A method and related device for analyzing female physiological signals in a smart ring

Abstract

The application discloses a female physiological signal analysis method of an intelligent ring and a related device, and is used for improving the accuracy of physiological cycle prediction based on physiological signals. In different physiological cycles of a target female user, physiological index collection data at different time nodes are acquired through the intelligent ring; physiological feature extraction is respectively performed on the physiological index collection data based on different physiological cycles, and physiological feature data is generated for each physiological cycle; a physiological cycle model is constructed according to the physiological feature data, physiological feature reference data and physiological feature distribution data of each physiological cycle; a physiological energy index set is generated for each physiological cycle according to the physiological cycle model; the physiological cycle model is trained according to the physiological energy index set; and dynamic correlation analysis is performed on actual measurement physiological index data of the target female user through the trained physiological stage sub-model.

Description

Technical Field

[0001] This application relates to the field of smart ring detection, and more particularly to a method and device for analyzing female physiological signals from a smart ring. Background Technology

[0002] With the popularization of smart wearable devices, the technology of collecting and analyzing vital signs through smart wearable devices has gradually developed. This application mainly focuses on the analysis of women's menstrual cycles using smart wearable devices.

[0003] Current technologies often use single-parameter threshold analysis, which employs a single physiological parameter to determine a woman's menstrual cycle. For example, the basal body temperature method continuously monitors morning body temperature to identify subtle temperature rises after ovulation. However, these methods are easily affected by various temperature-influencing factors, leading to a relatively high misjudgment rate of the female menstrual cycle.

[0004] While multi-parameter static fusion methods have been attempted to predict women's menstrual cycles, these approaches generally suffer from model rigidity. They employ fixed-weight algorithms, failing to adapt to individual differences among women. For example, the baseline HRV (heart rate variability) of active individuals is 2-3 times higher than that of resting individuals, and differences in age, height, and weight can also cause prediction errors. Secondly, although existing multi-signal fusion schemes utilize machine learning methods, they typically employ a globally uniform model without establishing separate state sub-models for different physiological stages or introducing state matching metrics based on energy functions. Without dynamically adjusting the weights of the menstrual cycle prediction model, the accuracy of menstrual cycle prediction based on physiological signals will decrease. Summary of the Invention

[0005] This application discloses a method and related device for analyzing female physiological signals in a smart ring, which is used to improve the accuracy of menstrual cycle prediction based on physiological signals.

[0006] In a first aspect, embodiments of this application provide a method for analyzing female physiological signals using a smart ring, comprising: During different menstrual cycles of the target female users, physiological indicator data at different time points are collected through smart rings. The physiological indicator data includes at least two types of vital signs data related to the menstrual cycle, including body temperature data. Body temperature data includes body temperature data or body temperature data estimated based on heat flux. Physiological features are extracted from the collected physiological index data based on different physiological cycles, and physiological feature data is generated for each physiological cycle. A physiological cycle model is constructed based on the physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data of each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. A set of physiological energy indicators is generated for each physiological cycle based on the physiological cycle model; The physiological cycle model is trained based on a set of physiological energy indicators; Dynamic correlation analysis was performed on the measured physiological index data of target female users using the trained physiological stage sub-model.

[0007] Optionally, the steps of performing dynamic correlation analysis on the measured physiological index data of target female users using the trained physiological stage sub-model specifically include: The smart ring collects vital signs data of the target female users in real time and generates initial measured data for different time periods. Obtain daily activity information of target female users, filter the detection interval of the initial test data based on the daily activity information of target female users, and generate measured physiological index data; Physiological features are extracted from measured physiological index data to generate measured physiological feature data. The measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data are respectively input into the physiological stage sub-model of the physiological cycle model to generate a set of energy indicators for different physiological cycles. The transition probability is generated based on the physiological cycle characteristic data of the target female users. The transition probability is the evolutionary pattern of the physiological cycle of the target female users. Based on the energy index set and transition probability of different physiological cycles, a dynamic correlation analysis is performed on the physiological cycle status of the target female user to generate the probability of the occurrence of the target physiological cycle.

[0008] Optionally, after the step of performing dynamic correlation analysis on the measured physiological index data of the target female users through the trained physiological stage sub-model, the female physiological signal analysis method further includes: Acquire short-term and long-term body temperature data from the target female users; Detect temperature jump events based on the temperature change sequence of short-term body temperature data within a preset short-term time window; A short-term correction factor is generated based on the magnitude of the temperature jump and the duration of the jump. The high / low characteristic temperature range is divided based on the temperature of at least one complete historical period in long-term body temperature data; Based on the statistical distribution characteristics of the duration and temperature difference of high temperature characteristic intervals in multiple historical cycles in long-term body temperature data, the periodic stability of target female users is assessed, and a long-term correction factor is generated. When the short-term correction factor and the long-term correction factor are in the same direction, the correction strength is increased; when the correction directions are not in the same direction, the weight of the short-term correction factor is reduced. The probability of the target physiological cycle is adjusted over time by using short-term correction factors, long-term correction factors, and correction directions.

[0009] Optionally, after the step of time-correcting the probability of the target menstrual cycle by using short-term correction factors, long-term correction factors, and correction directions, the steps of the female physiological signal analysis method further include: Acquire individual interaction information and sleep quality information of target female users; Generate individual state correction factors based on interaction information; Compensation factors are generated for short-term correction factors based on sleep quality information; Individual state / behavior is modified based on individual state correction factors and compensation factors to adjust the probability of occurrence of the target physiological cycle.

[0010] Optionally, the steps of inputting measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data into the physiological stage sub-model of the physiological cycle model to generate energy index sets for different physiological cycles specifically include: Generate time nodes for physiological characteristic changes based on the characteristics of physiological cycle changes of the target to be measured; The measured physiological characteristic data are divided into time intervals based on the time nodes of physiological characteristic changes; The physiological characteristic distribution data, the measured physiological characteristic data divided into time intervals, and the physiological characteristic reference data are input into the corresponding physiological stage sub-model to generate a set of energy indicators for different physiological cycles.

[0011] Optionally, the steps for constructing a menstrual cycle model based on physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data for each menstrual cycle specifically include: Based on the target female users’ body indicators, physiological characteristic reference data and physiological characteristic distribution data for each menstrual cycle in different time periods are obtained from the database. The physiological characteristic reference data includes the target female users’ historical menstrual cycle data. A menstrual cycle model for the target female user is constructed based on the physiological characteristic data of each menstrual cycle, the reference data of physiological characteristics of each menstrual cycle in different time periods, and the distribution data of physiological characteristics. The menstrual cycle model contains at least two related physiological stage sub-models.

[0012] Optionally, after acquiring physiological indicator data at different time points via a smart ring, and before extracting physiological features from the acquired physiological indicator data based on different menstrual cycles to generate physiological feature data for each menstrual cycle, the female physiological signal analysis method further includes: The collected physiological data were preprocessed, including noise reduction, outlier removal, and smoothing.

[0013] Secondly, embodiments of this application provide a female physiological signal analysis device for a smart ring, comprising: The menstrual cycle vital sign collection unit is used to collect physiological indicator data at different time points during different menstrual cycles of the target female user through a smart ring. The physiological indicator data includes at least two types of vital sign data related to the menstrual cycle, including body temperature data. The body temperature data includes body temperature data or body temperature data estimated based on heat flux. The physiological feature data generation unit is used to extract physiological features from the collected physiological indicator data based on different physiological cycles, and generate physiological feature data for each physiological cycle. The physiological cycle model construction unit is used to construct a physiological cycle model based on the physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data of each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. The physiological energy index set generation unit is used to generate a physiological energy index set for each physiological cycle based on the physiological cycle model. The physiological cycle model training unit is used to train the physiological cycle model based on a set of physiological energy indicators. The dynamic correlation analysis unit is used to perform dynamic correlation analysis on the measured physiological index data of target female users through the trained physiological stage sub-model.

[0014] Optionally, the dynamic correlation analysis unit specifically includes: The measured physiological index data generation module is used to obtain the daily activity information of the target female users, and to filter the detection interval of the initial measured data through the daily activity information of the target female users to generate measured physiological index data. The measured physiological feature data generation module is used to extract physiological features from measured physiological indicator data and generate measured physiological feature data. The energy index set generation module is used to input measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data into the physiological stage sub-model of the physiological cycle model to generate energy index sets for different physiological cycles. The transition probability generation module is used to generate transition probabilities based on the physiological cycle characteristic data of the target female users. The transition probability is the evolution pattern of the physiological cycle of the target female users over time. The target menstrual cycle occurrence probability generation module is used to perform dynamic correlation analysis on the menstrual cycle status of target female users based on the energy index set and transition probability of different menstrual cycles, and generate the target menstrual cycle occurrence probability.

[0015] Optionally, following the dynamic correlation analysis unit, the female physiological signal analysis device further includes: The body temperature data acquisition unit is used to acquire short-term and long-term body temperature data of the target female user. The temperature jump event detection unit is used to detect temperature jump events based on the temperature change sequence within a preset short-term time window using short-term body temperature data. The short-term correction factor generation unit is used to generate a short-term correction factor based on the magnitude of the temperature jump event and the duration after the jump. The high / low characteristic temperature interval division unit is used to divide high / low characteristic temperature intervals based on the temperature of at least one complete historical period in long-term body temperature data. The long-term correction factor generation unit is used to assess the periodic stability of the target female user and generate a long-term correction factor based on the statistical distribution characteristics of the duration and temperature difference amplitude of the high temperature characteristic interval in multiple historical cycles in long-term body temperature data. The correction direction adjustment unit is used to enhance the correction strength when the correction directions of the short-term correction factor and the long-term correction factor are consistent, and to reduce the weight of the short-term correction factor when the correction directions are inconsistent. The time correction unit is used to perform time correction processing on the probability of occurrence of the target physiological cycle through short-term correction factor, long-term correction factor and correction direction.

[0016] Optionally, after the time correction unit, the steps of the female physiological signal analysis device may further include: The user behavior information acquisition unit is used to acquire individual interaction information and sleep quality information of the target female users. The individual state correction factor generation unit is used to generate individual state correction factors based on interaction information. The compensation factor generation unit is used to generate compensation factors for short-term correction factors based on sleep quality information. The individual state / behavior modification unit is used to modify the probability of occurrence of the target physiological cycle based on the individual state modification factor and the compensation factor.

[0017] Optionally, the energy index set generation module specifically includes: Generate time nodes for physiological characteristic changes based on the characteristics of physiological cycle changes of the target to be measured; The measured physiological characteristic data are divided into time intervals based on the time nodes of physiological characteristic changes; Based on the physiological characteristic distribution data, the measured physiological characteristic data divided into time intervals, and the physiological characteristic reference data, the corresponding physiological stage sub-model is input to generate a set of energy indicators for different physiological cycles.

[0018] Optionally, the physiological cycle model building unit specifically includes: Based on the target female users’ body indicators, physiological characteristic reference data and physiological characteristic distribution data for each menstrual cycle in different time periods are obtained from the database. The physiological characteristic reference data includes the target female users’ historical menstrual cycle data. A menstrual cycle model for the target female user is constructed based on the physiological characteristic data of each menstrual cycle, the reference data of physiological characteristics of each menstrual cycle in different time periods, and the distribution data of physiological characteristics. The menstrual cycle model contains at least two related physiological stage sub-models.

[0019] Optionally, after the step of acquiring physiological indicator data at different time points through the smart ring, and before the step of extracting physiological features from the physiological indicator data based on different physiological cycles and generating physiological feature data for each physiological cycle, the female physiological signal analysis device further includes: The preprocessing unit is used to preprocess the physiological index data, including noise reduction, outlier removal, and smoothing.

[0020] Thirdly, embodiments of this application provide a female physiological signal analysis device for a smart ring, comprising: Processor, memory, input / output units, and bus; The processor is connected to memory, input / output units, and a bus; The memory stores a program, which the processor calls to execute, as in the first aspect and any optional female physiological signal analysis method of the first aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing a program that, when executed on a computer, performs the first aspect and any optional female physiological signal analysis method of the first aspect.

[0022] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: In this application, firstly, various physiological indicator data of the target female user are collected in advance through a smart ring during different menstrual cycles. Specifically, this includes body temperature data and vital sign data correlated with the menstrual cycle being measured. In particular, physiological indicator data needs to be collected at different time points. The physiological indicator data includes at least two types of vital sign data related to the menstrual cycle, including body temperature data or temperature data estimated based on heat flux. Next, physiological features are extracted from the collected physiological indicator data based on different menstrual cycles. That is, physiological features are selectively extracted for several menstrual cycles to be measured, ensuring that the physiological feature data corresponding to each menstrual cycle is correlated. Then, a sub-model of the physiological stage is constructed based on the correlation between each menstrual cycle and its corresponding physiological feature data. This construction requires the assistance of reference physiological feature data and physiological feature distribution data. Next, a set of physiological energy indicators is generated for each physiological cycle based on the physiological cycle model. The set of physiological energy indicators expresses the degree of difference between the physiological feature data used as training data and the reference data. Each physiological stage sub-model is trained based on the set of physiological energy indicators. The internal weights of the physiological stage sub-models are adjusted so that the physiological energy indicators output by the physiological stage sub-models meet the preset conditions, making the physiological stage sub-models more consistent with the physiological feature data of the target female users, thereby improving the accuracy of physiological cycle prediction based on physiological signals. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application, 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of the female physiological signal analysis method of the smart ring in this application; Figure 2 This is a schematic diagram of the dynamic correlation analysis method used in this application; Figure 3 This is a schematic diagram of the method for time-correcting the probability of occurrence of the target physiological cycle according to this application; Figure 4 This is another schematic diagram illustrating the method of this application for modifying the probability of occurrence of a target physiological cycle in an individual state / behavior. Figure 5 A schematic diagram of the method for generating a set of energy indices for this application; Figure 6 A schematic diagram of the method for constructing a physiological cycle model for this application; Figure 7 This is a schematic diagram of the preprocessing method for the physiological index data collected in this application; Figure 8 This is a schematic diagram of the female physiological signal analysis device of the smart ring in this application; Figure 9 This is another schematic diagram of the female physiological signal analysis device of the smart ring in this application. Detailed Implementation

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0026] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0027] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0028] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0029] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0030] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0031] This application discloses a method and related device for analyzing female physiological signals in a smart ring, which is used to improve the accuracy of menstrual cycle prediction based on physiological signals.

[0032] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0033] The method described in this application can be applied to servers, devices, terminals, or other devices with logical processing capabilities; therefore, this application does not limit its application. For ease of description, the following description uses a terminal as the executing entity.

[0034] Please see Figure 1 This application provides an embodiment of a method for analyzing female physiological signals using a smart ring, comprising: 101. During different menstrual cycles of the target female users, physiological indicator data at different time points are collected through a smart ring. The physiological indicator data includes at least two types of vital signs data related to the menstrual cycle, including body temperature data. Body temperature data includes body temperature data or body temperature data estimated based on heat flux. The terminal first collects vital signs data from the target female user during different menstrual cycles. These cycles include at least the luteal phase, follicular phase, and ovulation phase. These cycles not only have a relatively stable order of occurrence but may also overlap in time. Conventional smart wearable devices typically only use specific temperature characteristics for analysis, inputting internal temperature data into a pre-set detection model to predict the menstrual cycle based on changes in this data. This type of detection model does not consider the individual characteristics of the target female user; it uses a fixed, standard model for uniform analysis, leading to significant detection errors. Furthermore, the correlation between each menstrual cycle and numerous vital sign data points varies. Conventional detection models do not consider these correlations, further reducing the accuracy of the detection model.

[0035] Traditional smart rings typically only collect a single physiological signal (such as heart rate), which cannot distinguish gender-specific physiological characteristics. Furthermore, they often use static thresholds or simple time statistical prediction methods, failing to clearly describe the dynamic transition relationship between the follicular phase, menstrual phase, ovulation phase, and luteal phase in a woman's menstrual cycle, thus failing to provide women with deeper insights.

[0036] To address the aforementioned issues, in this embodiment, the terminal needs to collect long-term, continuous data from the user via a smart ring. The collected data includes at least core body temperature data (internal temperature data, specifically internal temperature data including body temperature data or temperature data including internal temperature estimated based on heat flux), and may further include physiological indicators such as sleep duration, sleep quality, heart rate, heart rate variability (HRV), and stress level. Based on the sensitivity of different physiological cycles to vital signs data, the data is collected uniformly via the smart ring. In the subsequent model building process, physiological indicator data with higher sensitivity for different physiological cycles is selected as the basis for construction.

[0037] 102. Based on different physiological cycles, physiological feature extraction is performed on the collected physiological indicator data to generate physiological feature data for each physiological cycle; In this embodiment, after the target female user wears the smart ring, physiological indicators are collected, and combined with other smart devices for auxiliary classification. While determining the collection time of physiological indicators, it is also necessary to confirm the activity status of the target female user during the collection process, so as to reduce data noise.

[0038] Specifically, in this embodiment, the terminal extracts features from the collected physiological indicator data based on different physiological cycles. For example, the luteal phase is more sensitive to the body temperature difference data during sleep, so the terminal needs to filter the collected physiological indicator data to select the body temperature data of the target female user during the sleep period. In this collection stage, the time node is determined, and the body temperature data collected at adjacent time nodes are subtracted to generate the body temperature difference.

[0039] The terminal needs to process the physiological indicator data collected by the smart ring, including locating user activity within a specific area and extracting data, in order to obtain physiological characteristic data that better matches the sensitivity of each physiological cycle.

[0040] In addition to adjusting the extraction method mentioned above, you can also choose to use a feature extraction model for extraction, which is not limited here.

[0041] 103. Construct a physiological cycle model based on the physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data for each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. In the current technology, there is a lack of specific detection models for women's special physiological periods. Existing algorithms have low accuracy in predicting their respective physiological cycles. The algorithm design is mainly aimed at half of the population and usually models are specifically designed for women's unique endocrine rhythms and physiological cycles, which cannot effectively reflect the true state of women at different physiological stages.

[0042] To address the aforementioned issues, in this embodiment, the terminal constructs a menstrual cycle model based on the target female user's internal temperature data. This model abstracts the female menstrual cycle into a finite number of discrete physiological stages, including the follicular phase and the luteal phase, among others. A probabilistic state-space modeling method is employed, treating the female menstrual cycle as a latent physiological state process that evolves over time. This allows the trained menstrual cycle model to dynamically estimate these latent physiological states.

[0043] In this embodiment, the terminal needs to obtain physiological feature reference data and physiological feature distribution data from the database. When the target female user first wears the smart ring to collect physiological indicators and build a model for her menstrual cycle, the terminal obtains physiological feature reference data and physiological feature distribution data from the reference database. After each training of the menstrual cycle model, the extracted physiological feature data is used to fuse or directly replace the physiological feature reference data, thereby updating the physiological feature reference data. Then, physiological feature distribution data is generated based on the updated physiological feature reference data.

[0044] The physiological characteristic reference data is physiological characteristic data extracted from a large number of female users' physiological indicator data. As the initial reference data, the physiological characteristic distribution data is the mean and standard deviation of these data.

[0045] The following example illustrates the construction of the physiological stage sub-model: The discrete set of physiological states is defined as the luteal phase. Body temperature and heart rate data are processed to extract temperature change features. Below are the physiological feature data (temperature change features) corresponding to the body temperature data:

[0046] Where k is the time window. This represents the lowest temperature rise in body temperature within the time period Q corresponding to the current time node t.

[0047] Next, physiological characteristic distribution data will be obtained from the database. It is assumed that during the luteal phase, changes in body temperature data follow a distribution. Heart rate during the luteal phase follows a distribution. Define what changes should theoretically be observed if we are in a certain stage.

[0048] The constructed energy function (physiological stage sub-model) is shown below:

[0049] in, This is a sub-model (energy function) of the physiological stages of the luteal phase. The energy weight corresponding to body temperature. The energy weight corresponding to heart rate. This refers to the time period during which body temperature data was collected. The data collection period for heart rate is defined as follows: during the luteal phase, body temperature data is more sensitive to specific activities or time periods, meaning the patterns of change are more pronounced. The same applies to heart rate; however, the time periods for both may not be the same. t represents a specific time point within that time period. These are physiological characteristic data corresponding to body temperature data. This is a reference data for physiological characteristics corresponding to body temperature data (corresponding to time periods). This refers to the distribution data of physiological characteristics (temperature standard deviation) corresponding to the body temperature data. This refers to the physiological characteristic data corresponding to heart rate data. This is a reference data for the physiological characteristics corresponding to heart rate data. This represents the physiological characteristic distribution data (heart rate standard deviation) corresponding to the heart rate data. d represents other body indicators matching the target female user within the same time period, such as height, weight, and age group, used to select more suitable physiological characteristic reference data from the database.

[0050] The above method generates a physiological stage sub-model for each menstrual cycle, and each physiological stage sub-model is constructed based on the collected data of the target female users.

[0051] In this embodiment, an implicit state model of the female menstrual cycle is constructed, which includes at least the follicular state, menstrual transition state, stable ovulation state, and luteal decay state. Multimodal physiological signals (physiological indicator data collection) are used to infer the current physiological state, and different prediction strategies are employed in different states to achieve switching of prediction model parameters or prediction logic based on the physiological state.

[0052] 104. Generate a set of physiological energy indicators for each physiological cycle based on the physiological cycle model; 105. Train the physiological cycle model based on the set of physiological energy indicators; The terminal generates physiological energy indicators based on the physiological cycle model. , The smaller the value, the more closely the physiological phase sub-model matches the changes in the target female user's menstrual cycle. Specifically, during the N days of the luteal phase, data from sleep periods are selected each day to generate physiological energy indicators. These indicators are then used to train the menstrual cycle model, specifically by updating and iterating the weights, so that the trained physiological phase sub-model better reflects the changes in the target female user's menstrual cycle.

[0053] 106. Dynamic correlation analysis was performed on the measured physiological index data of the target female users through the trained physiological stage sub-model.

[0054] Once all physiological stage sub-models have been trained, or the training data has been completed, the terminal uses the trained physiological stage sub-models to perform dynamic correlation analysis on the measured physiological indicator data of the target female user. The specific analysis method will be explained in detail in subsequent embodiments.

[0055] In this embodiment, firstly, various physiological indicator data of the target female user during different menstrual cycles are collected in advance using a smart ring. Specifically, this includes body temperature data and vital sign data correlated with the menstrual cycle being tested, especially data collected at different time points. Next, physiological features are extracted from the collected physiological indicator data based on different menstrual cycles. This means selectively extracting physiological features for each of the tested menstrual cycles, ensuring that the physiological feature data corresponding to each tested menstrual cycle is correlated. Next, a physiological stage sub-model is constructed based on the correlation between each tested menstrual cycle and its corresponding physiological feature data, aided by reference physiological feature data and physiological feature distribution data. Then, a set of physiological energy indicators is generated for each menstrual cycle based on the menstrual cycle model. This set of physiological energy indicators expresses the degree of difference between the physiological feature data used as training data and the reference data. Each physiological stage sub-model is trained based on the physiological energy indicator set. The internal weights of the physiological stage sub-model ensure that the output physiological energy indicators meet preset conditions, making the physiological stage sub-model more consistent with the physiological feature data of the target female user, thereby improving the accuracy of menstrual cycle prediction based on physiological signals.

[0056] Please see Figure 2 This application provides an embodiment of a method for dynamic correlation analysis, comprising: 201. Real-time collection of vital signs data of target female users through smart rings to generate initial measured data for different time periods; 202. Obtain daily activity information of target female users, filter the detection interval of the initial test data based on the daily activity information of target female users, and generate measured physiological index data; In existing technologies, many physiological signals are subject to detection interference. Certain activities can disrupt the analysis of some physiological signals, necessitating the incorporation of user activity events for noise reduction to improve accuracy. However, some physiological signals exhibit more pronounced changes under specific activities. Heart rate, skin temperature, and blood oxygenation signals are prone to abnormal fluctuations during exercise, emotional changes, and environmental changes. Existing technologies fail to incorporate dynamic weighted noise reduction based on real-world scenarios or dynamic filtering based on real-world activities, impacting the reliability of the results. Furthermore, existing technologies typically assume that the collected signals have equal reliability, failing to provide consistent evaluation of data quality across different time periods and physiological states. This leads to abnormal data interference unreasonably affecting the prediction results.

[0057] In this embodiment, after the terminal obtains the initial measured data of the target female user through the smart ring, since this data may have been collected over a long period of time, the user's activity status can be detected by other smart devices. This activity status can be actively provided by the target female user or detected by the smart wearable device; no limitation is made here. The initial measured data is filtered by intervals using the target female user's daily activity information to reduce the computational load in subsequent analysis stages. Furthermore, the target female user's daily activity information can be obtained first, and then the smart ring can be activated for data collection, reducing power consumption and excessive memory usage. The terminal uses the target female user's daily activity information to filter the initial measured data by detection intervals, generating measured physiological indicator data.

[0058] 203. Extract physiological features from the measured physiological index data to generate measured physiological feature data; 204. Input the measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data into the physiological stage sub-model of the physiological cycle model respectively to generate a set of energy indicators for different physiological cycles; Steps 203 and 204 are similar to steps 102 to 104 in the aforementioned embodiments, and will not be repeated here. It should be noted that the measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data need to be input into the sub-models of each physiological stage to generate energy indicators for the luteal phase, follicular phase, menstrual phase, etc., for subsequent analysis.

[0059] 205. Generate transition probabilities based on the physiological cycle characteristics data of target female users. The transition probabilities are the evolutionary patterns of the physiological cycle of target female users over time. 206. Based on the energy index set and transition probability of different physiological cycles, perform dynamic correlation analysis on the physiological cycle status of target female users to generate the probability of occurrence of the target physiological cycle.

[0060] In this embodiment, a transition probability is generated based on the physiological cycle characteristic data of the target female user. The transition probability is the evolution law of the physiological cycle time of the target female user. The transition probability can be personalized by inputting the user's physiological cycle characteristic data (cycle length, duration, etc.) into a preset model.

[0061] In this embodiment, the energy functions of different physiological cycles are used to characterize the matching degree between the observed physiological signals and the sub-models of each physiological stage, while the transition probability is used to constrain the evolution continuity of the physiological stages over time. Together, they constitute a dynamic estimation mechanism for the state of a woman's physiological cycle, which can further improve the accuracy of the probability of the occurrence of the target physiological cycle.

[0062] After determining the menstrual cycle by the probability of the target menstrual cycle, the measured physiological characteristic data can be used to update and replace the physiological characteristic reference data and physiological characteristic distribution data in the database. This allows for training in each menstrual cycle and makes the physiological stage sub-model more adaptable to the menstrual cycle changes of the target female users as the usage cycle increases.

[0063] Please see Figure 3 This application provides an embodiment of a method for time-correcting the probability of occurrence of a target physiological cycle, comprising: 301. Obtain short-term and long-term body temperature data of the target female users; In this embodiment, the consistency metric used in the previous embodiments is used to characterize the degree of matching between the physiological data observed at the current moment and the sub-models of each physiological stage. However, the female menstrual cycle has obvious temporal continuity and stage evolution characteristics, and relying solely on the consistency judgment at a single moment is insufficient to accurately reflect the true physiological state. Therefore, this embodiment further combines short-term physiological characteristics with long-term cycle evolution patterns to continuously model and predict the physiological state.

[0064] In practical applications, time-scale corrections, individual state corrections, and environmental / behavioral corrections can be made based on the basic prediction. In implementation, this step uses a basic physiological cycle model combined with a probability prediction model as the core prediction framework. The prediction results are corrected hierarchically across four dimensions: short-term body temperature changes, long-term cycle structure, external environmental disturbances, and internal health status. Specifically, short-term body temperature trends are used to correct the current cycle phase, long-term body temperature structure is used to correct cycle stability, user interaction information is used to incorporate environmental and emotional disturbances, and sleep and cardiovascular indicators are used to reflect the body's regulatory capacity, thereby dynamically updating the distribution parameters of menstrual probability or waiting time.

[0065] This embodiment provides a detailed description of time scale correction, which includes short-term time correction and long-term time correction.

[0066] 302. Detect temperature jump events based on the temperature change sequence within a preset short-term time window using short-term body temperature data; 303. Generate short-term correction factors based on the magnitude of the temperature jump event and the duration after the jump; In this embodiment, the short-term correction factor needs to analyze the temperature trend of the most recent small cycle (approximately a 30-day window). Specifically, the ovulation point is identified by monitoring short-term temperature changes, and the short-term correction factor is obtained based on the analyzed ovulation point. For example, after ovulation, the thermogenic effect (temperature spike) caused by progesterone secretion from the corpus luteum causes the basal body temperature to rise by approximately 0.3–0.5°C (the magnitude of the spike), lasting for approximately 12–14 days (the duration after the spike). Based on this, it is determined that the probability of ovulation increases and the probability of menstruation decreases. Conversely, when the temperature shows a downward trend, the probability of menstruation is determined to be increased. In addition to the above methods, the probability of a high temperature phase in a certain physiological cycle can also be calculated, and the temperature fluctuation analysis can be used to determine whether it meets the typical post-ovulation temperature rise criteria, thereby identifying the ovulation point and obtaining the short-term correction factor.

[0067] 304. Divide the high / low characteristic temperature ranges based on the temperature of at least one complete historical period in the long-term body temperature data; 305. Based on the statistical distribution characteristics of the duration and temperature difference of high temperature characteristic intervals in multiple historical cycles in long-term body temperature data, the periodic stability of target female users is evaluated, and a long-term correction factor is generated. 306. When the short-term correction factor and the long-term correction factor are in the same direction, the correction strength is increased; when the correction directions are not in the same direction, the weight of the short-term correction factor is reduced. 307. Time correction processing is performed on the probability of occurrence of the target physiological cycle by using short-term correction factors, long-term correction factors, and correction directions.

[0068] In this embodiment, long-term time correction uses machine learning to predict body temperature trends over the most recent quarter (approximately a 90-day window). The stability of the biphasic temperature structure across multiple short cycles reflects the health of luteal function. The aim is to utilize body temperature data over a longer timescale to divide it into high-temperature and low-temperature zones, corresponding to characteristic intervals of the luteal phase and other physiological cycles (such as the follicular phase), respectively. By analyzing the duration and temperature difference of these temperature zones, the status of luteal function is assessed, thereby assisting in judging the health of the cycle and the reliability of menstrual prediction. Based on the analysis, a long-term correction factor is provided. The terminal first acquires the temperature of at least one complete historical cycle from the long-term body temperature data. Next, each complete historical cycle is divided into high / low characteristic temperature intervals to generate high / low characteristic temperature intervals specific to the target female user. Then, based on the statistical distribution characteristics of the duration and temperature difference of the high-temperature characteristic intervals across multiple historical cycles in the long-term body temperature data, the cycle stability of the target female user can be assessed. A long-term correction factor is generated based on the magnitude of cycle stability. The long-term correction factor can be obtained by looking up a table, which is a statistical table of cycle stability and correction factors, and will not be elaborated here.

[0069] To avoid being misled by short-term noise, a reasonable range for long-term patterns is given. Historical data is used to determine whether the period is in the judgment stage. When the short-term and long-term judgments are consistent, that is, the correction direction is the same, it indicates that more weight needs to be added in that correction direction to increase the probability of menstruation through the long-term correction factor. If there is a deviation, that is, the correction direction is inconsistent, it means that long-term historical data should be used as the core benchmark, and the short-term correction weight needs to be reduced through the long-term correction factor.

[0070] After obtaining the short-term and long-term correction factors, the probability of the target physiological cycle is time-corrected using the short-term correction factor, the long-term correction factor, and the correction direction.

[0071] Please see Figure 4 This application provides an embodiment of a method for modifying the probability of occurrence of a target physiological cycle in an individual's state / behavior, comprising: 401. Obtain individual interaction information and sleep quality information of target female users; 402. Generate individual state correction factors based on interaction information; In this embodiment, the terminal integrates individual interaction information of the target female user with group statistical factors. That is, it collects information such as environment, emotion, and stress through the feedback channel of the target female user, and combines big data statistical patterns to use the relevant factors as input parameters (individual state correction factors) of the prediction model to correct the probability of the next menstrual period.

[0072] Specifically, a large amount of interaction information is collected during the design phase. This information is then fed into the menstrual cycle prediction model for training, and mapped to correction factors for physiological signals, which can amplify or suppress the current judgment. For example: high stress + abnormal body temperature fluctuations → reduce the confidence of "menstruation is imminent"; stable emotions + normal environment → enhance the credibility of body temperature signals.

[0073] 403. Generate compensating factors for short-term correction factors based on sleep quality information; 404. Adjust the probability of occurrence of the target physiological cycle based on the individual state correction factor and compensation factor.

[0074] The terminal can also generate compensation factors for short-term correction factors based on a comprehensive assessment of sleep and cardiovascular indicators. Specifically, it assesses the user's overall health status based on sleep duration, sleep quality, sleep stress level, and heart rate and HRV status at different times, and uses this as an important auxiliary feature for predicting the physiological cycle. For example, when a decline in sleep quality is detected, the short-term weight is reduced to avoid misjudgment under physiological stress.

[0075] After the terminal obtains the individual state correction factor and compensation factor, it can modify the individual state / behavior based on the probability of the target physiological cycle.

[0076] In this embodiment, a probabilistic state-space model is used to output the probability and prediction interval of the menstrual cycle to be measured, rather than the exact arrival date. Simultaneously, short-term and long-term cycle patterns are incorporated for fusion, balancing immediate response and long-term stability. This enables dynamic, probabilistic prediction of menstrual arrival time even in the presence of noise and individual differences, thereby significantly improving prediction accuracy, stability, and physiological interpretability.

[0077] Please see Figure 5 This application provides an embodiment of a method for generating a set of energy indices, comprising: 501. Generate time nodes for physiological characteristic changes based on the characteristics of physiological cycle changes of the target to be measured; 502. Divide the measured physiological characteristic data into time intervals according to the time nodes of physiological characteristic changes; 503. Input the physiological characteristic distribution data, the measured physiological characteristic data divided into time intervals, and the physiological characteristic reference data into the corresponding physiological stage sub-model to generate a set of energy indicators for different physiological cycles.

[0078] In this embodiment, the target physiological cycle to be measured refers to the physiological cycle that needs to be predicted, such as predicting the arrival date of menstruation. To predict this target physiological cycle, the currently collected data needs to be converted into physiological features and then input into different physiological stage sub-models. The energy indicators output by the different physiological stage sub-models are combined and analyzed. Because the stages of the physiological cycle alternate and overlap, energy indicators need to be generated for different time periods according to different physiological stage sub-models. The current physiological cycle node is analyzed based on the changing trends of the energy indicators of different models, thereby predicting the arrival time and probability of the target physiological cycle in the future.

[0079] The characteristics of physiological changes refer to the differences in physiological characteristics at different stages of the menstrual cycle. For example, in the early stage of ovulation, the target female user shows relatively small changes in the physiological signals of sleep heart rate and body temperature. However, as the period approaches ovulation (mid and late stages), the changes in these two physiological signals gradually increase and then slowly decrease. The degree of change in these two physiological characteristics exhibits unique features of the ovulation period. Therefore, it is necessary to divide the collected measured physiological characteristic data into time intervals and input the measured physiological characteristic data of different time intervals into the physiological stage sub-model to obtain multiple sets of energy indicators. Subsequently, the corresponding stage of the menstrual cycle is analyzed based on the changing patterns of these multiple sets of energy indicators.

[0080] Please see Figure 6 This application provides an embodiment of a method for constructing a menstrual cycle model, comprising: 601. Based on the target female users' body indicators, obtain physiological characteristic reference data and physiological characteristic distribution data for each menstrual cycle in different time periods from the database. The physiological characteristic reference data includes the target female users' historical menstrual cycle data. 602. Construct a menstrual cycle model for the target female user based on the physiological characteristic data of each menstrual cycle, the reference data of physiological characteristics of each menstrual cycle in different time periods, and the distribution data of physiological characteristics. The menstrual cycle model contains at least two related physiological stage sub-models.

[0081] In this embodiment, the terminal acquires the target female user's body indicator data, which may include height, weight, age, etc., and selects physiological characteristic reference data and physiological characteristic distribution data of similar groups and the target female user herself from the database using such body indicator data. The physiological characteristic reference data includes the target female user's historical menstrual cycle data, so that when using the physiological characteristic reference data and physiological characteristic distribution data to build the target female user's menstrual cycle model, the training time can be greatly reduced, and the menstrual cycle model can be more consistent with the target female user.

[0082] Please see Figure 7 This application provides an embodiment of a preprocessing method for physiological indicator collection data, comprising: 701. Preprocess the collected physiological index data, including noise reduction, outlier removal and smoothing.

[0083] In this embodiment, the collected physiological index data also needs to undergo a series of preprocessing steps to reduce interference. The preprocessing includes noise reduction, outlier removal, and smoothing to adjust or remove abnormal data.

[0084] Please see Figure 8 This application provides an embodiment of a female physiological signal analysis device for a smart ring, comprising: The menstrual cycle vital sign acquisition unit 801 is used to acquire physiological indicator data at different time points during different menstrual cycles of the target female user through a smart ring. The physiological indicator data includes at least two types of vital sign data related to the menstrual cycle, including body temperature data. The body temperature data includes body temperature data or body temperature data estimated based on heat flux. The preprocessing unit 802 is used to preprocess the physiological index data, including noise reduction, outlier removal and smoothing. The physiological feature data generation unit 803 is used to extract physiological features from the physiological indicator data collected based on different physiological cycles, and generate physiological feature data for each physiological cycle. The physiological cycle model construction unit 804 is used to construct a physiological cycle model based on the physiological characteristic data, physiological characteristic reference data and physiological characteristic distribution data of each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. Optionally, the menstrual cycle model construction unit 804 specifically includes: obtaining physiological characteristic reference data and physiological characteristic distribution data of each menstrual cycle in different time periods from the database based on the target female user's body indicators, wherein the physiological characteristic reference data includes the target female user's historical menstrual cycle data; constructing a menstrual cycle model of the target female user based on the physiological characteristic data of each menstrual cycle, the physiological characteristic reference data and physiological characteristic distribution data of each menstrual cycle in different time periods, wherein the menstrual cycle model contains at least two related physiological stage sub-models.

[0085] Physiological energy index set generation unit 805 is used to generate a physiological energy index set for each physiological cycle based on the physiological cycle model. The physiological cycle model training unit 806 is used to train the physiological cycle model based on a set of physiological energy indicators. The dynamic correlation analysis unit 807 is used to perform dynamic correlation analysis on the measured physiological index data of the target female users through the trained physiological stage sub-model; Optionally, the dynamic correlation analysis unit 807 specifically includes: The measured physiological index data generation module is used to obtain the daily activity information of the target female users, and to filter the detection interval of the initial measured data through the daily activity information of the target female users to generate measured physiological index data. The measured physiological feature data generation module is used to extract physiological features from measured physiological indicator data and generate measured physiological feature data. The energy index set generation module is used to input measured physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data into the physiological stage sub-model of the physiological cycle model to generate energy index sets for different physiological cycles. Optionally, the energy index set generation module specifically includes: Generate time nodes for physiological characteristic changes based on the characteristics of physiological cycle changes of the target to be measured; The measured physiological characteristic data and the physiological characteristic reference data are divided into time intervals according to the time nodes of physiological characteristic changes; Based on the physiological characteristic distribution data, the measured physiological characteristic data divided into time intervals, and the physiological characteristic reference data, the corresponding physiological stage sub-model is input to generate a set of energy indicators for different physiological cycles.

[0086] The transition probability generation module is used to generate transition probabilities based on the physiological cycle characteristic data of the target female users. The transition probability is the evolution pattern of the physiological cycle of the target female users over time. The target menstrual cycle occurrence probability generation module is used to perform dynamic correlation analysis on the menstrual cycle status of target female users based on the energy index set and transition probability of different menstrual cycles, and generate the target menstrual cycle occurrence probability.

[0087] The body temperature data acquisition unit 808 is used to acquire short-term and long-term body temperature data of the target female user. The temperature jump event detection unit 809 is used to detect temperature jump events based on the temperature change sequence within a preset short-term time window using short-term body temperature data. The short-term correction factor generation unit 810 is used to generate a short-term correction factor based on the magnitude of the temperature jump event and the duration after the jump. The high / low characteristic temperature interval division unit 811 is used to divide the high / low characteristic temperature interval based on the temperature of at least one complete historical period in the long-term body temperature data. The long-term correction factor generation unit 812 is used to assess the periodic stability of the target female user and generate a long-term correction factor based on the statistical distribution characteristics of the duration and temperature difference amplitude of the high temperature characteristic interval in multiple historical cycles in long-term body temperature data. The correction direction adjustment unit 813 is used to enhance the correction intensity when the correction directions of the short-term correction factor and the long-term correction factor are consistent, and to reduce the weight of the short-term correction factor when the correction directions are inconsistent. The time correction unit 814 is used to perform time correction processing on the probability of occurrence of the target physiological cycle through short-term correction factor, long-term correction factor and correction direction. User behavior information acquisition unit 815 is used to acquire individual interaction information and sleep quality information of the target female user; Individual state correction factor generation unit 816 is used to generate individual state correction factors based on interaction information; Compensation factor generation unit 817 is used to generate compensation factors for short-term correction factors based on sleep quality information; The individual state / behavior correction unit 818 is used to correct the individual state / behavior based on the individual state correction factor and the compensation factor to adjust the probability of occurrence of the target physiological cycle.

[0088] Please see Figure 9This application provides a female physiological signal analysis device for a smart ring, comprising: Processor 901, memory 902, input / output unit 903, and bus 904.

[0089] The processor 901 is connected to the memory 902, the input / output unit 903, and the bus 904.

[0090] The memory 902 stores a program, and the processor 901 calls the program to execute it, such as... Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 Methods for analyzing female physiological signals in Chinese.

[0091] This application provides a computer-readable storage medium on which a program is stored, and when the program is executed on a computer, it performs the following... Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 Methods for analyzing female physiological signals in Chinese.

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0093] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0094] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0095] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0096] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, 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, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for analyzing female physiological signals in a smart ring, characterized in that, include: During different menstrual cycles of the target female users, physiological indicator data at different time points are collected through a smart ring. The physiological indicator data includes at least two types of vital signs data related to the menstrual cycle, including body temperature data. The body temperature data includes body temperature data or body temperature data estimated based on heat flux. Physiological features are extracted from the collected physiological indicators based on different physiological cycles, and physiological feature data is generated for each physiological cycle. A physiological cycle model is constructed based on the physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data of each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. A set of physiological energy indicators is generated for each physiological cycle based on the physiological cycle model. The physiological cycle model is trained based on the set of physiological energy indicators; The physiological stage sub-model, once trained, is used to perform dynamic correlation analysis on the measured physiological index data of the target female users.

2. The method for analyzing female physiological signals according to claim 1, characterized in that, The steps of performing dynamic correlation analysis on the measured physiological index data of the target female user through the trained physiological stage sub-model specifically include: The smart ring collects the target female user's vital signs data in real time, generating initial measured data for different time periods. Obtain the daily activity information of the target female user, and use the daily activity information of the target female user to filter the detection interval of the initial measured data to generate measured physiological index data; Physiological features are extracted from the measured physiological index data to generate measured physiological feature data; The measured physiological characteristic data, the physiological characteristic reference data, and the physiological characteristic distribution data are respectively input into the physiological stage sub-model of the physiological cycle model to generate a set of energy indicators for different physiological cycles. A transition probability is generated based on the physiological cycle characteristic data of the target female user, and the transition probability is the time evolution law of the physiological cycle of the target female user. Based on the energy index set of different physiological cycles and the transition probability, a dynamic correlation analysis is performed on the physiological cycle status of the target female user to generate the probability of occurrence of the target physiological cycle.

3. The method for analyzing female physiological signals according to claim 2, characterized in that, After the step of performing dynamic correlation analysis on the measured physiological index data of the target female user through the trained physiological stage sub-model, the female physiological signal analysis method further includes: Obtain short-term and long-term body temperature data of the target female user; Based on the temperature change sequence of the short-term body temperature data within a preset short-term time window, detect temperature jump events; A short-term correction factor is generated based on the magnitude of the temperature jump and the duration after the jump. The high / low characteristic temperature range is divided based on the temperature of at least one complete historical period in the long-term body temperature data; Based on the statistical distribution characteristics of the duration and temperature difference of high temperature characteristic intervals in multiple historical cycles in the long-term body temperature data, the periodic stability of the target female user is evaluated, and a long-term correction factor is generated. When the short-term correction factor and the long-term correction factor are in the same direction, the correction strength is increased; when the correction directions are not in the same direction, the weight of the short-term correction factor is reduced. The probability of occurrence of the target physiological cycle is time-corrected using the short-term correction factor, the long-term correction factor, and the correction direction.

4. The method for analyzing female physiological signals according to claim 3, characterized in that, After the step of performing time-correction processing on the probability of occurrence of the target menstrual cycle using the short-term correction factor, the long-term correction factor, and the correction direction, the steps of the female physiological signal analysis method further include: Obtain the individual interaction information and sleep quality information of the target female user; Generate individual state correction factors based on interaction information; A compensation factor is generated for the short-term correction factor based on the sleep quality information; The probability of occurrence of the target physiological cycle is modified based on the individual state correction factor and the compensation factor.

5. The method for analyzing female physiological signals according to claim 2, characterized in that, The step of inputting the measured physiological characteristic data, the physiological characteristic reference data, and the physiological characteristic distribution data into the physiological stage sub-model of the physiological cycle model to generate energy index sets for different physiological cycles specifically includes: Generate time nodes for physiological characteristic changes based on the characteristics of physiological cycle changes of the target to be measured; The measured physiological characteristic data are divided into time intervals based on the time nodes of the changes in the physiological characteristics. The physiological characteristic distribution data, the measured physiological characteristic data divided into time intervals, and the physiological characteristic reference data are input into the corresponding physiological stage sub-model to generate a set of energy indicators for different physiological cycles.

6. The method for analyzing female physiological signals according to any one of claims 1 to 5, characterized in that, The steps for constructing a physiological cycle model based on physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data for each physiological cycle specifically include: Based on the target female user’s body indicators, physiological characteristic reference data and physiological characteristic distribution data for each physiological cycle in different time periods are obtained from the database. The physiological characteristic reference data includes the target female user’s historical physiological cycle data. The target female user's menstrual cycle model is constructed based on the physiological characteristic data of each menstrual cycle, the physiological characteristic reference data of each menstrual cycle in different time periods, and the physiological characteristic distribution data. The menstrual cycle model contains at least two related physiological stage sub-models.

7. The method for analyzing female physiological signals according to any one of claims 1 to 5, characterized in that, After the step of acquiring physiological indicator data at different time points through a smart ring, and before the step of extracting physiological features from the acquired physiological indicator data based on different physiological cycles to generate physiological feature data for each physiological cycle, the female physiological signal analysis method further includes: The collected physiological index data are preprocessed, including noise reduction, outlier removal, and smoothing.

8. A female physiological signal analysis device for a smart ring, characterized in that, include: The menstrual cycle vital sign acquisition unit is used to acquire physiological indicator data at different time points during different menstrual cycles of the target female user through a smart ring. The physiological indicator data includes at least two types of vital sign data related to the menstrual cycle, including body temperature data. The body temperature data includes body temperature data or body temperature data estimated based on heat flux. The physiological feature data generation unit is used to extract physiological features from the physiological indicator data collected based on different physiological cycles, and generate physiological feature data for each physiological cycle. A physiological cycle model construction unit is used to construct a physiological cycle model based on physiological characteristic data, physiological characteristic reference data, and physiological characteristic distribution data for each physiological cycle. The physiological cycle model contains at least two related physiological stage sub-models. A physiological energy index set generation unit is used to generate a physiological energy index set for each physiological cycle based on the physiological cycle model. A physiological cycle model training unit is used to train the physiological cycle model based on the set of physiological energy indicators. The dynamic correlation analysis unit is used to perform dynamic correlation analysis on the measured physiological index data of the target female user through the trained physiological stage sub-model.

9. A female physiological signal analysis device for a smart ring, characterized in that, The device includes: Processor, memory, input / output units, and bus; The processor is connected to memory, input / output units, and a bus; The memory stores a program, which the processor calls to execute the female physiological signal analysis method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium having a program stored thereon, which, when executed on a computer, performs the female physiological signal analysis method as claimed in any one of claims 1 to 7.

Images