-based storage media (e.g., EPROM, RAM, etc.), solid-state storage media (e.g., flash memory drive, etc.), cloud storage, and / or other electrically readable storage media. Electronic storage 22 may store software algorithms, information determined by processor 20, information received via object interface 24 and / or external computing systems (e.g., external resource 18), and / or other information that enables system 10 to function as described herein. Electronic storage 22 may be a separate component within system 10 (in whole or in part), or electronic storage 22 may be provided integratedly with one or more other components of system 10 (e.g., processor 20) (in whole or in part).

[0068] Object interface 24 is configured to provide an interface between system 10 and object 12 and / or other objects, through which object 12 and / or other objects can provide and receive information from system 10. This enables data, prompts, results and / or instructions, and any other communicable items (collectively, “information”) to communicate between an object (e.g., object 12) and one or more of sensors 14, sensory stimulators 16, external resources 18, processor 20, and / or other components of system 10. For example, sleep maps, EEG data, deep NREM sleep stage probabilities, and / or other information can be displayed to object 12 or other objects through object interface 24. As another example, object interface 24 may be and / or may 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 may run one or more electronic applications with a graphical object interface configured to provide and / or receive information from the object.

[0069] Examples of interface devices suitable for inclusion in object interface 24 include keypads, buttons, switches, keyboards, knobs, levers, displays, touchscreens, speakers, microphones, indicator lights, audible alarms, printers, haptic feedback devices, and / or other interface devices. In some embodiments, object interface 24 includes multiple separate interfaces. In some embodiments, object interface 24 includes at least one interface provided for integration with other components of processor 20 and / or system 10. In some embodiments, object interface 24 is configured to communicate wirelessly with processor 20 and / or other components of system 10.

[0070] It should be understood that this disclosure also contemplates other communication technologies, whether hardwired or wireless, as object interface 24. For example, this disclosure contemplates that object interface 24 can be integrated with a removable storage interface provided by electronic storage 22. In this example, information can be loaded into system 10 from removable storage (e.g., smart card, flash drive, removable disk, etc.) that enables objects to customize the implementation of system 10. Other exemplary input devices and technologies suitable for use as object interface 24 with system 10 include, but are not limited to, RS-232 ports, RF links, IR links, modems (telephone, cable, or others). In short, any technology used for communicating with system 10 is contemplated as object interface 24 in this disclosure.

[0071] Figure 7A method 700 for delivering sensory stimulation to a subject during sleep is illustrated. The system includes: one or more sensors, one or more sensory stimulators, one or more processors configured by machine-readable instructions, and / or other components. The one or more processors are configured to execute computer program components. The computer program components include information components, model components, control components, modulation components, and / or other components. The operation of method 700 given below is illustrative. In some embodiments, method 700 may be implemented with one or more additional operations not described and / or without using one or more of the operations discussed. Additionally, in Figure 7 The order of operations of method 700 shown and described below is not restrictive.

[0072] In some embodiments, method 700 may be implemented in one or more processing devices such as one or more processors 20 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). The one or more processing devices may include one or more devices that perform some or all of the operations of method 700 in response to instructions stored electronically on an electronic storage medium. The one or more processing devices may include one or more devices configured by hardware, firmware, and / or software that are specifically designed to perform one or more operations of method 700.

[0073] In operation 702, an output signal is generated that conveys information related to the subject's brain activity during sleep periods. The output signal is generated during the subject's sleep periods and / or at other times. In some embodiments, operation 702 is controlled by sensor 14 (in...). Figure 1 (The same or similar sensors shown and described herein) are used.

[0074] In some embodiments, operation 702 includes providing information from the output signal to the neural network in a time set corresponding to various time periods during the sleep period. In some embodiments, operation 710 includes causing the trained neural network to output, based on the time set of information, the depth of NREM sleep detected for the subject during the sleep period. In some embodiments, operation 702 is performed by the model component 32 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0075] In operation 704, sensory stimulation is provided to the subject during the sleep period. The sensory stimulation is applied to the subject in the form of stimulation blocks, each block having intervals between stimuli within each block and intervals between blocks. In some embodiments, in response to determining that the subject is in deep NREM sleep, one or more sensory stimulators are caused to provide sensory stimulation to the subject. In some embodiments, the sensory stimulation may be in the form of auditory vibrations, tactile vibrations, light pulses, and / or other types of sensory stimulation. In some embodiments, operation 704 is controlled by control component 34 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0076] In operation 706, unstimulated slow-wave activity in the subject during the sleep period is detected. In some embodiments, unstimulated slow-wave activity may be slow-wave activity in the time period prior to stimulation (e.g., two seconds before stimulation). In some embodiments, operation 706 is controlled by control component 34 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0077] In operation 708, stimulated slow-wave activity in the subject during sleep periods is detected. Stimulated slow-wave activity includes slow-wave activity during the application of block stimulation. In some embodiments, operation 708 is controlled by control component 34 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0078] In operation 710, stimulated slow-wave activity is compared with unstimulated slow-wave activity. This comparison may include filtering the stimulated slow-wave activity 410 using a frequency band, squaring the filtered data, and / or calculating a moving average over a time period. In some embodiments, unstimulated slow-wave activity may include the average of unstimulated slow-wave activity over a period of time (e.g., two seconds) prior to the application of the sensory stimulus. The comparison may include calculating differences, percentage differences, and / or any other comparison. In some embodiments, operation 710 is controlled by control component 34 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0079] In operation 712, one or more sensory stimulators update the magnitude, timing, inter-block spacing, intra-block spacing, and / or intensity of the sensory stimulus based on a comparison of stimulated slow-wave activity with unstimulated slow-wave activity. One or more sensory stimulators update the stimulation parameters based on one or more brain activity parameters and / or values ​​output from one or more recurrent layers of a trained neural network. In some embodiments, operation 712 is performed with modulation component 36 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0080] In some embodiments, sensory stimulation includes audible tones, tactile vibrations, light pulses, and / or other stimuli. Causing one or more sensory stimulators to update the timing and / or intensity of the sensory stimulation includes adjusting the inter-block interval, intra-block interval, number of stimulations, and / or stimulation volume in response to the detection of deep NREM sleep. In some embodiments, block stimulation is timed to be synchronized with the detection of the rising state of slow waves in the EEG.

[0081] In operation 714, one or more sensory stimulators are controlled based on updated stimulus parameters. One or more sensory stimulators can deliver sensory stimuli to an object using the updated stimulus parameters (i.e., as updated in operation 712). In some embodiments, operation 714 is controlled by control component 34 (in... Figure 1 (shown and described herein) to perform the same or similar processor components.

[0082] In the claims, any reference numerals placed in parentheses should not be construed as limiting the claims. The words "comprising" or "including" do not exclude the presence of elements or steps other than those listed in the claims. In a device claim enumerating multiple means, these means may be implemented by the same hardware. The word "a" or "an" preceding an element does not exclude the presence of multiple such elements. In any device claim enumerating multiple means, these means may be implemented by the same hardware. The fact that certain elements are described in mutually different dependent claims does not mean that these elements cannot be used in combination.

[0083] While the description provided above offers details for illustrative purposes based on embodiments currently considered to be the most practical and preferred, it should be understood that such details are for that purpose only, and that this disclosure is not limited to the explicitly disclosed embodiments, but rather is intended to cover modifications and equivalent arrangements falling within the spirit and scope of the appended claims. For example, it should be understood that this disclosure contemplates, to the extent possible, that one or more features of any embodiment may be combined with one or more features of any other embodiment.

Claims

1. A system (10) for delivering sensory stimuli to a subject (12) during a sleep period, the system comprising: One or more sensors (14) are configured to generate output signals that convey information related to the subject’s brain activity during sleep. One or more sensory stimulators (16) are configured to provide sensory stimulation to the object; as well as One or more processors (20), coupled to the one or more sensors and the one or more sensory stimulators, the one or more processors being configured by machine-readable instructions to: A deep neural network with multiple convolutional layers and multiple recurrent layers is implemented, wherein the deep neural network is configured to detect a deep NREM sleep of the object based on the output signal, wherein the multiple convolutional layers produce multiple convolutional outputs, and the multiple convolutional outputs are fed into the multiple recurrent layers, and The multiple cyclic layers output multiple cyclic outputs, each cyclic output including a predicted probability for each of the multiple individual sleep stages; Based on the output signal, detect unstimulated slow wave activity in the object during the sleep period when no sensory stimulation is provided to the object; The sensory stimulation is provided to the subject during the sleep period; Based on the output signal, stimulated slow-wave activity in the object during the sleep period when the sensory stimulus has been provided to the object; Compare the stimulated slow wave activity with the unstimulated slow wave activity; The stimulation parameters of the sensory stimulus are updated based on the comparison and based on at least one of the convolutional output and the loop output. as well as The one or more sensory stimulators are controlled based on the updated stimulation parameters.

2. The system of claim 1, wherein the stimulus is applied to the object in the form of blocks of repetitive vibrations.

3. The system of claim 2, wherein the blocks are separated from each other by inter-block intervals, and the repetitive vibrations are separated from each other by intra-block intervals, wherein the inter-block intervals are longer than the intra-block intervals.

4. The system of claim 3, wherein the unstimulated slow-wave activity comprises slow-wave activity in the object during the inter-block interval, or wherein the stimulated slow-wave activity comprises slow-wave activity in the object during the repetitive vibration of the blocks.

5. The system of claim 3, wherein, in order to update the stimulation parameters of the sensory stimulus, the one or more processors are further configured to change the duration, intensity, vibration frequency, inter-block interval, or intra-block interval of the stimulus.

6. The system of claim 1, wherein the sensory stimulus comprises: Auditory vibrations, tactile vibrations, or light pulses.

7. The system of claim 1, wherein the comparison comprises determining the difference between the stimulated slow-wave activity and the unstimulated slow-wave activity.

8. A method for delivering sensory stimuli to a subject (12) during a sleep period using a system (10), said system comprising one or more sensors (14), one or more sensory stimulators (16), and one or more processors (20), said sensory stimulators being configured to provide sensory stimuli to the subject, said method comprising: The one or more sensors are used to generate an output signal that conveys information related to the subject's brain activity during the sleep period; The deep NREM sleep of the object is detected using a deep neural network based on the output signal. The deep neural network has multiple convolutional layers and multiple recurrent layers, wherein the multiple convolutional layers produce multiple convolutional outputs, the multiple convolutional outputs are fed into the multiple recurrent layers, and wherein the multiple recurrent layers output multiple recurrent outputs, each recurrent output including a predicted probability for each of multiple individual sleep stages. The one or more processors are used to detect unstimulated slow-wave activity in the object during the sleep period when sensory stimuli are not provided to the object. The sensory stimulation is provided to the subject during the sleep period; The one or more processors are used to detect stimulated slow-wave activity in the object during the sleep period when the sensory stimulus is provided to the object based on the output signal; The stimulated slow-wave activity is compared with the unstimulated slow-wave activity using the one or more processors; The stimulation parameters of the stimulus are updated using the one or more processors based on the comparison and based on at least one of the convolutional output and the loop output; as well as The one or more processors are used to control the one or more sensory stimulators based on the updated stimulation parameters.

9. The method of claim 8, wherein the stimulus is applied to the object in the form of a block of repetitive vibrations.

10. The method of claim 9, wherein the repetitive vibrations are separated from each other by intra-block intervals, wherein the inter-block intervals are longer than the intra-block intervals.

11. The method of claim 10, wherein the unstimulated slow-wave activity comprises slow-wave activity in the object during the inter-block interval, or wherein the stimulated slow-wave activity comprises slow-wave activity in the object during the repetitive vibration of the blocks.

12. The method of claim 10, wherein updating the stimulation parameters of the sensory stimulus comprises changing the duration, intensity, vibration frequency, inter-block interval, or intra-block interval of the stimulus.

13. The method of claim 8, wherein the comparison comprises determining the difference between the stimulated slow-wave activity and the unstimulated slow-wave activity.

14. The method of claim 13, further comprising: Compare the differences with the threshold; In response to determining that the difference has not exceeded the threshold, the stimulation parameters of the stimulus are updated; as well as In response to determining that the difference exceeds the threshold, the stimulation parameters of the stimulus are applied to subsequent sleep periods.

15. The method of claim 14, wherein the threshold is determined based on the minimum difference used to indicate the effectiveness of the stimulus.

Images