An effective sleep state recognition method and system fusing sleep staging and drowsiness features, a terminal and a medium

Abstract

This invention discloses an effective sleep state identification method, system, terminal, and medium that integrates sleep stages and drowsiness characteristics. The method includes: during a user's sleep, collecting the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals using a wearable device. The EEG signals reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals reflect changes in sleep posture and the number of times the user turns over. For valid signal segments, based on the EEG signals and the IMU signals, determining the percentage of an effective resting state and the percentage of a highly active, awake state; when the percentage of an effective resting state reaches a first preset threshold and the percentage of a highly active, awake state is less than a second preset threshold, the current sleep state is determined to be an effective sleep state. This invention can accurately identify effective sleep states, avoiding missed or false positives.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface technology, and in particular to an effective method, system, terminal, and medium for sleep state recognition that integrates sleep stages and drowsiness characteristics. Background Technology

[0002] At present, sleep state recognition technology is mainly based on the analysis of EEG data, but it generally has the following technical defects: (1) Relying solely on sleep stages, the misjudgment rate is high: Relying solely on sleep stages to judge the rest state has obvious limitations. For example, when a user is awake but closes their eyes and relaxes, the EEG signal may show characteristics similar to light sleep. If only the stages are used, it will be misjudged as having entered a rest state.

[0003] (2) Lack of dedicated identification logic for “effective rest”: Existing systems either regard any non-awake state as “sleep” or regard all waking states as “awake”, without effectively identifying the “quasi-rest” state that is actually due to high drowsiness and low body movement in the waking state, which can easily lead to misjudgment.

[0004] Therefore, existing technologies still need improvement. Summary of the Invention

[0005] To address the aforementioned deficiencies in existing technologies, this invention provides an effective sleep state identification method, system, terminal, and medium that integrates sleep stages and drowsiness characteristics. The technical solution adopted by this invention is as follows: In a first aspect, the present invention provides an effective sleep state identification method that integrates sleep stages and drowsiness characteristics, the method comprising: The system collects the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals using wearable devices. The EEG signals are used to reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals are used to reflect changes in sleep posture and the number of times the user turns over. For valid signal segments, based on the EEG signal and the inertial measurement unit signal, determine the percentage of valid resting state and the percentage of awake and highly active state; When the percentage of the effective rest state reaches a first preset threshold and the percentage of the awake and highly active state is less than a second preset threshold, the current sleep is determined to be an effective sleep state.

[0006] In one implementation, the method further includes: The signal quality of the EEG signals and inertial measurement unit signals collected in each time segment is judged to determine the effective signal segments. In the effective signal segments, the electrode contact state of the EEG signals is stable.

[0007] In one implementation, for valid signal segments, based on the EEG signal and the inertial measurement unit signal, the percentage of valid resting state and the percentage of awake, highly active state are determined, including: Based on the EEG signals, the ratio between effective EEG signal segments and total EEG signal segments is calculated to obtain the signal integrity. The device usage data, device status, and device usage duration of the wearable device are obtained, and verification is performed based on the device usage data, device status, device usage duration, and signal integrity. Once the verification is successful, the percentage of effective resting state and the percentage of awake and highly active state are determined based on the EEG signal and the inertial measurement unit signal.

[0008] In one implementation, after the verification is passed, based on the EEG signal and the inertial measurement unit signal, the percentage of effective resting state and the percentage of awake and highly active state are determined, including: Based on the electroencephalogram (EEG) signals, the electrode contact status, sleep stages, and drowsiness index are determined. Based on the signals from the inertial measurement unit, the number of times the user turns over and the changes in sleep posture are determined; Obtain effective rest signal segments for the sleep stages of light sleep, deep sleep, and rapid eye movement; or, obtain effective rest signal segments for the sleep stage of wakefulness where the drowsiness index reaches a preset range, the number of times the person turns over within a preset time period is less than a first preset number, the number of sleep posture changes is less than a second preset number, and the electrode contact state is stable. The effective rest state percentage is obtained based on the ratio between the effective rest signal segment and the total signal segment. The system acquires a high-activity wakefulness signal segment when the sleep stage is a waking state and the drowsiness index is less than a fourth preset threshold; or, it acquires a high-activity wakefulness signal segment when the sleep stage is a waking state and the number of times the sleeper turns over and the number of sleep posture changes within a preset time period are both greater than a third preset number. The proportion of the awake and highly active state is obtained based on the ratio between the awake and highly active signal segments and the total signal segments.

[0009] In one implementation, the method further includes: After identifying effective sleep states, a rest effectiveness score is calculated based on the effective rest duration coefficient, the proportion of effective rest states, and the rest proximity coefficient. The rest proximity coefficient is calculated based on the drowsiness index. Awakening alertness score is calculated based on wake-up completion coefficient, wake-up timing coefficient, and post-wake stability coefficient. The reliability score is calculated based on signal integrity, contact state stability, body motion coefficient, and equipment continuous use coefficient. The final sleep score is obtained based on the rest effectiveness score, wakefulness score, and process credibility score.

[0010] In one implementation, the method further includes: Obtain a preset evaluation text library, which contains evaluation tags corresponding to different sleep score ranges; The final sleep score is matched with the evaluation text library to determine the target evaluation tag corresponding to the final sleep score; Obtain the explanatory text of the sleep state corresponding to the target evaluation label to obtain the sleep state evaluation result.

[0011] In one implementation, the method further includes: Based on the sleep state evaluation results, a sleep aid plan and a wake-up strategy are generated; When the user goes to sleep next time, the sleep aid audio is adaptively adjusted based on the sleep aid scheme, and the wake-up audio is adaptively adjusted based on the wake-up strategy.

[0012] Secondly, embodiments of the present invention also provide an effective sleep state recognition system that integrates sleep stages and drowsiness characteristics, wherein the system is used to implement the steps of the effective sleep state recognition method that integrates sleep stages and drowsiness characteristics described in any of the above claims, and the system includes: A multi-dimensional data acquisition module is used to collect the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals based on a wearable device during sleep. The EEG signals are used to reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals are used to reflect changes in sleep posture and the number of times the user turns over. The multi-dimensional signal analysis module is used to determine the percentage of effective resting state and the percentage of awake and highly active state based on the EEG signal and the inertial measurement unit signal for effective signal segments. The sleep effectiveness determination module is used to determine the current sleep as an effective sleep state when the proportion of the effective rest state reaches a first preset threshold and the proportion of the awake and highly active state is less than a second preset threshold.

[0013] Thirdly, embodiments of the present invention also provide a terminal, wherein the terminal includes a memory, a processor, and an effective sleep state recognition program that integrates sleep stages and drowsiness features stored in the memory and can run on the processor. When the processor executes the effective sleep state recognition program that integrates sleep stages and drowsiness features, it implements the steps of the effective sleep state recognition method that integrates sleep stages and drowsiness features of any of the above-mentioned schemes.

[0014] Fourthly, embodiments of the present invention also provide a computer-readable storage medium, wherein the computer-readable storage medium stores an effective sleep state recognition program that integrates sleep stages and drowsiness features, and the effective sleep state recognition program that integrates sleep stages and drowsiness features implements the steps of the effective sleep state recognition method that integrates sleep stages and drowsiness features as described in any of the above schemes on the computer-readable storage medium.

[0015] Beneficial Effects: Compared with existing technologies, this invention provides an effective sleep state identification method that integrates sleep stages and drowsiness characteristics. First, during the user's sleep, wearable devices collect the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals. The EEG signals reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals reflect changes in sleep posture and the number of times the user turns over. Next, for valid signal segments, based on the EEG and IMU signals, the proportion of an effective rest state and the proportion of a highly active, awake state are determined. Then, when the proportion of an effective rest state reaches a first preset threshold and the proportion of a highly active, awake state is less than a second preset threshold, the current sleep is determined to be an effective sleep state. This invention can accurately identify whether a user is in an effective sleep state in short-term sleep scenarios, avoiding misjudging closed-eye wakefulness as an effective sleep state or light sleep as an ineffective sleep state. Attached Figure Description

[0016] Figure 1 This is a flowchart of a preferred embodiment of the effective sleep state identification method that integrates sleep stages and drowsiness features according to an embodiment of the present invention.

[0017] Figure 2 This is a technical framework diagram of an effective sleep state recognition system that integrates sleep stages and drowsiness features according to an embodiment of the present invention.

[0018] Figure 3 A schematic diagram of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content, operations, or steps, nor does it require execution in the described order. For example, some operations or steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0021] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. For example, the first control information and the second control information are only used to distinguish different control information and do not limit their order.

[0023] Those skilled in the art will understand that the words "first" and "second" do not limit the quantity or the order of execution, and that the words "first" and "second" do not necessarily imply that they are different.

[0024] It should also be understood that the term “and / or” as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] The method in this embodiment can be applied to a terminal, which can be an intelligent electronic device such as a computer. Figure 1 As shown in the figure, the effective sleep state identification method that integrates sleep stages and drowsiness features in this embodiment specifically includes the following steps: Step S100: Collect the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals based on the wearable device, wherein the EEG signals are used to reflect electrode contact status, sleep stages, and drowsiness index, and the IMU signals are used to reflect changes in sleep posture and the number of times the user turns over.

[0026] During their afternoon nap, users wear wearable devices that collect multi-dimensional physiological signals, including electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals. The EEG signals reflect electrode contact status, sleep stages, and drowsiness index. Sleep stages specifically include Light (light sleep), Deep (deep sleep), and REM (rapid eye movement) sleep. The drowsiness index is an objective quantitative indicator derived from EEG signal analysis. It reflects a person's physiological tendency or degree of transition from wakefulness to sleep, and is academically often referred to as the Drowsiness Index or Fatigue Index. The IMU signals reflect changes in sleep posture and the number of times the user turns over during sleep.

[0027] In this embodiment, after acquiring device data, the acquired EEG and inertial measurement unit (IMU) signals are further segmented and preprocessed to determine signal quality. For example, the acquired EEG and IMU signals are segmented into uniform signal segments within a fixed time window (e.g., 1 minute). Then, the intensity of the EEG and IMU signals is filtered to remove weak signals, thereby selecting valid signal segments. Alternatively, this embodiment can also determine the electrode contact state during EEG signal acquisition. If the electrode contact state is stable, the acquired EEG signal is considered valid. In other implementations, the wearable device in this embodiment can also acquire photoplethysmography (PPG) data to assist in verifying the signal quality and contact state of the data.

[0028] Step S200: For the effective signal segments, based on the EEG signal and the inertial measurement unit signal, determine the percentage of effective resting state and the percentage of awake and highly active state.

[0029] Next, for the effective signal segments, this embodiment can combine EEG signals and / or the inertial measurement unit signals to determine the proportion of effective rest states and the proportion of awake and highly active states. In this embodiment, the proportion of effective rest states is the ratio between effective rest signal segments and the total number of signal segments. Effective rest segments are signal segments in the sleep stages of light sleep, deep sleep, and REM sleep, or, when the sleep stage is awake, signal segments with a drowsiness index reaching a preset range and exhibiting low body movement. The proportion of awake and highly active states is the ratio between awake and highly active signal segments and the total number of signal segments. Awakening and highly active signal segments are awake and highly active signal segments in the sleep stage of awake state where the drowsiness index is less than a fourth preset threshold, or, when the sleep stage is awake state, signal segments exhibiting significant body movement.

[0030] In practical applications, a dual-threshold verification mechanism is established to identify the valid sleep state in this embodiment. This mechanism includes a valid reporting threshold and a valid sleep threshold. In actual applications, the reliability of the collected data is first determined by verifying the valid reporting threshold, and then the valid sleep threshold is verified to confirm the actual existence of the sleep state. This effectively distinguishes between two scenarios: wearing a wearable device but not resting and truly entering a resting state, thus avoiding misjudgment of valid sleep.

[0031] Specifically, this embodiment establishes a valid reporting threshold verification mechanism based on the wearable device's usage data, device status, usage duration, and signal integrity. The valid reporting threshold is defined as follows: the device usage data reflects that the wearable device has completed a full usage cycle; the device status reflects that the wearable device is functioning normally; the usage duration reaches a preset duration; and the signal integrity reaches a third preset threshold. This embodiment also establishes a valid sleep threshold verification mechanism based on the user's effective rest state percentage and awake / high-activity state percentage. The valid sleep threshold is defined as follows: the effective rest state percentage reaches a first preset threshold, and the awake / high-activity state percentage is less than a second preset threshold. The effective rest state percentage is the ratio of effective rest signal segments to total signal segments; the awake / high-activity state percentage is the ratio of awake / high-activity signal segments to total signal segments. During dual threshold verification, the valid reporting threshold is verified first. Only after the valid reporting threshold verification passes is the valid sleep threshold verification performed. Only after both threshold verifications pass is the user's valid sleep state identified.

[0032] Specifically, in verifying the effective reporting threshold, this embodiment first filters the EEG signal for signal quality and, in conjunction with the stability of electrode contact, filters out effective EEG signal segments. Then, it calculates the ratio between effective EEG signal segments and total EEG signal segments to obtain the signal integrity. Next, it analyzes the wearable device's usage data, device status, and device usage duration. When the device usage data reflects that the wearable device has completed a full use, the device status reflects that the wearable device is functioning normally, the device usage duration reaches a preset duration, and the signal integrity reaches a first preset threshold (e.g., signal integrity ≥ 60%), the effective reporting threshold verification is considered passed. In other implementations, if the signal integrity is lower than the first preset threshold, the device usage data, device status, and device usage duration are not analyzed, and subsequent effective sleep threshold verification and sleep effect evaluation are not performed; only the collected data is recorded. This means that even if the user is in a clear resting state, if there are insufficient effective EEG segments, the subsequent sleep effect evaluation process will not be triggered.

[0033] Furthermore, after the valid reporting threshold is verified, the valid sleep threshold is verified. This embodiment determines the percentage of effective rest and the percentage of awake, highly active states based on the EEG signals and the inertial measurement unit signals. Specifically, this embodiment first determines the electrode contact status, sleep stage, and drowsiness index based on the EEG signals. When calculating the drowsiness index, this embodiment extracts key EEG features related to drowsiness and fatigue from the collected EEG signals, including the EEG power spectrum and specific frequency band power wall ratios, such as the β / (θ+α) ratio, (θ+α) / β ratio, and θ / α ratio. β waves are associated with wakefulness and alertness, while α / θ waves are associated with relaxation and drowsiness. Therefore, an increased β / (θ+α) ratio may indicate greater alertness, while a decreased ratio may indicate greater fatigue. An increased (θ+α) / β ratio is generally associated with increased fatigue; furthermore, increased θ wave activity and decreased α wave activity are also considered markers of drowsiness. Therefore, by fusing one or more of the extracted features using a specific algorithm model (such as optimization algorithms, neural networks, etc.), a drowsiness index that changes over time is finally calculated. Next, based on the inertial measurement unit signal, this embodiment can determine the number of times the user turns over and the changes in sleep posture. Then, this embodiment obtains effective rest signal segments for the sleep stages of light sleep, deep sleep, and REM sleep, or, when the sleep stage is wakefulness, the drowsiness index reaches a preset range (e.g., 700-100), and the number of turns over within a preset time period is less than a first preset number (e.g., ≤1 time), and the changes in sleep posture are less than a second preset number (e.g., ≤2 times), and the electrode contact state is stable. Therefore, the percentage of effective rest states can be obtained based on the ratio between the effective rest signal segments and the total signal segments.

[0034] When determining the highly active awake signal segment, this embodiment can acquire a highly active awake signal segment when the sleep stage is a waking state and the drowsiness index is less than a fourth preset threshold (e.g., <50). Alternatively, it can acquire a highly active awake signal segment when the sleep stage is a waking state and the number of times the user turns over and the changes in sleep posture within a preset time period are both greater than a third preset number (e.g., both >2 times). In this case, it indicates that the user has not entered a resting state. Therefore, the proportion of the highly active awake state can be obtained based on the ratio between the highly active awake signal segments and the total number of signal segments.

[0035] Step S300: When the proportion of the effective rest state reaches a first preset threshold and the proportion of the awake and highly active state is less than a second preset threshold, the current sleep is determined to be an effective sleep state.

[0036] In this embodiment, the effective sleep threshold verification is considered passed when the percentage of the effective rest state reaches a first preset threshold and the percentage of the awake and highly active state is less than a second preset threshold. For example, when the effective rest state is ≥20% and the percentage of the awake and highly active state is less than 80%, the effective sleep threshold verification is passed, and the user's effective sleep state is identified. If the percentage of the effective rest state does not reach the first preset threshold, or the percentage of the awake and highly active state is greater than the second preset threshold, it means that no effective sleep state has been detected.

[0037] Therefore, this embodiment can analyze the proportion of effective rest state and the proportion of awake and highly active state based on sleep stages and drowsiness characteristics, and then identify the effective sleep state based on the proportion of effective rest state and the proportion of awake and highly active state. Compared with the existing technology that relies solely on sleep stages or solely on drowsiness index, this embodiment can significantly improve the accuracy of identifying the effective rest state in short sleep scenarios.

[0038] Furthermore, after identifying an effective rest state, this embodiment can evaluate the sleep rest state to obtain a sleep state evaluation result. This embodiment determines the sleep state evaluation result by calculating a final sleep score, which is based on a rest effectiveness score, an arousal and wakefulness score, and a process reliability score.

[0039] Specifically, this embodiment first calculates the rest effectiveness score based on the effective rest duration coefficient, the percentage of effective rest states, the rest proximity coefficient, and their respective weights. In practical applications, the effective rest duration coefficient is determined based on the user's specific effective rest minutes. For example, if the effective rest minutes are less than 10 minutes, the effective rest duration coefficient is 0; if the effective rest minutes are 10-20 minutes, the effective rest duration coefficient = (effective rest minutes - 10) / 10; if the effective rest minutes are 20-30 minutes, the effective rest duration coefficient is 1; if the effective rest minutes are 30-45 minutes, the effective rest duration coefficient = 1 - (effective rest minutes - 30) / 30; and if the effective rest minutes are ≥45 minutes, the effective rest duration coefficient is 0.5. The rest proximity coefficient is calculated as: (average drowsiness index in the second half - average drowsiness index in the first 5 minutes + 20) / 40. Preferably, the rest proximity coefficient in this embodiment does not exceed 0.1. After calculating the effective rest duration coefficient, the effective rest state percentage, and the rest approach coefficient, this embodiment can obtain the weights corresponding to the effective rest duration coefficient, the effective rest state percentage, and the rest approach coefficient, and perform a weighted summation to obtain the rest effectiveness score. For example, the calculation formula is: Rest effectiveness score = 0.45 × effective rest duration coefficient + 0.40 × rest state percentage + 0.15 × rest approach coefficient.

[0040] Next, this embodiment calculates the wake-up alertness score based on the wake-up completion coefficient, wake-up timing coefficient, and post-wake stability coefficient. Specifically, the wake-up completion coefficient is related to whether the user is awake. When the user has been woken up, or the wake-up function of the wearable device (such as an alarm clock or music playback) is triggered, and the user actively removes the wearable device after the wake-up function has been completed, the wake-up completion coefficient is 1. If the wake-up function of the wearable device is triggered but prematurely terminated, and the user is already awake, the wake-up completion coefficient is 0.8. If the wake-up function is executed, but the user is not woken up, the wake-up completion coefficient is 0.5. If the wake-up function of the wearable device is not triggered, the wake-up completion coefficient is 0. The wake-up timing coefficient in this embodiment is related to the proportion of wake-up risk segments in the three minutes before wake-up. The three minutes before wake-up refers to the three-minute time window before the wake-up function is triggered. The proportion of wake-up risk segments = the number of segments identified as "wake-up risk segments" in the three minutes before wake-up ÷ the total number of segments in the three minutes before wake-up (calculated in a 1-minute window, a total of 3 segments). In this embodiment, the wake-up risk segments are derived from deep sleep segments, segments with significant jumps in drowsiness index, and segments with obvious body movement. In practical applications, the wake-up timing coefficient can be set as shown in Table 1.

[0041] Table 1

[0042] In this embodiment, the post-awakening stability coefficient is calculated as: (Number of stable post-awakening segments within 2 minutes of awakening) ÷ (Number of effective post-awakening signal segments). The core logic of this post-awakening stability coefficient is that within 2 minutes of awakening, the higher the proportion of usable effective signals that are judged as "stable," the closer the coefficient is to 1, indicating a smoother transition from rest to wakefulness. The number of effective post-awakening signal segments refers to the number of segments within these 2 minutes where the EEG signal quality is normal when the EEG electrodes are in contact with the skin. The number of stable post-awakening segments must simultaneously meet the following three conditions: the sleep stage is awake, the drowsiness index is less than 50 or the drowsiness index has decreased by ≥20 compared to before awakening, and there is no continuous obvious body movement. No continuous obvious body movement means that no sustained body movement occurs within this 1-minute segment; occasional single turning over or minor posture adjustments do not constitute continuous obvious body movement. Therefore, the post-awakening stability coefficient can be obtained based on the ratio between the number of stable post-awakening segments and the number of effective post-awakening signal segments within 2 minutes of awakening. In practical applications, the value of the post-awakening stability coefficient can be shown in Table 2.

[0043] Table 2

[0044] After calculating the wake-up completion coefficient, wake-up timing coefficient, and post-wake stability coefficient, this embodiment can perform a weighted sum based on the wake-up completion coefficient, wake-up timing coefficient, and post-wake stability coefficient to obtain the wake-up alertness score. The calculation formula for the wake-up alertness score is: Wake-up alertness score = 0.40 × Wake-up completion coefficient + 0.35 × Wake-up timing coefficient + 0.25 × Post-wake stability coefficient.

[0045] Furthermore, this embodiment can calculate the reliability score based on signal integrity, contact stability, low body movement coefficient, and continuous equipment use coefficient. Specifically, signal integrity = effective EEG signal segments ÷ total EEG signal segments. Effective EEG signal segments refer to signal segments where the EEG electrodes are in contact with the skin and the signal quality is normal. Contact stability = 1 Number of failed contact segments ÷ Total number of EEG signal segments. The number of failed contact segments refers to EEG electrodes not making contact with the skin or continuous abnormal EEG signals, such as consecutive excessively large or small EEG signals. Low body movement coefficient = 1 The formula `clamp(obvious body movement segment percentage / 20%, 0, 1)` means that `x` is limited to between 0 and 1, with values ​​less than 0 counted as 0 and values ​​greater than 1 counted as 1. In practical applications, low body movement refers to turning over ≤1 times and changing posture ≤2 times per minute. Obvious body movement refers to turning over ≥2 times or changing posture ≥3 times per minute. Obvious body movement segment percentage - number of signal segments with obvious body movement ÷ total number of EEG signal segments. Device continuous use coefficient = 1 - `clamp((number of device pauses × 2 + number of minutes the device was actively removed) ÷ total number of minutes the device was used, 0, 1)`. After calculating the signal integrity, contact stability, low body movement coefficient, and device continuous use coefficient, this embodiment can perform a weighted sum based on the signal integrity, contact stability, low body movement coefficient, and device continuous use coefficient to obtain the process reliability score. In practical applications, the process reliability score can be calculated as follows: Process reliability score = 0.40 × signal integrity + 0.30 × contact stability + 0.20 × low body movement coefficient + 0.10 × continuous conversation coefficient.

[0046] In other implementations, this embodiment can also collect photoplethysmography (PPG) signals, which reflect the user's blood oxygenation and heart rate during sleep. Then, the EEG signals, inertial unit signals, and PPG signals are combined to comprehensively analyze the wearability of the wearable device, accurately identify electrode connection status, reduce the false positive rate of contact stability, and improve the accuracy of process reliability calculation. Further, after calculating the rest effectiveness score, wakefulness score, and process reliability score, this embodiment can perform a weighted sum to obtain the final sleep score. The weight of the rest effectiveness score is 50, the weight of the wakefulness score is 25, and the weight of the process reliability score is 25. The final sleep score is calculated as follows: Final sleep score = 50 × rest effectiveness score + 25 × wakefulness score + 25 × process reliability score.

[0047] Furthermore, in this embodiment, after obtaining the final sleep score, the final sleep score is mapped to a sleep state evaluation result. In practical applications, this embodiment can obtain a preset evaluation text library, which contains evaluation tags corresponding to different sleep score ranges. This embodiment matches the final sleep score with the evaluation text library to determine the target evaluation tag corresponding to the final sleep score. Then, the sleep state explanatory text corresponding to the target evaluation tag is obtained to obtain the sleep state evaluation result. For example, when the final sleep score is in the score range of 80-100, the obtained sleep state explanatory text reflects that the sleep state evaluation result is "very good sleep effect". When the final sleep score is in the score range of 70-80, the obtained sleep state explanatory text reflects that the sleep state evaluation result is "good sleep effect". When the final sleep score is in the score range of 60-70, the obtained sleep state explanatory text reflects that the sleep state evaluation result is "average sleep effect". If the final sleep score is between 50 and 60, the explanatory text for the sleep state indicates that the sleep state needs improvement. If the final sleep score is between 0 and 50, the explanatory text for the sleep state indicates that effective rest was not achieved.

[0048] Furthermore, in other implementations, this embodiment can also perform a brain-charging visual representation of the final sleep score to obtain brain-charging visual representation information. Similarly, the brain-charging visual representation information can be determined based on the numerical range of the final sleep score. The higher the final sleep score, the better the sleep state evaluation information, and the more fully the brain rests; therefore, the more positive the brain-charging visual representation information will be. Finally, this embodiment can also visualize the final sleep score, the sleep state evaluation result, and the brain-charging visual representation information to provide an intuitive display of the user's sleep state.

[0049] In other implementations, wearable devices also possess sleep aid and wake-up functions. The sleep aid function is achieved by playing sleep-aid audio, while the wake-up function is achieved by playing wake-up audio combined with vibration stimulation. Therefore, this embodiment can also generate a sleep aid plan and a wake-up strategy based on the sleep state evaluation results. During the user's next sleep period, the sleep aid audio is adaptively adjusted based on the sleep aid plan, and the wake-up audio is adaptively adjusted based on the wake-up strategy. For example, in practical applications, the wearable device can collect the user's sleep state evaluation results over a period of time (e.g., 3 days). If the user's sleep quality is found to be poor, guided relaxation sleep-aid audio can be played during the user's next sleep period to help the user fall asleep faster. If the user's sleep quality is found to be good and difficult to wake up, vibration pre-wake-up can be used, followed by playing gradually increasing, long-cycle wake-up audio to effectively wake the user.

[0050] The method in this embodiment can be applied to long-term sleep scenarios, nap scenarios, or other short-term rest scenarios.

[0051] Based on the above embodiments, the present invention also provides an effective sleep state recognition system that integrates sleep stages and drowsiness characteristics. This system is used to implement the steps in the above method embodiments. Specifically, as... Figure 2 As shown, the system includes: a multi-dimensional data acquisition module 10, a multi-dimensional signal analysis module 20, and a sleep effectiveness determination module 30. Specifically, the multi-dimensional data acquisition module 10 is used to collect the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals based on a wearable device. The EEG signals reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals reflect changes in sleep posture and the number of times the user turns over. The multi-dimensional signal analysis module 20 is used to determine the percentage of effective rest state and the percentage of awake and highly active state based on the EEG signals and the IMU signals for effective signal segments. The sleep effectiveness determination module 30 is used to determine that the current sleep is a valid sleep state when the percentage of effective rest state reaches a first preset threshold and the percentage of awake and highly active state is less than a second preset threshold.

[0052] The principles of each module in the effective sleep state recognition system embodiment that integrates sleep stages and drowsiness features are the same as the implementation process of each step in the above method embodiment, and will not be elaborated further here.

[0053] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 3 As shown. The terminal may include one or more processors 100 ( Figure 3(Only one is shown in the image), memory 101, and computer program 102 stored in memory 101 and executable on one or more processors 100. For example, an effective sleep state recognition program that integrates sleep stages and drowsiness characteristics. When one or more processors 100 execute computer program 102, they can implement the various steps in the effective sleep state recognition method embodiment that integrates sleep stages and drowsiness characteristics. Alternatively, when one or more processors 100 execute computer program 102, they can implement the functions of various modules / units in the effective sleep state recognition device embodiment that integrates sleep stages and drowsiness characteristics, which is not limited here.

[0054] In one embodiment, the processor 100 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0055] In one embodiment, memory 101 can be an internal storage unit of the terminal, such as a hard disk or RAM. Memory 101 can also be an external storage terminal of the terminal, such as a plug-in hard disk, smart media card (SM), secure digital card (SD), flash card, etc., all equipped on the terminal. Furthermore, memory 101 can include both internal and external storage units. Memory 101 is used to store computer programs and other programs and data required by the terminal. Memory 101 can also be used to temporarily store data that has been output or will be output.

[0056] Those skilled in the art will understand that Figure 3 The block diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0057] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), direct memory bus RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An effective sleep state identification method integrating sleep stages and drowsiness characteristics, characterized in that, The method includes: The system collects the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals using wearable devices. The EEG signals are used to reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals are used to reflect changes in sleep posture and the number of times the user turns over. For valid signal segments, based on the EEG signal and the inertial measurement unit signal, determine the percentage of valid resting state and the percentage of awake and highly active state; When the percentage of the effective rest state reaches a first preset threshold and the percentage of the awake and highly active state is less than a second preset threshold, the current sleep is determined to be an effective sleep state.

2. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 1, characterized in that, The method further includes: The signal quality of the EEG signals and inertial measurement unit signals collected in each time segment is judged to determine the effective signal segments. In the effective signal segments, the electrode contact state of the EEG signals is stable.

3. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 1, characterized in that, For valid signal segments, based on the EEG signals and the inertial measurement unit signals, the percentage of valid resting states and the percentage of awake, highly active states are determined, including: Based on the EEG signals, the ratio between effective EEG signal segments and total EEG signal segments is calculated to obtain the signal integrity. The device usage data, device status, and device usage duration of the wearable device are obtained, and verification is performed based on the device usage data, device status, device usage duration, and signal integrity. Once the verification is successful, the percentage of effective resting state and the percentage of awake and highly active state are determined based on the EEG signal and the inertial measurement unit signal.

4. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 3, characterized in that, Once the verification is successful, based on the EEG signals and the inertial measurement unit signals, the percentage of effective resting states and the percentage of awake, highly active states are determined, including: Based on the electroencephalogram (EEG) signals, the electrode contact status, sleep stages, and drowsiness index are determined. Based on the signals from the inertial measurement unit, the number of times the user turns over and the changes in sleep posture are determined; Obtain effective rest signal segments for the sleep stages of light sleep, deep sleep, and rapid eye movement; or, obtain effective rest signal segments for the sleep stage of wakefulness where the drowsiness index reaches a preset range, the number of times the person turns over within a preset time period is less than a first preset number, the number of sleep posture changes is less than a second preset number, and the electrode contact state is stable. The effective rest state percentage is obtained based on the ratio between the effective rest signal segment and the total signal segment. The system acquires a high-activity wakefulness signal segment when the sleep stage is a waking state and the drowsiness index is less than a fourth preset threshold; or, it acquires a high-activity wakefulness signal segment when the sleep stage is a waking state and the number of times the sleeper turns over and the number of sleep posture changes within a preset time period are both greater than a third preset number. The proportion of the awake and highly active state is obtained based on the ratio between the awake and highly active signal segments and the total signal segments.

5. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 4, characterized in that, The method further includes: After identifying effective sleep states, a rest effectiveness score is calculated based on the effective rest duration coefficient, the proportion of effective rest states, and the rest proximity coefficient. The rest proximity coefficient is calculated based on the drowsiness index. Awakening alertness score is calculated based on wake-up completion coefficient, wake-up timing coefficient, and post-wake stability coefficient. The reliability score is calculated based on signal integrity, contact state stability, body motion coefficient, and equipment continuous use coefficient. The final sleep score is obtained based on the rest effectiveness score, wakefulness score, and process credibility score.

6. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 5, characterized in that, The method further includes: Obtain a preset evaluation text library, which contains evaluation tags corresponding to different sleep score ranges; The final sleep score is matched with the evaluation text library to determine the target evaluation tag corresponding to the final sleep score; Obtain the explanatory text of the sleep state corresponding to the target evaluation label to obtain the sleep state evaluation result.

7. The effective sleep state identification method integrating sleep stages and drowsiness characteristics according to claim 6, characterized in that, The method further includes: Based on the sleep state evaluation results, a sleep aid plan and a wake-up strategy are generated; When the user goes to sleep next time, the sleep aid audio is adaptively adjusted based on the sleep aid scheme, and the wake-up audio is adaptively adjusted based on the wake-up strategy.

8. An effective sleep state recognition system integrating sleep stages and drowsiness characteristics, characterized in that, The system is used to implement the steps of the effective sleep state identification method that integrates sleep stages and drowsiness characteristics as described in any one of claims 1-7, the system comprising: A multi-dimensional data acquisition module is used to collect the user's electroencephalogram (EEG) signals and inertial measurement unit (IMU) signals based on wearable devices. The EEG signals are used to reflect electrode contact status, sleep stages, and drowsiness index, while the IMU signals are used to reflect changes in sleep posture and the number of times the user turns over. The multi-dimensional signal analysis module is used to determine the percentage of effective resting state and the percentage of awake and highly active state based on the EEG signal and the inertial measurement unit signal for effective signal segments. The sleep effectiveness determination module is used to determine the current sleep as an effective sleep state when the proportion of the effective rest state reaches a first preset threshold and the proportion of the awake and highly active state is less than a second preset threshold.

9. A terminal, characterized in that, The terminal includes a memory, a processor, and an effective sleep state recognition program that integrates sleep stages and drowsiness features, stored in the memory and executable on the processor. When the processor executes the effective sleep state recognition program that integrates sleep stages and drowsiness features, it implements the steps of the effective sleep state recognition method that integrates sleep stages and drowsiness features as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an effective sleep state recognition program that integrates sleep stages and drowsiness features. The effective sleep state recognition program that integrates sleep stages and drowsiness features implements the steps of the effective sleep state recognition method that integrates sleep stages and drowsiness features as described in any one of claims 1-7 on the computer-readable storage medium.

Images