Method and system for improving sleep quality based on control of REM stage stimulation

Abstract

The invention provides a method for improving sleep quality based on control of REM stage stimulation, and the method comprises the steps: collecting the objective total sleep duration and subjective total sleep duration of a user, calculating to obtain a sleep error perception index, and determining whether the user is a contradictory insomnia; applying preset REM stimulation parameters to the user when the user enters the REM period and meets safety conditions, and monitoring physiological parameters of the user after being stimulated; the total sleep duration of the user at the night is collected, and the evaluation effect E of the intervention is obtained through calculation based on the evaluation model; when E is greater than 0, the intervention is valid, and when E is less than or equal to 0, the intervention is invalid. According to the method, the sleep error perception index is calculated through the sum of the objective total sleep duration and the subjective total sleep duration, and is used for screening and marking the crowd with the contradictory insomnia, so that intervention resources can be concentrated on the crowd with subjective feeling long-term deviation, and the intervention efficiency is improved.

Description

Technical Field

[0003] This invention belongs to the field of sleep monitoring and intervention technology, specifically relating to a method and system for improving sleep quality based on controlling REM stage stimulation. Background Technology

[0005] Among patients with insomnia, there is a group whose objective sleep duration and efficiency are close to normal, but whose subjective experience is extremely poor; this is often referred to as paradoxical insomnia or sleep state misperception. Even if traditional medications or CBT-I improve PSG / actigraphy indicators, the subjective feeling of "not sleeping well" remains persistent, severely impacting quality of life and treatment adherence. Currently, most sleep aids / apps on the market only focus on objective indicators such as sleep onset delay, number of awakenings, and slow-wave sleep ratio, with virtually no targeted design for addressing long-term deviations in "subjective sleep experience."

[0006] Some existing technologies estimate sleep stages or sleep stability by collecting physiological signals and body movement data, but they cannot, or lack, adjusting the volume and pitch of music / audio / video playback based on stability to improve the sleep experience. The above solutions mainly target environmental volume adjustment before / during sleep onset and a smooth mapping between sleep stability and volume throughout the night. However, they have at least three shortcomings: 1) They do not differentiate between special populations such as ambivalent insomnia: they do not utilize the "objective-subjective sleep difference" for population identification and individualized strategy design; they do not provide targeted intervention for the REM stage: they focus more on overall sleep stability, sleep onset, and wakefulness, rather than modulating dream / emotional processing during REM; 2) They lack a closed-loop optimization based on "subjective sleep experience" as feedback; 3) The adjustment strategies are usually based only on physiological / body movement data, without considering the long-term trend of the patient's morning subjective scores, and cannot truly optimize "misperception of sleep state".

[0007] Therefore, there is an urgent need for a stimulation control method and system that can identify contradictory insomnia characteristics, implement safe and controllable sensory stimulation during the REM stage, and perform long-term adaptive optimization in combination with the user's subjective sleep evaluation, so as to improve the user's subjective perception of their own sleep quality. Summary of the Invention

[0009] To address the problems existing in the prior art, this invention provides a method and system for improving sleep quality based on controlled REM (Rapid Eye Movement) stage stimulation. The aim is to identify users with contradictory insomnia characteristics and improve intervention efficiency by calculating the user's sleep misperception index through multi-night baseline monitoring and questionnaires / subjective ratings. Furthermore, during the intervention, wearable or non-contact devices are used to collect physiological signals and body movement data in real time, identify sleep stages and locate the REM stage. Under preset safety conditions, sound stimuli and / or other sensory stimuli conforming to preset parameters or adaptively updated parameters are output to the user. This improves the user's subjective evaluation of sleep the following day without significantly increasing awakening events. Further, by adopting the types of sound and / or light actively fed back by the user, the stimulation effect can be specifically improved. Even further, by introducing "next morning subjective sleep score" as a key feedback quantity, a long-term adaptive optimization loop is constructed between "REM stimulation parameters—objective sleep structure—subjective perception." Through multi-night iterative updates of the parameter set, the optimal REM stimulation scheme for the individual can be gradually converged, thereby improving subjective sleep experience in a targeted manner with limited changes in objective indicators.

[0010] According to one aspect of the present invention, a method for improving sleep quality based on controlling REM stage stimulation is provided, comprising the following steps:

[0011] 1) Collect sleep monitoring data to obtain the user's objective total sleep duration;

[0012] 2) Collect users' subjective sleep quality evaluation data on the sleep in step 1) to obtain the total subjective sleep duration;

[0013] 3) Based on the objective total sleep duration from step 1) and the subjective total sleep duration from step 2), obtain the sleep misperception index. In the formula, subjective TST is the subjective total sleep duration, and objective TST is the objective total sleep duration; then the |MI| value is compared with the preset threshold to determine whether the user has ambivalent insomnia; if yes, proceed to step 4); if not, end the process.

[0014] 4) Apply preset REM stimulation parameters to the user when the user enters the REM phase and safety conditions are met;

[0015] 5) Monitor the user's physiological parameters after receiving stimulation in step 4); if the monitored physiological parameters exceed the preset threshold, adjust the stimulation parameters or stop the intervention according to the extent of the exceedance; if the preset threshold is not exceeded, stop the intervention when the user's expected REM period transitions to other sleep phases or a predetermined time (e.g., 10-15 minutes) before the expected wake-up.

[0016] 6) Collect the user's total sleep duration for that night, and calculate the evaluation effect of this intervention based on the evaluation model, wherein the evaluation model is... In the formula, w1:w2:w3 = 0.4:0.4:0.2, ΔSE=SE'- The For sleep efficiency after intervention, The average sleep efficiency is defined as the average sleep efficiency over at least two nights prior to the intervention, whereby the sleep efficiency SE = , The MI b The MI' is the average paradoxical sleep index for at least two nights prior to the intervention, and the objective TST is the post-intervention paradoxical sleep index. ΔObjective TST = Objective TST' - The objective TST' refers to the objective total sleep duration after intervention. The average total objective sleep duration was measured over at least two nights prior to the intervention. When E > 0, the intervention was considered effective; when E ≤ 0, there was no significant improvement.

[0017] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 2), the subjective total sleep duration is specifically the total sleep duration subjectively perceived by the user.

[0018] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 3), the preset threshold is ≥30%, more preferably 30%~50%, and for example, it can also be 30%, 35%, 40%, 45%, or 50%.

[0019] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 3), the criteria for determining whether a user has ambivalent insomnia include that the |MI| value exceeds a preset threshold for more than 3 consecutive nights.

[0020] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 4), the REM stage specifically refers to the REM stage of the 2nd to 4th sleep cycle, and the indicators for determining the safety conditions include at least one of the following: heart rate ≤ 110% of the pre-sleep baseline and body acceleration < 0.1g in the last 60 seconds.

[0021] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 4), the preset REM stimulation parameters include sound and / or light stimulation, as well as the corresponding stimulation duration and triggering cycle; wherein the specific type of sound and / or light stimulation is selected according to the user's subjective feedback, the initial value of the stimulation duration is 30 to 60 seconds, the initial value of the triggering cycle is 1 to 2 times per sleep cycle, and the stimulation parameters can be iteratively updated gradually with user feedback.

[0022] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, the initial decibel of the sound is 50~60 and the initial color temperature of the light is 2700~3000K.

[0023] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, in step 5), the threshold conditions include at least one of the following: heart rate > 105% of the pre-sleep baseline, body acceleration > 0.1g, continuous occurrence of EEG α waves > 3s, and respiratory rate change exceeding ±20% of the baseline.

[0024] As a preferred embodiment of the method for improving sleep quality based on controlled REM stage stimulation according to the present invention, in step 5), when the user's heart rate is detected to be 5-10% higher than the initial heart rate of stimulation, the volume of the audio stimulation and / or the brightness of the light stimulation are reduced by 5-10%; when the heart rate increases by 10-20%, the volume and / or brightness are reduced by 10-20%, and the interval between the two stimulations is doubled; when the heart rate increases by more than 20%, the stimulation in the current cycle is stopped, and the stimulation parameters can be iteratively updated according to user feedback. When the user's body movement acceleration is detected to be >0.1g (60s), the current REM stimulation is immediately paused; if no new large-amplitude body movement (acceleration >0.1g) is detected within the following at least 60 seconds, the stimulation cycle is resumed. When the EEG alpha wave is continuously >3 seconds (which may indicate mild arousal), the stimulation output should be immediately turned off until it returns to a level within the threshold condition before the stimulation cycle is resumed.

[0025] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, the method further includes the step of establishing an optimal parameter model, specifically including: obtaining objective sleep data and a sample library of subjective total sleep duration from at least 1000 users and at least 5 nights, and using regression or neural network prediction to predict the parameters of the group with the highest E value as the preset REM stimulation parameters.

[0026] As a preferred embodiment of the method for improving sleep quality based on controlling REM stage stimulation according to the present invention, the method further includes the steps of inputting optimized parameter values ​​into the aforementioned sample library, and removing parameters that are ineffective from the aforementioned sample library.

[0027] As a preferred embodiment of the method for improving sleep quality based on controlled REM stage stimulation according to the present invention, the method further includes the steps of repeating the stimulation when it is determined that there is no significant improvement, and reducing the initial audio volume of the night by 5% to 20% and / or reducing the light stimulation brightness of the night by 5% to 20% compared with the previous night. At the same time, the stimulation interval can be adaptively lengthened (e.g., the number of times is halved, the number of times needs to be rounded down, the minimum is 1 time or the interval time is doubled) until the improvement effect is achieved, or the stimulation is stopped when a preset iteration threshold (e.g., 3 times, 4 times, 5 times, etc.) is reached.

[0028] As a preferred embodiment of the method for improving sleep quality based on controlled REM stage stimulation of the present invention, the method further includes the step of optimizing and executing the stimulation program when it is determined that there is no significant improvement. The step includes: perturbing the stimulation sound of the night within ±5dB of the initial value of the previous night and / or the color temperature of the stimulation light of the night within ±200K of the initial value of the previous night, adjusting the stimulation duration within ±5 seconds, observing the changing trend of the E value in multiple (e.g., 50) perturbation experiments, obtaining the optimal parameter combination under the perturbation experiment, and using it as the initial value of a new round of stimulation.

[0029] According to another aspect of the present invention, a system for improving sleep quality based on controlling REM stage stimulation is provided, comprising:

[0030] The sleep monitoring module is used to collect sleep monitoring data, including a total sleep duration monitoring submodule, a heart rate monitoring submodule, a respiratory rate monitoring submodule, a body motion acceleration monitoring submodule, an EEG alpha wave monitoring submodule, and a bed rest time monitoring submodule.

[0031] A paradoxical insomnia identification module, the identification module includes a subjective total sleep duration input port and a port for receiving sleep monitoring data, the paradoxical insomnia identification module is used to acquire the sleep monitoring data and subjective total sleep duration and determine whether the user has paradoxical insomnia;

[0032] The REM detection module includes a port for receiving instructions from the paradoxical insomnia identification module. The REM detection module is used to detect whether the user has entered the REM phase and whether the safety conditions are met.

[0033] The stimulation control module includes a port for receiving instructions from the REM detection module, and the stimulation control module is used to apply REM phase stimulation to the user;

[0034] The sleep detection module, paradoxical insomnia identification module, REM detection module, and stimulation control module are connected by electricity or electrical signals.

[0035] Studies have shown that applying appropriate external stimuli during REM sleep can influence dream content without causing arousal. For example, providing specific light or sound cues during REM stages (preferably stages 2-4) can induce lucid dreaming, in which the dreamer is aware that they are dreaming and thus has the opportunity to modulate dream content. This suggests that moderate intervention in REM sleep can reconstruct dream content without causing significant sleep disruption, potentially improving patients' subjective experience of sleep.

[0036] Certain oscillation frequencies in the brain during REM sleep are related to functions such as memory consolidation. To address this, this invention delivers sound during the REM phase to influence specific brainwave phases. Notably, the sound stimulation, after precise phase alignment, does not awaken the sleeper. By reconstructing the brainwave rhythm during REM sleep through sound stimulation, REM function is enhanced without disrupting sleep. This supports the intervention principle of this invention from a neuromodulation perspective.

[0037] A key challenge in sleep research is providing interventions without inducing sleep arousal. This invention utilizes safety condition detection (heart rate, body movement, etc.) to ensure that stimulation is applied only when physiologically stable, aiming to avoid stimulation-induced micro-arousals. Appropriate REM stimulation can influence the sleep experience.

[0038] Because patients with ambivalent insomnia have relatively normal objective sleep structure but poor subjective experience, this invention proposes an intervention targeting the REM stage without reducing deep sleep or increasing wakefulness. This intervention is expected to improve subjective sleep quality by enriching dream content or stabilizing the REM physiological state. Moderate auditory and visual stimulation can influence dreams and brain activity without interfering with sleep continuity, potentially correcting the subjective experience of "not sleeping well" in ambivalent insomnia sufferers.

[0039] It should be noted that the techniques used in this invention to monitor parameters such as total sleep duration, heart rate, respiratory rate, body acceleration, and EEG alpha waves can all be achieved through existing monitoring methods.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. This invention calculates the sleep misperception index by combining "objective total sleep duration + subjective total sleep duration" to screen and label people with paradoxical insomnia. This allows intervention resources to be concentrated on people with "long-term subjective perception bias" and improves intervention efficiency.

[0042] 2. This invention implements stimulation during the REM phase, rather than solely based on overall sleep stability. It utilizes the characteristics of the REM phase in relation to dreams, emotional processing, and memory integration to attempt to alter users' emotional experiences and memory representations of sleep, thereby improving users' subjective evaluation of sleep the following day without significantly increasing arousal events.

[0043] 3. By employing the type of sound and / or light that the user actively provides feedback on, this invention can specifically enhance the stimulation effect on the user.

[0044] 4. By employing appropriate stimulation mechanisms, this invention can effectively enhance the stimulation effect. Specifically, stimulation under conditions where the heart rate is not higher than 110% of the baseline before falling asleep can prevent awakening; stimulation under conditions where the average body acceleration in the last 60 seconds is <0.1g can ensure sleep stability; stimulation during the REM phase of the 2nd to 4th sleep cycle can eliminate early unstable REM; and selecting an appropriate number of stimulations can prevent cumulative effects.

[0045] 5. This invention introduces "subjective sleep score the following morning" as a key feedback quantity to construct a long-term adaptive optimization loop among "REM stimulation parameters - objective sleep structure - subjective perception"; and by iteratively updating the parameter set over multiple nights, it can gradually converge to the optimal REM stimulation scheme for the individual, thereby improving the subjective experience of sleep in a targeted manner with limited changes in objective indicators. Detailed Implementation

[0047] The technical solutions described below in conjunction with the embodiments will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] The system used in the method of this invention in specific embodiments includes:

[0049] The sleep monitoring module is used to collect sleep monitoring data, including a total sleep duration monitoring submodule, a heart rate monitoring submodule, a respiratory rate monitoring submodule, a body motion acceleration monitoring submodule, an EEG alpha wave monitoring submodule, and a bed rest time monitoring submodule. Each submodule can be set individually or integrated.

[0050] A paradoxical insomnia identification module, the identification module includes a subjective total sleep duration input port and a port for receiving sleep monitoring data, the paradoxical insomnia identification module is used to acquire the sleep monitoring data and subjective total sleep duration and determine whether the user has paradoxical insomnia;

[0051] The REM detection module includes a port for receiving instructions from the paradoxical insomnia identification module. The REM detection module is used to detect whether the user has entered the REM phase and whether the safety conditions are met.

[0052] The stimulation control module includes a port for receiving instructions from the REM detection module, and the stimulation control module is used to apply REM phase stimulation to the user;

[0053] The sleep detection module, paradoxical insomnia identification module, REM detection module, and stimulation control module are connected by electricity or electrical signals.

[0054] This embodiment provides a specific implementation of the method of the present invention, including the following steps:

[0055] 1) Sleep monitoring data was collected, and the objective total sleep duration of a user for three consecutive nights was 420 minutes, 400 minutes and 390 minutes, respectively, with 470 minutes spent in bed for each night;

[0056] 2) Collect the user's subjective total sleep duration for the three consecutive nights in step 1), which is 280 minutes, 270 minutes and 240 minutes respectively;

[0057] 3) Based on the objective total sleep duration in step 1) and the subjective total sleep duration in step 2), calculate the user's sleep misperception index using the same method described above. In the formula, subjective TST is the subjective total sleep duration, and objective TST is the objective total sleep duration. The user's MI for three consecutive nights was -33.33%, -32.50%, and -38.46%, respectively. The user was judged to have ambivalent insomnia (the MI for three consecutive nights exceeded 30%, which is consistent with the characteristics of ambivalent insomnia).

[0058] 4) When the user enters the REM phase of the 2nd to 4th sleep cycle and meets the safety conditions of heart rate ≤ 110% of pre-sleep baseline and average body acceleration < 0.1g in the last 60 seconds, preset REM stimulation parameters are applied to the user. This embodiment selects initial stimulation parameters using a pre-established sample library and prediction model: A sample library containing at least 1000 different users, each with at least 5 nights of sleep data, is obtained. A neural network algorithm is used to predict the intervention effect E-value of each parameter combination. The parameter combination with the highest E-value from the user group of the same gender and closest age is selected as the user's initial REM stimulation parameters. Specifically, the initial REM stimulation parameters selected in this embodiment are: light stimulation using a warm white light source (color temperature approximately 3000K), sound stimulation using soothing background music (volume approximately 50 decibels), each stimulation lasting 30 seconds, and two stimulations are performed per sleep cycle.

[0059] 5) Monitor changes in the user's physiological parameters after receiving stimulation in step 4). If the user's heart rate increases by 5-10% (excluding)% relative to the initial heart rate, reduce the audio volume and light intensity by 10% each. If the heart rate increases by 10-20% (excluding)%, reduce the volume and light intensity by 20% each, and double the interval between subsequent stimulations. If the heart rate increases by more than 20%, stop stimulation for the remaining time in the current sleep cycle. This embodiment uses the above dynamic adjustment strategy to intervene with REM stimulation until the user's expected REM period transitions to NREM period, at which point the intervention is stopped.

[0060] 6) The following morning, the user's subjective total sleep duration for that night is collected. The evaluation effect of this intervention is calculated based on an evaluation model, wherein the evaluation model is... In the formula, w1:w2:w3 = 0.4:0.4:0.2, ΔSE=SE'- The Sleep efficiency after intervention (90%) The average sleep efficiency (85%) for the three nights prior to the intervention is defined as the sleep efficiency SE = , The MI b The mean paradoxical sleep index (MTS) for the three nights prior to the intervention was -35%, and MI' was the mean paradoxical sleep index (MTS) after the intervention (-26.2%). ΔObjective TST = Objective TST' - The objective TST' refers to the objective total sleep duration after intervention (420 minutes). The average objective total sleep duration (400 minutes) for the three nights prior to the intervention was used. The calculated E after the intervention was 0.4×0.0588+0.4×0.251+0.2×0.05=0.134>0, indicating that the intervention was effective.

[0061] In another specific embodiment of the present invention, a method for improving sleep quality based on controlling REM stage stimulation is provided, comprising the following steps:

[0062] 1) Sleep monitoring data was collected, and the objective total sleep duration of a user for three consecutive nights was 360 minutes, 380 minutes, and 400 minutes, respectively, with a bed rest time of 470 minutes in each night;

[0063] 2) Collect the user's subjective total sleep duration (the user's subjective perception of total sleep duration) for the three consecutive nights of sleep in step 1), which are 240 minutes, 260 minutes and 280 minutes respectively;

[0064] 3) Calculate the Misperception Index (MI) based on the objective total sleep duration from step 1) and the subjective total sleep duration from step 2). Using the user's data, the calculated MI values ​​for each night are approximately -33.33%, -31.58%, and -30.00%. Then, the |MI| values ​​are compared with a preset threshold of 30%. The results show that the user's |MI| exceeded 30% for three consecutive nights, thus confirming that the user suffers from ambivalent insomnia.

[0065] 4) When the user enters the REM phase of the 2nd to 4th sleep cycle and their physiological state meets the safety conditions of heart rate ≤ 110% of pre-sleep baseline and average body acceleration < 0.1g in the last 60 seconds, the preset REM stimulation parameters are applied to the user. Based on the user's feedback preferences, the preset REM stimulation parameters are set as follows: light stimulation uses a warm-toned soft light source (color temperature approximately 2800K), and sound stimulation uses soft background music (volume approximately 55 decibels). Each stimulation lasts 30 seconds and is performed once during the REM phase of each sleep cycle;

[0066] 5) During the application of REM stimulation to the user, the changes in the user's physiological parameters were monitored in real time. In this embodiment, the user's physiological parameters did not exceed the preset threshold conditions during the stimulation process (heart rate was always ≤ 105% of the initial heart rate of stimulation, body acceleration ≤ 0.1g, no EEG α wave lasting more than 3 seconds, and respiratory rate changes were kept within ±20% of the baseline), indicating that the stimulation did not cause an abnormal response. The intervention was stopped 10 minutes before the expected awakening.

[0067] 6) The following morning, the user's subjective total sleep duration for that night is collected. The evaluation effect of this intervention is calculated based on an evaluation model, wherein the evaluation model is... In the formula, w1:w2:w3 = 0.4:0.4:0.2, ΔSE=SE'- The Sleep efficiency after intervention (90%) The average sleep efficiency (81%) for the three nights prior to the intervention was defined as SE = , The MI b The mean paradoxical sleep index (MTS) for the three nights prior to the intervention was -31.6%, and MI' was the mean paradoxical sleep index (MTS) after the intervention (-26.2%). ΔObjective TST = Objective TST' - The objective TST' refers to the objective total sleep duration after intervention (400 minutes). The average objective total sleep duration (380 minutes) for the three nights prior to the intervention was used. The calculated E after the intervention was 0.4×0.09 + 0.4×0.054 + 0.2×0.0833 = 0.036 + 0.0216 + 0.01666 = 0.07426 > 0, indicating that the intervention was effective.

[0068] It should be noted that, based on the above embodiments of the present invention, those skilled in the art can fully realize the scope of the independent claims and dependent claims of the present invention, and the implementation process and methods are the same as those in the above embodiments; and the parts of the present invention not described in detail belong to the well-known technology in the art. However, the protection scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A method for improving sleep quality based on controlling REM stage stimulation, characterized in that, Includes the following steps: 1) Collect sleep monitoring data to obtain the user's objective total sleep duration; 2) Collect users' subjective sleep quality evaluation data on the sleep in step 1) to obtain the total subjective sleep duration; 3) Based on the objective total sleep duration from step 1) and the subjective total sleep duration from step 2), obtain the sleep misperception index. In the formula, subjective TST is the subjective total sleep duration, and objective TST is the objective total sleep duration; then the |MI| value is compared with the preset threshold to determine whether the user has ambivalent insomnia; if yes, proceed to step 4); if not, end the process. 4) Apply preset REM stimulation parameters to the user when the user enters the REM phase and safety conditions are met; 5) Monitor the user's physiological parameters after receiving stimulation in step 4); if the monitored physiological parameters exceed the preset threshold, adjust the stimulation parameters or stop the intervention according to the extent of the exceedance; if the preset threshold is not exceeded, stop the intervention when the user's expected REM period transitions to other sleep phases or a predetermined time before expected wake-up. 6) Collect the user's total sleep duration for that night, and calculate the evaluation effect of this intervention based on the evaluation model, wherein the evaluation model is... In the formula, w1:w2:w3 = 0.4:0.4:0.2, ΔSE=SE'- The For sleep efficiency after intervention, The average sleep efficiency is defined as the average sleep efficiency over at least two nights prior to the intervention, whereby the sleep efficiency SE = , The MI b The MI' is the average paradoxical sleep index for at least two nights prior to the intervention, and the objective TST is the post-intervention paradoxical sleep index. ΔObjective TST = Objective TST' - The objective TST' refers to the objective total sleep duration after intervention. The average total objective sleep duration was measured over at least two nights prior to the intervention. When E > 0, the intervention was considered effective; when E ≤ 0, there was no significant improvement.

2. The method for improving sleep quality based on controlling REM stage stimulation as described in claim 1, characterized in that, In step 2) and in step 3), the preset threshold is ≥30%.

3. The method for improving sleep quality based on controlling REM stage stimulation as described in claim 1, characterized in that, In step 3), the criteria for determining whether a user has ambivalent insomnia include having a |MI| value exceeding a preset threshold for more than 3 consecutive nights.

4. A method for improving sleep quality based on controlling REM stage stimulation as described in any one of claims 1-3, characterized in that, In step 4), the REM period specifically refers to the REM period of the 2nd to 4th sleep cycles, and the indicators for determining the safety conditions include at least one of the following: heart rate ≤ 110% of the pre-sleep baseline and body acceleration < 0.1g in the last 60 seconds.

5. A method for improving sleep quality based on controlling REM stage stimulation as described in any one of claims 1-3, characterized in that, In step 4), the preset REM stimulation parameters include sound and / or light stimulation, as well as the corresponding stimulation duration and triggering cycle; wherein the initial parameters of sound and / or light can be any of the following: 1) subjective feedback from the user, 2) the initial decibel of sound is 50~60db, the initial color temperature of light is 2700~3000K, the initial value of the stimulation duration is 30~60 seconds, and the initial value of the triggering cycle is 1~2 times per sleep cycle.

6. The method for improving sleep quality based on controlling REM stage stimulation as described in claim 5, characterized in that, When E≤0, the initial audio volume and / or light stimulation brightness are reduced by 5% to 20% compared to the previous time and the stimulation is reapplied until E>0.

7. A method for improving sleep quality based on controlling REM stage stimulation as described in any one of claims 1-3 and 6, characterized in that, In step 5), the threshold conditions include at least one of the following: heart rate > 105% of the pre-sleep baseline, body acceleration > 0.1g, continuous EEG α waves > 3s, and respiratory rate change exceeding ±20% of the baseline.

8. The method for improving sleep quality based on controlling REM stage stimulation as described in claim 7, characterized in that, In step 5), when the user's heart rate is detected to be 5-10% higher than the initial heart rate of the stimulus, the volume of the audio stimulus and / or the brightness of the light stimulus are reduced by 5-10%; when the heart rate increases by 10-20%, the volume and / or brightness are reduced by 10-20%, and the interval between the two stimuli is doubled; when the heart rate increases by more than 20%, the stimulation in the current cycle is stopped; when the user's body acceleration is detected to be >0.1g, the current REM stimulation is immediately paused, and if no new body movement >0.1g is detected within the following at least 60 seconds, the stimulation cycle is resumed; when the user continuously exhibits EEG α waves for >3 seconds, the stimulation output is turned off until it returns to a level within the threshold condition, and then the stimulation cycle is resumed.

9. A method for improving sleep quality based on controlling REM stage stimulation as described in any one of claims 1-3, 6, and 8, characterized in that, The method also includes the step of establishing an optimal parameter model, specifically including: obtaining objective sleep data and a sample library of subjective total sleep duration from at least 1,000 users and at least 5 nights, using regression or neural network prediction to predict, and obtaining the parameters of the group with the highest E value as the preset REM stimulation parameters.

10. A system for implementing the method as described in any one of claims 1-9, characterized in that, include: The sleep monitoring module is used to collect sleep monitoring data, including a total sleep duration monitoring submodule, a heart rate monitoring submodule, a respiratory rate monitoring submodule, a body motion acceleration monitoring submodule, an EEG alpha wave monitoring submodule, and a bed rest time monitoring submodule. A paradoxical insomnia identification module, the identification module includes a subjective total sleep duration input port and a port for receiving sleep monitoring data, the paradoxical insomnia identification module is used to acquire the sleep monitoring data and subjective total sleep duration and determine whether the user has paradoxical insomnia; The REM detection module includes a port for receiving instructions from the paradoxical insomnia identification module. The REM detection module is used to detect whether the user has entered the REM phase and whether the safety conditions are met. The stimulation control module includes a port for receiving instructions from the REM detection module, and the stimulation control module is used to apply REM phase stimulation to the user; The sleep detection module, paradoxical insomnia identification module, REM detection module, and stimulation control module are connected by electricity or electrical signals.

Images