Sleep quality evaluation method and device based on intelligent ring multi-mode sensing

Abstract

The invention provides a sleep quality evaluation method and device based on intelligent ring multi-mode sensing, the method is applied to an intelligent ring, and the method comprises the following steps: collecting multi-mode sleep data of a user in a sleep cycle; determining a multi-dimensional sleep score according to the multi-modal sleep data based on a preset sleep model; and determining a comprehensive sleep score according to the multi-dimensional sleep score. According to the method, the anti-interference performance and medical credibility of the sleep staging and scoring result are fundamentally improved, and the evaluation misaccuracy caused by misjudgment of a single signal is avoided, so that the evaluation depth and reliability close to professional medical analysis are realized on consumer equipment, and unprecedented accurate health insight is provided for users.

Description

Technical Field

[0001] This invention relates to the field of wearable device technology, specifically to a method and device for assessing sleep quality based on multimodal sensing in a smart ring. Background Technology

[0002] Sleep is an important physiological process for the human body, and accurate assessment of sleep quality is crucial for physical and mental health. The traditional "gold standard" for sleep monitoring is polysomnography (PSG), but it requires wearing a large number of sensors in a professional laboratory, is expensive, has a poor user experience, and cannot be used for daily monitoring.

[0003] Currently, while consumer wearable devices (such as smart bracelets and watches) offer sleep monitoring functions, they have significant limitations. These devices typically rely solely on motion analysis to roughly determine sleep status, providing only a single data dimension and failing to accurately distinguish fine sleep structures such as deep and light sleep stages. The sleep scores they provide are also mostly based on simple calculations of sleep duration, resulting in a crude assessment that fails to deeply integrate and quantify sleep quality from multiple physiological parameter levels. This leads to discrepancies between the assessment results and the user's actual sleep experience, limiting their reference value.

[0004] Furthermore, these devices are bulky and uncomfortable to wear, which may actually disrupt sleep. Therefore, there is an urgent need in this field for a solution that can collect multi-dimensional physiological data in a seamless and comfortable manner, and then perform in-depth fusion analysis based on this data to provide a scientific and accurate assessment of sleep quality. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a sleep quality assessment method based on multimodal sensing in a smart ring, applicable to the smart ring, comprising: Collect multimodal sleep data from users during their sleep cycles; Based on a preset sleep model, a multi-dimensional sleep score is determined according to the multimodal sleep data; A comprehensive sleep score is determined based on the multi-dimensional sleep score.

[0006] In one embodiment, the above-described collection of multimodal sleep data of the user during the sleep cycle includes: The duration of a user's time in bed is determined based on the user's initial action of getting into bed and their final action of getting out of bed; The duration of sleep for a user is determined based on their initial lying down posture and the timing of their fall asleep. Collect users' heart rate, heart rate variability, body movement information, core body temperature changes, and blood oxygenation information; Multimodal sleep data is determined based on the user's described time spent in bed, time to fall asleep, heart rate, heart rate variability, body movement information, core body temperature changes, and blood oxygenation information.

[0007] In one embodiment, the preset sleep model includes a sleep staging model and a multi-dimensional scoring engine. The step of determining a multi-dimensional sleep score based on the sleep staging model and the multimodal sleep data includes: After a sleep cycle is completed, the multimodal sleep data is uploaded to the server so that the server can perform data cleaning and feature extraction. Based on the sleep staging model, sleep stages are segmented according to the multimodal sleep data and a preset time interval, wherein the preset time interval is no more than 3 minutes. Based on a multi-dimensional sleep scoring engine, multi-dimensional sleep scores are determined based on multimodal sleep data after data cleaning and feature extraction, as well as staged sleep data. The multi-dimensional sleep scores include sleep efficiency score, sleep structure score, sleep stability score, sleep onset score, and sleep consistency score.

[0008] In one embodiment, the multi-dimensional sleep scoring engine determines a multi-dimensional sleep score based on multimodal sleep data after data cleaning and feature extraction, as well as staged sleep data. The multi-dimensional sleep score includes a sleep efficiency score, a sleep structure score, a sleep stability score, a sleep onset score, and a sleep consistency score, including: Based on the staged sleep data, wakefulness time information and sleep time information are determined; the sleep time information includes the duration and characteristic nodes of each stage of characteristic sleep, which includes deep sleep, light sleep, REM sleep and wakefulness. A sleep efficiency score should be determined based on at least total sleep duration and time spent in bed; At least the proportion and absolute duration of each sleep stage in the total sleep duration should be compared with the conventional model for peers and the individual baseline to determine the sleep structure score; A sleep stability score should be determined based on at least heart rate variability, core body temperature changes, duration of the wake phase, and key features. A sleep onset score should be determined based on at least the duration of sleep onset. A sleep consistency score should be determined based on sleep patterns from at least multiple sleep cycles.

[0009] In one embodiment, determining the comprehensive sleep score based on the multi-dimensional sleep score includes: Based on a preset weighted fusion formula, a comprehensive sleep score is determined according to the multi-dimensional sleep scores. A sleep report is pushed to the user based on the first UI interface of the device with a screen. The sleep report includes a comprehensive sleep score, duration of each stage, trend comparison, and a first personalized suggestion report. The first personalized suggestion report is generated based on at least the dimension with the lowest score in the multi-dimensional sleep score.

[0010] In one embodiment, after determining the comprehensive sleep score based on the multi-dimensional sleep score, the method further includes: Personalized sleep baseline data for users is determined based on multiple consecutive comprehensive sleep scores; Based on the user's unusual behavior before sleep, obtain information on the user's unusual behavior characteristics before sleep; Based on the user's multi-dimensional sleep score that night, determine the overall sleep score after the anomaly; The overall sleep score after the change is compared with the personalized sleep baseline data to assess the independent impact of the change on sleep quality and generate a second personalized recommendation report.

[0011] In one embodiment, based on the user's unusual movements before sleep, information on the user's unusual movements before sleep is obtained, including: Collect the user's vital signs information during a preset time period before sleep. The vital signs information includes heart rate, heart rate variability, body movement information, core body temperature change information, and blood oxygen information. Based on the vital signs information, the user's first pre-sleep behavior is matched in a preset behavior database. The first pre-sleep behavior includes getting angry, aerobic / anaerobic exercise before bed, meditation, or foot bath. The user's environmental information before going to sleep is obtained from the device with a screen. The environmental information includes lighting information, noise information and temperature information. The system obtains a second pre-sleep behavior input by the user on a second UI interface of a device with a screen. The second pre-sleep behavior includes drinking coffee / tea, eating a late-night snack, working, and playing games.

[0012] Based on the first pre-sleep behavior, environmental information, and the second pre-sleep behavior, compare them with the routine phenomena of personalized sleep baseline data to identify abnormal phenomena; Based on the abnormal activity, the characteristic parameters of the abnormal activity are obtained, and the abnormal activity characteristic information of the user before going to sleep is generated.

[0013] This invention also provides a sleep quality assessment device based on multimodal sensing in a smart ring, applied to a smart ring, comprising: The data acquisition module is used to collect multimodal sleep data from users during their sleep cycles. The first determining module is used to determine a multi-dimensional sleep score based on the multimodal sleep data according to a preset sleep model. The second determining module is used to determine the comprehensive sleep score based on the multi-dimensional sleep score.

[0014] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described sleep quality assessment method based on multimodal sensing of a smart ring.

[0015] The present invention also provides a computer storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned sleep quality assessment method based on multimodal sensing of a smart ring.

[0016] This invention utilizes a smart ring to collect multimodal sleep data encompassing multiple physiological states such as heart rate, body movement, and body temperature without being detected. Based on a preset model, it performs in-depth fusion analysis and ultimately outputs a comprehensive sleep score. This not only achieves refined and objective quantification of sleep quality under natural sleep conditions, but more importantly, by leveraging cross-validation and physiological correlations among multimodal data, it fundamentally improves the anti-interference capability and medical reliability of sleep staging and scoring results. It avoids inaccurate assessments caused by misjudgments of a single signal (such as body movement), thus achieving near-professional medical-grade analysis depth and reliability on consumer-grade devices, providing users with unprecedentedly accurate health insights. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the sleep quality assessment method based on multimodal sensing of a smart ring according to the first embodiment of the present invention; Figure 2 This is a detailed flowchart of S11 in the first embodiment of the present invention; Figure 3 This is a detailed flowchart of S12 in the first embodiment of the present invention; Figure 4 This is a detailed flowchart of S123 of the first embodiment of the present invention; Figure 5 This is one of the detailed flowcharts of S13 in the first embodiment of the present invention; Figure 6 This is one of the detailed flowcharts of S15 in the first embodiment of the present invention; Figure 7 A structural block diagram of a sleep quality assessment device based on multimodal sensing of a smart ring according to a second embodiment of the present invention; Figure 8 This is a schematic diagram of the internal structure of a computer according to another embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Well-known modules, units, and their connections, links, communications, or operations are not shown or described in detail. Furthermore, the described features, architectures, or functions can be combined in any way in one or more embodiments. Those skilled in the art should understand that the various embodiments described below are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. It is also readily understood that the modules, units, or processing methods in the various embodiments described herein and shown in the accompanying drawings can be combined and designed in various different configurations. 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.

[0020] Please refer to Figures 1 to 6 As shown, this embodiment of the invention discloses a sleep quality assessment method based on multimodal sensing of a smart ring, including steps S11 to S13, wherein: S11 collects multimodal sleep data of users during their sleep cycles.

[0021] As a specific solution and not a limitation, the execution subject in this embodiment is a smart ring, which can work collaboratively with a cloud server system and devices with screens. This step is completed by the sensor array and embedded program on the smart ring. The sleep cycle can be the complete time period from when the user prepares to fall asleep to when they wake up the next day; or it can be another complete time period from a preset time before going to sleep (which may include one or two hours before going to bed) to when the user wakes up the next day, depending on sleep habits. Multimodal sleep data refers to a set of data collected through different types of sensors that can reflect sleep states from multiple physiological dimensions.

[0022] As a specific solution, please refer to Figure 2 As shown, this step S11 includes sub-steps S111 to S114, wherein: S111, determine the user's bedtime based on the user's initial action of getting into bed and final action of getting out of bed.

[0023] In this step, the smart ring continuously monitors the user's posture using its built-in three-axis accelerometer and gyroscope. It can identify the initial stepping into bed by recognizing a specific acceleration and angular velocity pattern of the wrist or wearing finger changing from vertical to horizontal and remaining stationary; conversely, it can identify the final step of getting out of bed by recognizing a pattern of changing from a stationary horizontal to a vertical position accompanied by displacement.

[0024] It should be noted that when a user experiences sleep apnea, such as waking up at night, sleepwalking, or sitting up in the middle of the night, the final action of getting out of bed is usually determined based on the duration of the sleep apnea and whether sleep resumes afterward. The duration in bed is the difference between these two points in time.

[0025] S112 determines the user's sleep duration based on the user's initial lying down action and the timing of falling asleep.

[0026] The initial lying-down action is a more detailed description of the initial bed-getting action in S111, specifically referring to the body entering a lying position. The onset of sleep is not based on the user's subjective feeling, but rather determined by a subsequent sleep stage model that analyzes the entire night's data to identify the start of the first sleep stage (usually N1 stage (early light sleep) or N2 stage (mid-light sleep)). Sleep latency is the time interval from the initial lying-down action to the onset of sleep latency, and is a key indicator for evaluating sleep efficiency.

[0027] S113 collects the user's heart rate, heart rate variability, body movement information, core body temperature change information, and blood oxygen information.

[0028] This step is the core of data acquisition, and it is synchronously completed by the multimodal sensor array of the smart ring, including: Heart rate and heart rate variability were acquired using a PPG (Photoplethysmography) sensor. This sensor emits green light (sensitive to changes in blood flow) and / or red light (commonly used for blood oxygenation) into the skin and receives the reflected light signals. By analyzing the periodic and morphological fluctuations of the pulse wave, heart rate and heart rate variability were calculated.

[0029] Body motion information is collected via a triaxial accelerometer to monitor macroscopic movements (such as turning over) and microscopic movements (such as limb tremors).

[0030] Core body temperature changes are continuously monitored using high-precision skin temperature sensors (typically thermistors or infrared sensors) to track fingertip skin temperature. In a constant sleep environment, the trend of fingertip skin temperature changes is considered an effective proxy for the rhythmic decline in core body temperature, an important physiological signal related to sleep depth.

[0031] Blood oxygen information is transmitted via red and infrared light emitted by a PPG sensor. Blood oxygen saturation is calculated by utilizing the difference in absorption rates of hemoglobin and oxyhemoglobin to different wavelengths of light.

[0032] S114. Based on the user's description of time spent in bed, time to fall asleep, heart rate, heart rate variability, body movement information, core body temperature change information, and blood oxygen information, determine multimodal sleep data.

[0033] This step involves integrating all the aforementioned data and information to form a structured data package or string that can be used for subsequent analysis.

[0034] The steps S111-S114 above, by comprehensively utilizing motion, optical and temperature sensors, achieve all-round and non-invasive collection of sleep-related physiological and behavioral signals, providing a rich and multi-dimensional data foundation for subsequent refined analysis, fundamentally overcoming the shortcomings of existing technologies with single data dimensions.

[0035] S12, Based on a preset sleep model, determine a multi-dimensional sleep score according to the multimodal sleep data.

[0036] This step utilizes the powerful computing resources of cloud servers to run complex sleep models. These sleep models can be machine learning or deep learning models pre-trained using a large amount of PSG-annotated sleep data. In this step, the preset sleep models include a sleep staging model and a multi-dimensional scoring engine.

[0037] Please refer to Figure 3 As shown, this step S12 includes sub-steps S121 to S123, wherein: S121, after a sleep cycle is completed, the multimodal sleep data is uploaded to the server so that the server can complete data cleaning and feature extraction.

[0038] In this step, a sleep cycle typically refers to a complete sleep cycle of one night. Data cleaning includes removing noise segments from the signal caused by motion artifacts (such as violent turning over) or brief sensor detachment, and filtering and smoothing the signal. Feature extraction involves extracting quantitative indicators related to sleep structure from the cleaned raw signal. For example, extracting time-domain indicators (such as RMSSD) and frequency-domain indicators (such as the LF / HF ratio) of HRV from PPG signals, extracting body kinetic energy from acceleration signals, and extracting the nighttime temperature drop slope from temperature signals.

[0039] S122, Based on the sleep stage model, the sleep stages are staged according to the multimodal sleep data and a preset time interval, wherein the preset time interval is no more than 3 minutes.

[0040] The sleep staging model, such as a machine learning model (e.g., random forest, XGBoost) or a deep learning model (e.g., CNN, LSTM), receives the feature sequence extracted in step S121. The model processes data in units of preset time intervals, preferably 30 seconds, to ensure the precision of the staging. The model outputs the sleep stages for each preset time interval, including: wakefulness, light sleep (N1 / N2 stage), deep sleep (N3 stage), and REM sleep.

[0041] S123, based on a multi-dimensional sleep scoring engine, determines multi-dimensional sleep scores based on multimodal sleep data after data cleaning and feature extraction, as well as staged sleep data. The multi-dimensional sleep scores include sleep efficiency score, sleep structure score, sleep stability score, sleep onset score, and sleep consistency score.

[0042] The multi-dimensional sleep scoring engine in this embodiment is a set of rule-based or lightweight machine learning algorithm-based calculation logic. It uses the stage results and extracted features to calculate scores for multiple dimensions according to a preset formula.

[0043] As a preferred option and not a limitation, in this embodiment, please refer to... Figure 4 As shown, this step S123 includes S1231 to S1236, wherein: S1231, Based on the staged sleep data, determine the wakefulness time information and sleep time information; the sleep time information includes the duration and characteristic nodes of each staged characteristic sleep stage, which includes deep sleep, light sleep, REM (rapid eye movement sleep) and wakefulness.

[0044] The feature nodes include the start time, end time, and duration of each sleep stage. A staged sleep stage can contain multiple sleep stages with preset time intervals. For example, a staged sleep stage can be a continuous period of wakefulness, light sleep, deep sleep, and REM sleep.

[0045] S1232, determine a sleep efficiency score based at least on total sleep duration and time spent in bed.

[0046] In this step, sleep efficiency = (total sleep time / time in bed) × 100%. This score directly reflects how much of the time spent in bed is actually used for sleep or sleep preparation.

[0047] S1233, at least based on the proportion and absolute duration of each sleep stage in the total sleep duration, compare with the conventional model of peers and personal baseline to determine the sleep structure score.

[0048] Peer baselines, derived from statistical data in large-scale population studies, provide the normal range for the proportion of each sleep stage. An individual baseline is the average of a user's sleep structure over a past period (e.g., a week). By comparing a user's sleep structure to these two, it's possible to assess whether their current sleep structure is healthy and whether it deviates from their normal range.

[0049] S1234, at least based on heart rate variability, core body temperature change information, duration of wakefulness and characteristic nodes, determine sleep stability score.

[0050] Since the high-frequency component of heart rate variability is positively correlated with parasympathetic activity (degree of physical relaxation), its overall high and stable level at night indicates good sleep depth. A sustained, smooth decrease in core body temperature is another physiological marker of sleep stability. Conversely, frequent and prolonged awakenings significantly reduce stability scores. This step quantifies sleep depth and coherence by integrating these physiological signals.

[0051] S1235, determine the sleep onset score based at least on sleep onset time.

[0052] Generally, the shorter the time it takes to fall asleep, the higher the sleep onset score.

[0053] S1236, Determine the sleep consistency score based on sleep patterns from at least multiple sleep cycles.

[0054] This step assesses the regularity and consistency of a user's sleep patterns by analyzing the standard deviation of their sleep and wake times over several consecutive days.

[0055] The steps S121-S123 above, by introducing a sleep staging model and a multi-dimensional scoring engine, transform the raw multimodal data into quantitative scores with clear physiological significance, achieving a leap in sleep quality assessment. In particular, by integrating heart rate variability and body temperature to assess sleep stability, they provide in-depth insights that go beyond traditional body movement analysis.

[0056] S13, Determine the comprehensive sleep score based on the multi-dimensional sleep score.

[0057] Please refer to Figure 5 As shown, this step S13 also includes sub-steps S131-S132, wherein: S131, Based on the preset weighted fusion formula, determine the comprehensive sleep score according to the multi-dimensional sleep score.

[0058] This step combines multi-dimensional sleep scores (sleep efficiency, sleep structure, sleep stability, sleep onset, and sleep consistency) into a single, easily understood, and observable comprehensive sleep score (e.g., 0-100). Weighting can be based on clinical research or fine-tuned according to individual user circumstances to highlight the dimensions most important to them.

[0059] S132, a sleep report is pushed to the user based on the first UI interface of the device with a screen. The sleep report includes a comprehensive sleep score, duration of each stage, trend comparison, and a first personalized suggestion report. The first personalized suggestion report is generated based on at least the dimension with the lowest score in the multi-dimensional sleep score.

[0060] The device with a screen in this step can be a user's smartphone, smartwatch, or tablet, or wearable devices such as smartwatches, VR / AR / MR glasses, etc. Sleep reports can visually display data graphically. Trend comparisons can be made with the user's own historical data or with data from peers. The first personalized recommendation report can employ a strategy to address weaknesses. For example, if the sleep stability score is the lowest, the recommendation might be "avoid caffeine and strenuous exercise before bed"; if the sleep onset score is the lowest, the recommendation might be "establish a consistent bedtime relaxation ritual."

[0061] The steps S131 and S132 above generate a comprehensive sleep score and targeted suggestions, transforming complex data analysis results into a user-understandable and intuitive graphical report. Combined with the first personalized suggestion report, this forms a complete closed loop from monitoring to intervention, greatly improving the system's usability and enhancing the user experience.

[0062] As a preferred option and not a limitation, steps S14 to S17 are included after step S13, wherein: S14 determines the user's personalized sleep baseline data based on multiple consecutive comprehensive sleep scores.

[0063] This step is performed by a cloud server. The system collects the user's comprehensive sleep score and scores for each dimension over multiple consecutive sleep cycles (e.g., a week). By calculating statistics such as moving average and standard deviation, it establishes a personalized, dynamically updated sleep quality baseline for the user. This quantitative baseline more scientifically reflects the user's individual normal level, avoiding the bias that may arise from simple comparisons with a single, fixed group standard.

[0064] S15: Based on the user's unusual activity before going to sleep, obtain the user's unusual activity characteristics information before going to sleep.

[0065] The anomalous phenomena in this step refer to behaviors, events, or environmental variables that significantly deviate from the user's usual bedtime habits or physiological state within a period of time before sleep (e.g., 1-2 hours before bedtime). These phenomena may have a negative impact on sleep quality (e.g., anger, strenuous exercise) or a positive impact (e.g., meditation, foot bath). Identifying these phenomena is key to achieving accurate attribution and personalized recommendations.

[0066] Please refer to this as a preferred option rather than a limitation. Figure 6 As shown, this step S15 includes sub-steps S151 to S155, wherein: S151, Collect the user's vital signs information during a preset period before sleep. The vital signs information includes heart rate, heart rate variability, body movement information, core body temperature change information, and blood oxygen information.

[0067] This step is similar to nighttime monitoring, but focuses on the pre-sleep period. The smart ring continues to work, collecting physiological signals before sleep. For example, anger or stress may cause a short-term increase in heart rate and a decrease in heart rate variability; strenuous exercise will cause a gradual increase in heart rate and a sustained high level, while body temperature will rise; and meditation or foot baths may result in a stable heart rate within a preset range and an increase in heart rate variability.

[0068] S152, Match the user's first pre-sleep behavior in the preset behavior database based on the vital signs information. The first pre-sleep behavior includes getting angry, aerobic / anaerobic exercise before bed, meditation, or foot bath.

[0069] The preset behavior database is a cloud-based database that contains the correspondence between different pre-sleep behaviors and physiological patterns. The system matches the real-time physiological patterns collected in S151 with the patterns in the behavior database to infer the activities the user is most likely to engage in, thus achieving seamless, physiological data-based pre-sleep behavior recognition.

[0070] S153, Obtain the user's environmental information before going to sleep from the screen device, the environmental information including light information, noise information and temperature information.

[0071] This step supplements data collection using sensors on the user's smartphone or other screen-equipped device. Excessive blue light, high ambient noise, or unsuitable ambient temperature are all environmental factors that can affect sleep.

[0072] S154, obtain the second bedtime behavior input by the user on the second UI interface of the device with a screen, the second bedtime behavior includes drinking coffee / tea, eating a late-night snack, working and playing games.

[0073] This step supplements the system's automatic detection of pre-sleep behavior information through user-generated logs. Users can manually record these subjective behaviors before bed through the app interface, providing the system with a more comprehensive information foundation.

[0074] S155, Based on the first pre-sleep behavior, environmental information, and the second pre-sleep behavior, compare them with the routine phenomena of the personalized sleep baseline data to identify abnormal phenomena.

[0075] The system compares the pre-sleep behavior and environmental data identified in S152-S154 with the user's typical pre-sleep patterns included in the personalized sleep baseline data established in S14. For example, if the user's baseline shows that their heart rate is usually stable and they do not consume caffeine before bed, and the system detects an abnormal increase in their heart rate on a certain day and they manually record drinking coffee, then an anomaly is determined to have occurred.

[0076] The steps S151-S155 above, by integrating physiological sensing, environmental perception and user input, construct a multi-dimensional pre-sleep behavior recognition system that can accurately and automatically capture various abnormal factors affecting sleep quality, providing solid data support for subsequent causal analysis.

[0077] S16. Based on the user's multi-dimensional sleep score that night, determine the comprehensive sleep score after the anomaly.

[0078] The comprehensive sleep score following the abnormality here refers to the comprehensive sleep score calculated using the method in step S13 on the night the abnormality was identified in S15.

[0079] S17. Compare the overall sleep score after the change with the personalized sleep baseline data, assess the independent impact of the change on sleep quality, and generate a second personalized recommendation report.

[0080] This step doesn't simply report a lower score; rather, it quantifies the independent impact of that specific anomaly on sleep by comparing the score after the anomaly to the baseline level. For example, the report might explicitly state: "Consuming caffeine before bed last night may have reduced your deep sleep percentage by 15% compared to your usual average." Based on this assessment, the system will generate a more targeted second personalized recommendation report, such as: "Data shows that caffeine significantly affects your sleep depth; it is recommended that you limit your last cup of coffee to at least 6 hours before bedtime." Steps S14-S17 above establish a personal baseline, intelligently identify pre-sleep anomalies, and perform attribution analysis, achieving an intelligent leap from monitoring sleep outcomes to tracing factors influencing sleep. This makes the system's recommendations more than just general statements; they become highly personalized health guidance based on the user's individual physiological responses and empirical data. More importantly, after a user's pre-sleep habits are adjusted according to the first personalized recommendation report, the system can continuously track their sleep data through steps S14-S17, objectively quantifying the actual effect of the improvement measures. If the effect is significant, the recommendation is reinforced; if the effect is poor, the system can further analyze the reasons, fine-tune and optimize the initial recommendation, thereby generating a more targeted second personalized recommendation report. This continuous closed loop of "assessment-recommendation-reassessment-optimization" greatly improves the system's intelligence level, the accuracy of intervention, and long-term user engagement.

[0081] This invention utilizes a smart ring to collect multimodal sleep data encompassing multiple physiological states such as heart rate, body movement, and body temperature without being detected. Based on a preset model, it performs in-depth fusion analysis and ultimately outputs a comprehensive sleep score. This not only achieves refined and objective quantification of sleep quality under natural sleep conditions, but more importantly, by leveraging cross-validation and physiological correlations among multimodal data, it fundamentally improves the anti-interference capability and medical reliability of sleep staging and scoring results. It avoids inaccurate assessments caused by misjudgments of a single signal (such as body movement), thus achieving near-professional medical-grade analysis depth and reliability on consumer-grade devices, providing users with unprecedentedly accurate health insights.

[0082] Second embodiment: Please refer to Figure 7 As shown, the present invention also provides a sleep quality assessment device 100 based on multimodal sensing of a smart ring, applied to a smart ring, comprising: The acquisition module 110 is used to collect multimodal sleep data of the user during the sleep cycle; The first determining module 120 is used to determine a multi-dimensional sleep score based on the multimodal sleep data according to a preset sleep model. The second determining module 130 is used to determine a comprehensive sleep score based on the multi-dimensional sleep score.

[0083] The modules in this embodiment are the same as the corresponding steps in the first embodiment described above, and will not be repeated here.

[0084] This invention utilizes a smart ring to collect multimodal sleep data encompassing multiple physiological states such as heart rate, body movement, and body temperature without being detected. Based on a preset model, it performs in-depth fusion analysis and ultimately outputs a comprehensive sleep score. This not only achieves refined and objective quantification of sleep quality under natural sleep conditions, but more importantly, by leveraging cross-validation and physiological correlations among multimodal data, it fundamentally improves the anti-interference capability and medical reliability of sleep staging and scoring results. It avoids inaccurate assessments caused by misjudgments of a single signal (such as body movement), thus achieving near-professional medical-grade analysis depth and reliability on consumer-grade devices, providing users with unprecedentedly accurate health insights.

[0085] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0086] This invention also provides a computer storage medium storing a computer program that, when executed by a processor, implements the sleep quality assessment method based on multimodal sensing of a smart ring as described in the above embodiments.

[0087] Those skilled in the art will understand that all or part of the processes in 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 sleep quality assessment method based on multimodal sensing of a smart ring as described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0088] Alternatively, if the integrated units of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, terminal, or network device, etc.) to execute all or part of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, RAM, ROM, magnetic disks, or optical disks.

[0089] Corresponding to the computer storage medium described above, one embodiment also provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the sleep quality assessment method based on multimodal sensing of a smart ring as described in the above embodiments.

[0090] This computer device can be a terminal, and its internal structure diagram can be as follows: Figure 8 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When executed by the processor, the computer program implements a sleep quality assessment method based on multimodal sensing of a smart ring. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device casing, or an external keyboard, touchpad, or mouse.

[0091] This invention utilizes a smart ring to collect multimodal sleep data encompassing multiple physiological states such as heart rate, body movement, and body temperature without being detected. Based on a preset model, it performs in-depth fusion analysis and ultimately outputs a comprehensive sleep score. This not only achieves refined and objective quantification of sleep quality under natural sleep conditions, but more importantly, by leveraging cross-validation and physiological correlations among multimodal data, it fundamentally improves the anti-interference capability and medical reliability of sleep staging and scoring results. It avoids inaccurate assessments caused by misjudgments of a single signal (such as body movement), thus achieving near-professional medical-grade analysis depth and reliability on consumer-grade devices, providing users with unprecedentedly accurate health insights.

[0092] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0093] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A sleep quality assessment method based on multimodal sensing in a smart ring, characterized in that, Applications in smart rings include: Collect multimodal sleep data from users during their sleep cycles; Based on a preset sleep model, a multi-dimensional sleep score is determined according to the multimodal sleep data; A comprehensive sleep score is determined based on the multi-dimensional sleep score.

2. The method as described in claim 1, characterized in that, The collection of multimodal sleep data from users during their sleep cycles includes: The duration of a user's time in bed is determined based on the user's initial action of getting into bed and their final action of getting out of bed; The duration of sleep for a user is determined based on their initial lying down posture and the timing of their fall asleep. Collect users' heart rate, heart rate variability, body movement information, core body temperature changes, and blood oxygenation information; Multimodal sleep data is determined based on the user's described time spent in bed, time to fall asleep, heart rate, heart rate variability, body movement information, core body temperature changes, and blood oxygenation information.

3. The method as described in claim 2, characterized in that, The preset sleep model includes a sleep staging model and a multi-dimensional scoring engine. The determination of the multi-dimensional sleep score based on the sleep staging model and the multi-modal sleep data includes: After a sleep cycle is completed, the multimodal sleep data is uploaded to the server so that the server can perform data cleaning and feature extraction. Based on the sleep staging model, sleep stages are segmented according to the multimodal sleep data and a preset time interval, wherein the preset time interval is no more than 3 minutes. Based on a multi-dimensional sleep scoring engine, multi-dimensional sleep scores are determined based on multimodal sleep data after data cleaning and feature extraction, as well as staged sleep data. The multi-dimensional sleep scores include sleep efficiency score, sleep structure score, sleep stability score, sleep onset score, and sleep consistency score.

4. The method as described in claim 3, characterized in that, The multi-dimensional sleep scoring engine determines multi-dimensional sleep scores based on cleaned and feature-extracted multimodal sleep data and staged sleep data. These multi-dimensional sleep scores include sleep efficiency score, sleep structure score, sleep stability score, sleep onset score, and sleep consistency score. Based on the staged sleep data, wakefulness time information and sleep time information are determined; the sleep time information includes the duration and characteristic nodes of each stage's characteristic sleep stages, which include deep sleep, light sleep, REM sleep, and wakefulness. A sleep efficiency score should be determined based on at least total sleep duration and time spent in bed; At least the proportion and absolute duration of each sleep stage in the total sleep duration should be compared with the conventional model for peers and the individual baseline to determine the sleep structure score; A sleep stability score should be determined based on at least heart rate variability, core body temperature changes, duration of the wake phase, and key features. A sleep onset score should be determined based on at least the duration of sleep onset. A sleep consistency score should be determined based on sleep patterns from at least multiple sleep cycles.

5. The method as described in claim 4, characterized in that, The determination of the comprehensive sleep score based on the multi-dimensional sleep score includes: Based on a preset weighted fusion formula, a comprehensive sleep score is determined according to the multi-dimensional sleep scores. A sleep report is pushed to the user based on the first UI interface of the device with a screen. The sleep report includes a comprehensive sleep score, duration of each stage, trend comparison, and a first personalized suggestion report. The first personalized suggestion report is generated based on at least the dimension with the lowest score in the multi-dimensional sleep score.

6. The method as described in claim 1, characterized in that, After determining the comprehensive sleep score based on the multi-dimensional sleep score, the process also includes: Personalized sleep baseline data for users is determined based on multiple consecutive comprehensive sleep scores; Based on the user's unusual behavior before sleep, obtain information on the user's unusual behavior characteristics before sleep; Based on the user's multi-dimensional sleep score that night, determine the overall sleep score after the anomaly; The overall sleep score after the change is compared with the personalized sleep baseline data to assess the independent impact of the change on sleep quality and generate a second personalized recommendation report.

7. The method as described in claim 6, characterized in that, Based on the user's unusual activity patterns before sleep, obtain information on the user's unusual activity patterns before sleep, including: Collect the user's vital signs information during a preset time period before sleep. The vital signs information includes heart rate, heart rate variability, body movement information, core body temperature change information, and blood oxygen information. Based on the vital signs information, the user's first pre-sleep behavior is matched in a preset behavior database. The first pre-sleep behavior includes getting angry, aerobic / anaerobic exercise before bed, meditation, or foot bath. The user's environmental information before going to sleep is obtained from the device with a screen, including lighting information, noise information and temperature information; The system obtains a second pre-sleep behavior input by the user on a second UI interface of a device with a screen. The second pre-sleep behavior includes drinking coffee / tea, eating a late-night snack, working, and playing games. Based on the first pre-sleep behavior, environmental information, and the second pre-sleep behavior, compare them with the routine phenomena of personalized sleep baseline data to identify abnormal phenomena; Based on the abnormal activity, the characteristic parameters of the abnormal activity are obtained, and the abnormal activity characteristic information of the user before going to sleep is generated.

8. A sleep quality assessment device based on multimodal sensing in a smart ring, characterized in that, Applications in smart rings include: The data acquisition module is used to collect multimodal sleep data from users during their sleep cycles. The first determining module is used to determine a multi-dimensional sleep score based on the multimodal sleep data according to a preset sleep model. The second determining module is used to determine the comprehensive sleep score based on the multi-dimensional sleep score.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the sleep quality assessment method based on multimodal sensing of a smart ring as described in any one of claims 1 to 7.

10. A computer storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the sleep quality assessment method based on multimodal sensing of a smart ring as described in any one of claims 1 to 7.

Images