Health monitoring system based on pulse signal multi-domain feature extraction and deep learning

Abstract

The invention discloses a health monitoring system based on pulse signal multi-domain feature extraction and deep learning, and the system is characterized in that a multi-dimensional feature extraction module employs an FMCW radar and a sensor device to collect and obtain pulse signals, carries out the multi-level processing and feature extraction of the obtained pulse signals, and extracts features related to a health state; the feature association mapping module matches and associates the pulse signal features of different frequency bands with the health state; the state recognition module automatically recognizes and classifies pulse signal features and health states based on a deep learning model; and the health information feedback module analyzes the health state of the user according to the health state recognition result of the state recognition module, provides health feedback and generates corresponding personalized health guidance information. By using the system, comprehensive analysis of health conditions can be realized, continuous and highly-targeted health tracking and evaluation services can be provided, and modern personalized and intelligent traditional Chinese medicine health monitoring and management requirements can be met.

Description

Technical Field

[0001] The present invention relates to the technical field of traditional Chinese medicine health monitoring and artificial intelligence, and particularly relates to a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning for assisting in the analysis and evaluation of traditional Chinese medicine health. Background Art

[0002] At present, as an important part of the four diagnostic methods in traditional Chinese medicine, pulse diagnosis mainly relies on the experience of traditional Chinese medicine practitioners and fingertip sensations to judge and evaluate the health status of patients in traditional practice. However, due to subjective factors and the lack of standardization, etc., the accuracy and repeatability of traditional pulse diagnosis are limited. In addition, most of the existing instrument systems rely on time-domain analysis, can only extract the surface features of the pulse waveform, fail to deeply analyze information such as frequency components and phase changes, and do not even combine the systematic scientific thinking of amplification and feedback, resulting in insufficient and comprehensive interpretation of the health status of zang-fu organs, and being unable to mine key information such as traditional Chinese medicine syndrome elements in physical features, making it difficult to meet the refined monitoring requirements for the functional status of zang-fu organs, and also difficult to accurately identify the changing trends of health status. Moreover, it restricts the improvement of the diagnosis and treatment efficiency of integrated traditional Chinese and Western medicine. The current technology lacks the adaptability to individual differences and is difficult to perform personalized health analysis based on the unique physiological characteristics of users. This limits the application of the technology in personalized health management, making it difficult to provide continuous and highly targeted health tracking and evaluation services, and unable to meet the modern personalized and intelligent health monitoring and management needs. Summary of the Invention

[0003] In view of this, the present invention provides a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning that solves at least the above partial technical problems. Through multi-domain feature extraction of pulse signals and deep learning technology, the system overcomes the limitations of the prior art, facilitates providing a scientific basis for health assessment, realizes accurate and comprehensive analysis of health conditions, and provides continuous and highly targeted health tracking and evaluation services, which is beneficial to meeting the modern personalized and intelligent traditional Chinese medicine health monitoring and management needs.

[0004] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning, the system includes: a multi-dimensional feature extraction module, a feature association mapping module, a state recognition module, and a health information feedback module, wherein:

[0006] The multi-dimensional feature extraction module acquires a pulse signal by using an FMCW radar and a sensor device, and performs multi-level processing and feature extraction on the acquired pulse signal to extract features related to the health status.

[0007] The feature association mapping module is used to match and associate the features of pulse signals in different frequency bands with the health status;

[0008] The status recognition module automatically recognizes and classifies the features of pulse signals and the health status based on a deep learning model;

[0009] The health information feedback module is used to analyze the user's health status and provide health feedback according to the health status recognition result of the status recognition module, and generate corresponding personalized health guidance information.

[0010] In an optional embodiment, the system further includes: a status tracking and trend analysis module, which is used to perform time series analysis on the pulse signal data and draw a health trend graph changing with time.

[0011] In an optional embodiment, the system further includes: a personalized baseline measurement module, which is used to establish an initial reference baseline for the user's health status, and give priority to comparing with the initial reference baseline when analyzing the user's each detection result to judge abnormal health changes.

[0012] In an optional embodiment, the system further includes: an AR display module, which uses augmented reality technology to present the information collected by the FMCW radar and the sensor device, the health feedback provided by the health information feedback module and the generated personalized health guidance information, and the health trend graph drawn by the status tracking and trend analysis module on the real-world view.

[0013] In an optional embodiment, the operating frequency range of the FMCW radar covers 24 GHz to 77 GHz; the sensor device includes: an infrared sensor and a pressure sensor, where: the infrared sensor collects the optical waveform information of the pulse by detecting the change in light absorption caused by blood flow; the pressure sensor directly captures the pressure change caused by the pulse fluctuation through mechanical contact.

[0014] In an optional embodiment, the multi-dimensional feature extraction module performs multi-level processing and feature extraction on the acquired pulse signal, specifically including: time-domain analysis, frequency-domain analysis and non-linear feature analysis, where:

[0015] In time-domain analysis, the basic features of the pulse signal are extracted, including the amplitude, peak, valley of the waveform and the pulse transit time;

[0016] In frequency-domain analysis, the pulse signal is transformed from the time domain to the frequency domain, its frequency decomposition is carried out, and the energy distribution of different frequency bands is extracted;

[0017] In non-linear feature analysis, the entropy value analysis and Lyapunov exponent method are used to evaluate the dynamic complexity of the pulse signal and the stability of the viscera function.

[0018] In an alternative embodiment, in the frequency-domain analysis, first, the pulse signal is transformed into frequency-domain information through Fourier transform to obtain the energy distribution of the signal at different frequencies; then, it is restored to the time-domain signal through inverse Fourier transform, and the Hilbert transform is used to generate an analytic signal; finally, the analytic signal is combined with the original signal to obtain a signal containing amplitude and phase information; the calculation formulas include:

[0019]

[0020] In the formula, X(f) represents the signal spectrum after Fourier transform of the time-domain signal x(t); f represents the frequency variable; t represents the time variable, which is the time position of the signal in the time domain; j represents the imaginary unit; e represents the complex exponential function; X n (t) represents the frequency-domain information obtained after Fourier transform of the signal x n (t), that is, the spectral representation of the signal, x n (t) represents the pulse signal obtained after preprocessing the original pulse signal x(t); represents the analytic signal generated by the Hilbert transform, τ represents the time delay or lag, and is used as the virtual integration variable in the Hilbert transform process; z(t) represents the analytic signal The signal combined with the signal x(t); A(t) represents the instantaneous amplitude information of the signal; φ(t) represents the instantaneous phase information of the signal.

[0021] In an alternative embodiment, the feature correlation mapping module includes: a frequency-domain and zang-fu function correlation mapping unit, an amplitude and fluctuation amplitude analysis unit, and a waveform feature database, where:

[0022] The frequency-domain and zang-fu function correlation mapping unit is used to associate the pulse signal characteristics in different frequency bands with the zang-fu function states, and analyze the zang-fu states according to the fluctuations at specific frequencies;

[0023] The amplitude and fluctuation amplitude analysis unit is used to analyze the envelope of the pulse signal after Hilbert transform, obtain the instantaneous amplitude and phase information, judge the strength of the zang-fu functions through the change in the fluctuation amplitude of the amplitude, and observe the flow state of qi and blood through the dynamic change of the phase;

[0024] The waveform feature database is used to store the pulse diagnosis data and pulse characteristics, and provide a pattern reference for judging the zang-fu function states; the database contains the standard pulse signal waveform data in various constitutions and health states, and has diverse qi-blood and zang-fu state patterns.

[0025] In an alternative embodiment, the state recognition module uses the ResNet101 deep learning model to automatically recognize and classify the pulse signal features and health status through layer-by-layer convolution.

[0026] In an alternative embodiment, when analyzing the user's health status, the health information feedback module also comprehensively evaluates by combining the user's pulse information, breathing rate, and body temperature information. The evaluation method is as follows:

[0027] Q = α0 + α1|F - F′| + α2|B - B′| + α3|T - T′|

[0028] In the formula, Q is the health index, representing the health status value after comprehensive evaluation; α0 is the constant term, representing the basic health index when there is no deviation; α1 represents the weight coefficient of the pulse deviation; F represents the actual pulse value, F' represents the pulse standard value; α2 represents the weight coefficient of the breathing rate deviation, B represents the actual breathing rate, B' represents the breathing rate standard value in the ideal health state; α3 represents the weight coefficient of the body temperature deviation, T represents the current body temperature value, and T' represents the body temperature standard value.

[0029] Compared with the prior art, the present invention provides a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning. The system overcomes the limitations of the prior art through multi-domain feature extraction and machine learning techniques. The present invention has at least the following beneficial effects:

[0030] More comprehensive information acquisition: Through the combination of time domain, frequency domain, and short-time transformation, the system can not only capture the intensity information of the pulse wave but also extract the instantaneous energy distribution and phase changes of the internal organs, facilitating more comprehensive, in-depth, and accurate health monitoring and analysis.

[0031] Innovative multi-domain signal processing method: Using the multiple combinations of wavelet, Fourier, and Hilbert transforms, a multi-dimensional and accurate pulse signal analysis system is formed, which has significant innovation in multi-domain signal processing.

[0032] Accurate correlation between internal organ health: Using frequency decomposition and feature mapping, the pulse signal is accurately correlated with the health status of the five internal organs and six hollow organs and traditional Chinese medicine syndrome elements, facilitating the provision of more targeted health assessment results.

[0033] Automation and personalization: Based on the machine learning model, the system realizes the automatic recognition of health status and can be adjusted according to the user's personalized baseline, facilitating the adaptation to the health needs of different users.

[0034] Dynamic monitoring and health feedback: The system long-term tracks the user's health status and gives timely warnings when there are health fluctuations, facilitating the provision of dynamic health monitoring management and feedback support for users.

[0035] Intelligent: The system adopts AR augmented reality, which can directly present information on the real-world view, enhancing the intelligence of traditional Chinese medicine health monitoring, making it more intuitive and efficient; it is conducive to meeting the modern personalized and intelligent requirements for traditional Chinese medicine health monitoring management.

[0036] Other features and advantages of the present invention will be described in the subsequent specification, and some of them will become obvious from the specification or be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in the written specification and the drawings.

[0037] The technical solution of the present invention will be further described in detail below with reference to the drawings and embodiments. Description of the Drawings

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0039] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention.

[0040] Figure 1 It is a schematic structural diagram of a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning provided for an embodiment of the present invention.

[0041] Figure 2 It is a schematic structural diagram of an electronic device provided for an embodiment of the present invention. Detailed Embodiments

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0043] In the description of the present invention, it should be noted that: in some processes described in the specification and drawings of the present application, there are multiple operations that appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. In addition, various serial numbers, etc., are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0044] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 As shown, the present invention provides a health monitoring system based on multi-domain feature extraction and deep learning of pulse signals, which uses a multi-dimensional signal processing method to extract subtle features in pulse signals (such as fingertips), combines machine learning models and feature association mapping, automatically analyzes the health status of the internal organs, and realizes the health assessment of the upper, middle and lower jiao and the extraction of TCM syndrome elements. The system realizes autonomic arterial diagnosis and health monitoring assessment through a combination of software and hardware. The system mainly includes: a multi-dimensional feature extraction module, a feature association mapping module, a state recognition module based on deep learning, and a health information feedback module; in some preferred embodiments, the system also includes: a state tracking and trend analysis module, a personalized baseline measurement module, and an AR display module.

[0046] In a specific embodiment, by adopting FMCW frequency modulation continuous wave technology, the present invention overcomes the inconvenience of contact measurement in traditional detection methods and realizes the non-sensing collection of pulse signals. FMCW can not only accurately measure the speed change of pulse waves, but also synchronously realize distance measurement, providing a reliable technical guarantee for the comprehensive acquisition of pulse signals. The present invention uses a multi-dimensional signal processing method to extract weak features in pulse signals to reveal potential health status information. Explore the feature extraction method in non-stationary and nonlinear pulse signals, explore the key features that can reflect the functional state of the five internal organs, and combine these features with the amplification feedback theory to achieve a comprehensive analysis of the user's health status and support the refined classification prediction of health status. The present invention uses specific frequency decomposition and signal processing technology to separate the different frequency components of the pulse signal, and explore the matching and association between each frequency component and the specific internal organ function. The goal is to identify the frequency bands related to the specific functional activities of the five internal organs in the pulse signal spectrum, so as to provide a scientific basis for the health assessment of the internal organs and realize the deep integration of pulse signals and the traditional Chinese medicine diagnosis and treatment system. The present invention improves the intelligence level and evaluation accuracy of pulse signal processing through deep learning algorithms (such as ResNet). Through the application of deep learning networks, it supports personalized health management and realizes accurate monitoring of human health status and yin and yang change trends, providing technical support for the modernization of traditional Chinese medicine diagnosis and treatment.

[0047] The working principle and specific implementation methods of each module of the system of the present invention are described in detail below:

[0048] 1. Multidimensional feature extraction module:

[0049] In the embodiments of the present invention, the hardware structure adopts a dual - sensor system, namely a millimeter - wave FMCW radar and a fingertip sensor. The two work together to ensure the comprehensive capture of pulse signals and data reliability. The fingertip sensor part includes an infrared sensor (OSRAM SFH 7050) and a pressure sensor (Honeywell FSS - SMT), and the two complement each other: the infrared sensor collects the optical waveform information of the pulse by detecting the change in light absorption caused by blood flow; the pressure sensor directly captures the pressure change caused by pulse fluctuations through mechanical contact. This dual - mode data acquisition can more comprehensively reflect the true biological information of the pulse. On the other hand, the millimeter - wave FMCW radar realizes non - contact pulse signal detection, avoiding the compression effect or measurement error that may be brought by traditional contact methods. The FMCW radar uses the BGT60 series millimeter - wave radar chips of Infineon, and the operating frequency range covers 24 GHz to 77 GHz. The relatively high operating frequency significantly improves its detection sensitivity to small displacements, and it can capture the micrometer - level micro - motion signals caused by pulse fluctuations on the skin surface of the human radial artery.

[0050] The working principle of the FMCW radar is to transmit a linearly frequency - modulated signal, whose frequency increases or decreases linearly with time. After the signal irradiates the target surface, it is reflected back to the radar receiving end. The micro - motion caused by the pulse leads to changes in the phase and frequency of the reflected signal. Through mixing processing, a beat - frequency signal (BeatFrequency) is generated, and the frequency difference of the beat - frequency signal is related to the target distance and speed information. The specific data acquisition process includes: the radar antenna continuously transmits the frequency - modulated signal, and at the same time receives the echo signal reflected by the skin surface; after the received signal is mixed with the transmitted signal, a beat - frequency signal containing the target distance and micro - motion information is generated. Through fast Fourier transform (FFT) for frequency - domain analysis of the beat - frequency signal, the distance between the human skin surface and the radar is accurately calculated, and environmental noise interference is removed. Subsequently, using the phase - difference method and I / Q demodulation technology, the phase change of the reflected signal is tracked in real time, and the micro - displacement information of the skin is further extracted, which is converted into time - series data, thereby reconstructing the pulse waveform.

[0051] After the data acquisition is completed, the system performs multi - level processing and feature extraction on the obtained pulse signals, specifically including three major parts: time - domain analysis, frequency - domain analysis, and non - linear feature analysis. First, in time - domain analysis, basic features of the pulse signal are extracted, such as the amplitude, wave peaks, wave valleys of the waveform, and pulse transit time (PTT). These features can intuitively reflect the undulating shape, strength change of the pulse, and the dynamic characteristics of blood flow, providing basic data for the assessment of heart rhythm and vascular elasticity.

[0052] In frequency-domain analysis, the pulse signal is transformed from the time domain to the frequency domain through Fourier transform, its frequency decomposition is carried out, the energy distribution of different frequency bands is extracted, and its correlation analysis is carried out with the functional activities of the five zang-organs in traditional Chinese medicine. In addition, to further capture the instantaneous dynamic changes in the pulse signal, wavelet transform is used to decompose the signal at multiple levels. Wavelet transform has good short-time analysis ability, which can decompose the pulse signal into detailed features of different time scales and frequency bands, and analyze the dynamic changes from low frequency to high frequency step by step, so as to reveal the pulse details closely related to the functional activities of the five zang-organs. In frequency-domain analysis, the specific calculation formulas are as follows:

[0053] The signal is transformed from the time domain to the frequency domain through Fourier transform to obtain the distribution of the signal at different frequencies:

[0054]

[0055] In the formula, X(f): the Fourier transform of the signal x(t) in the frequency domain, representing the spectrum of the signal. f: the frequency variable, representing different frequency components of the signal in the frequency domain. x(t): the time-domain signal, representing the signal that changes with time (such as pulse wave, pressure wave, etc.). t: the time variable, representing the time position of the signal in the time domain. j: the imaginary unit; e: the complex exponential function, representing the modulation process, through which the changes of the signal at different frequency components can be analyzed.

[0056] Then, the signal decomposition is carried out. In the frequency domain, X(f) is decomposed into different parts according to different frequency components. Through biological experiments, the frequency bands matching the twelve zang-fu organs are selected to extract a part of the signal, and then the inverse Fourier transform is performed on these different frequency components to convert them into time-domain signals:

[0057]

[0058] In the formula, X n (t) represents the frequency-domain information obtained after the signal x n (t) is Fourier-transformed, that is, the spectrum representation of the signal, and x n (t) represents a processed version of the signal x(t), usually the pulse signal obtained after time-domain analysis, filtering or other preprocessing.

[0059] Then, the Hilbert transform is carried out: restored to the time-domain signal through the inverse Fourier transform, and the analytic signal is generated using the Hilbert transform Analyze the instantaneous amplitude and phase changes to further reveal the dynamic health status of the twelve zang-fu organs:

[0060]

[0061] In the formula, τ represents the time delay or lag, which serves as a dummy integration variable in the Hilbert transform process. This variable is used for the integration processing of signals and the generation process of analytic signals, and is used to describe the relative changes of signals between different time points.

[0062] Construct the analytic signal z(t), which is the combination of the original signal and the Hilbert transform:

[0063]

[0064] Among them, the instantaneous amplitude of the envelope signal: The instantaneous phase of the signal:

[0065] In nonlinear feature analysis, the fuzzy Lyapunov exponent is introduced to solve the problem of noise sensitivity in the calculation of traditional Lyapunov exponents; to evaluate the dynamic complexity of pulse signals and the stability of the system. The calculation formula is:

[0066]

[0067] Among them: λ f(t) is the fuzzy Lyapunov exponent, which represents the local dynamic characteristics of the system at time t. It measures the rate of change of the system state and is used to judge the stability of the system. The larger the Lyapunov exponent, the more unstable the system dynamics; the smaller, the more stable the system. μ(·) represents the fuzzy membership function, which is used to handle the uncertainty of signals; Δt represents the time window or time step, and this value controls the time step or time sequence difference in the signal sequence. By changing the size of Δt, the system stability at different time scales can be analyzed. ∣x(t + Δt) - x(t)∣ represents the difference between the pulse signal at time t + Δt and time t, measuring the degree of change of the signal at this time step. This difference reflects the instantaneous change of the pulse signal. ∣x(t) - x(t - Δt)∣ represents the difference between the pulse signal at time t and time t - Δt, similarly measuring the degree of change of the pulse signal. ∑ i represents the summation over all time instants i. The summation here is a statistical operation on each pair of time differences t and t + Δt in the signal sequence.

[0068] Specifically, approximate entropy and sample entropy are used to calculate the complexity of pulse signals, revealing the regularity and randomness of the signals, and then evaluating the coordination degree of visceral function activities, especially the state of kidney function. The Lyapunov exponent judges the stability and self-excitation oscillation phenomenon of the visceral function system by analyzing the nonlinear dynamic characteristics of the pulse wave signal, providing a basis for the in-depth evaluation of the health state.

[0069] Combined with the results of time-domain, frequency-domain, and non-linear feature analysis, the system can effectively extract characteristic indicators closely related to the functional activities of the five internal organs, realizing multi-dimensional assessment and dynamic monitoring of the health status of the zang-fu organs.

[0070] Finally, to ensure the accuracy and reliability of the data, the system synchronously fuses the pulse waveform extracted by the FMCW radar with the infrared and pressure signals collected by the fingertip sensor. Preferably, a feature fusion method using the weighted average method is adopted to weight and combine the feature information of the pulse waveform extracted by the FMCW radar and the infrared and pressure signals collected by the fingertip sensor. The formula is:

[0071]

[0072] Where: w i is the weight of the characteristics of each signal source to obtain a more accurate pulse signal characteristic; n represents the number of signal source characteristics; after feature extraction and fusion processing, a pulse signal f(t) that combines the characteristics of multiple sensors is obtained, which has higher stability and accuracy. This fused signal can effectively reflect the change characteristics of the pulse signal, providing reliable data support for subsequent health monitoring and analysis.

[0073] Furthermore, the fused data undergoes filtering processing and baseline correction to remove external interference and noise, obtaining a more stable and accurate pulse signal.

[0074] 2. Feature Association Mapping Module:

[0075] In the embodiment of the present invention, the feature association mapping module consists of a frequency-domain and zang-fu function association mapping unit, an amplitude and fluctuation amplitude analysis unit, and a waveform feature database, aiming to reveal the deep relationship between the pulse waveform and the five zang-organs and six fu-organs through multi-dimensional feature association. The specific functions are as follows:

[0076] Frequency-domain and zang-fu function association mapping unit: Conducts frequency-domain analysis on the pulse signal and associates the characteristics of different frequency bands with the functional states of the five zang-organs and six fu-organs. According to traditional Chinese medicine theory, the characteristics of the liver meridian, kidney meridian, spleen meridian, lung meridian, etc. are usually more obvious in the low-frequency band, while the triple energizer meridian and small intestine meridian often show high-frequency band characteristics. Through this frequency feature association mapping, the system can analyze the zang-fu state based on the fluctuations of specific frequencies, providing an accurate frequency basis for health assessment.

[0077] Amplitude and fluctuation amplitude analysis unit: Analyzes the envelope of the pulse signal after Hilbert transform to obtain instantaneous amplitude and phase information. The system judges the strength of zang-fu functions through the change in the fluctuation amplitude of the amplitude and observes the flow state of qi and blood through the dynamic change of the phase. This unit can depict the decline and prosperity changes of the zang-fu organs in detail, especially capturing abnormalities in the dynamic balance of the state of qi, blood, yin, and yang, providing precise data support for health assessment.

[0078] Waveform Feature Database: This database is used to store a large amount of pulse diagnosis data and pulse characteristics, providing a pattern reference for the diagnosis of the five zang-organs and six fu-organs. The database contains standard waveform data under various constitutions and health states, and has diverse qi-blood and zang-fu state patterns. The system can identify the abnormal health state of the user through matching analysis with these standard waveforms and generate personalized health suggestions.

[0079] 3. State Recognition Module Based on Deep Learning:

[0080] In the embodiments of the present invention, the state recognition module based on deep learning adopts the ResNet101 deep learning algorithm to realize the automatic recognition and classification of pulse signal characteristics and health states. The specific implementation method is as follows:

[0081] ResNet101 Model Training Unit: This unit uses historical data of multi-dimensional features for model training. The model can identify the health states of different parts such as the upper jiao (heart and lung), middle jiao (spleen and stomach), lower jiao (liver, kidney, and bladder), and three yin and three yang.

[0082] Automatic Recognition and Classification Unit: The system automatically classifies the input data into the corresponding health states and outputs the evaluation results. In addition, based on the mapping relationship between waveforms and syndrome elements, the model associates the pulse waveforms and specific TCM syndromes with health states.

[0083] The present invention utilizes multi-domain feature data, combines deep learning algorithms, and realizes the accurate recognition of health states through layer-by-layer convolution. Based on the multi-dimensional feature mapping of qi-blood and zang-fu function states, the model can identify syndrome elements such as "qi deficiency" and "lung", and further form comprehensive syndromes such as "syndrome of lung qi deficiency". In the health monitoring system of the present invention, the state recognition module uses a deep learning model based on the theory of three yin and three yang to classify the detected physiological characteristics to accurately evaluate the human health state. Shaoyang triple energizer and Taiyin spleen earth form the Taiji, respectively dominating the basic temperature and humidity of the human body. The two interact with each other to generate nutrient qi, defensive qi, qi-blood, and spirit, maintaining life activities. Pathologically, yang deficiency of Shaoyang leads to the failure of yin essence to rise, easily causing heart and lung diseases; weakness of the spleen in Taiyin results in the failure of water and dampness to transform, triggering damp diseases. The system identifies states such as yang deficiency of Shaoyang and weakness of the spleen in Taiyin through the analysis of pulse conditions and health characteristics, and combines the laws of the ascending of the spring yang of Jueyin liver and the descending of the autumn yin of Yangming lung to identify the dynamic balance of Jueyin wind-fire and Yangming dryness-dampness, thereby judging the disease trend, such as jueyin heat relapse or internal injury dryness disease. In addition, the system combines the law of the extreme transformation of yin and yang between Taiyang and Shaoyin, analyzes the cold-heat complexity during the winter solstice and summer solstice, and identifies the health risks caused by the alternation of cold and heat.

[0084] 4. Health Information Feedback Module:

[0085] In the embodiments of the present invention, the health information feedback module includes: a health feedback unit and a conditioning suggestion unit, where:

[0086] Health feedback unit: This unit is used to analyze the user's viscera state based on the system detection results and provide accurate health feedback. The feedback content is refined to the qi-blood-yin-yang balance state of the upper jiao (heart, lungs), middle jiao (spleen, stomach), and lower jiao (liver, kidneys, bladder) triple energizer regions, and the specific functional states of the five zang-organs and six fu-organs are differentiated. The feedback content includes common traditional Chinese medicine syndromes such as heart yin deficiency, heart yang deficiency, heart qi deficiency, and lung qi deficiency, so that users can intuitively understand the health status of their respective viscera.

[0087] Conditioning suggestion unit: According to the differentiation results of the health feedback unit, the conditioning suggestion unit generates corresponding personalized health guidance. The suggestion content is specific to diet, lifestyle, and work and rest adjustments, and specific conditioning methods are provided for users according to the health states of different viscera. For example, for users with heart yin deficiency, the system will recommend diet conditioning for nourishing heart yin; for users with weak spleen and stomach, it will suggest diet choices for warming and tonifying the spleen and stomach and work and rest adjustments.

[0088] The specific feedback methods are as follows:

[0089] Upper jiao (heart, lungs) feedback: The upper jiao is mainly responsible for the distribution of qi. For example, heart yang deficiency will be manifested as weak heart palpitations, shortness of breath, etc. The feedback unit will prompt users that their cardiopulmonary function is not good, and there may be situations such as qi deficiency and yang deficiency. The conditioning suggestion unit will suggest that users do more deep breathing exercises, consume foods that nourish the heart and lungs (such as lotus seeds, tremella), and appropriately carry out cardiopulmonary function exercises.

[0090] Middle jiao (spleen, stomach) feedback: The middle jiao is mainly responsible for digestion and transportation. The functional state of the spleen and stomach corresponds to the source of qi and blood generation in traditional Chinese medicine theory. If it is detected that the middle jiao has the characteristics of spleen qi deficiency or stomach qi deficiency, the system will feedback the situation of weak spleen and stomach function. The conditioning suggestion will recommend that users add easily digestible and spleen-benefiting foods (such as Chinese yam, coix seed) to their diet and reduce the intake of cold and raw foods. At the same time, it is recommended to appropriately strengthen exercise and maintain a regular diet.

[0091] Lower jiao (liver, kidneys, bladder) feedback: The lower jiao corresponds to the excretory and reproductive functions and is related to the health of kidney yang and kidney essence. If it is detected that the lower jiao function is abnormal, such as kidney qi deficiency or kidney yin deficiency, the system will feedback the situation of insufficient kidney function, which may be accompanied by symptoms such as fatigue and soreness in the waist and knees. The conditioning suggestion will guide users to choose foods that tonify the kidney and strengthen the foundation (such as black sesame seeds, wolfberries), and appropriately adjust the work and rest to improve the energy state.

[0092] In the process of syndrome differentiation feedback of qi, blood, yin, and yang in the health monitoring system of the present invention, the health information feedback module refines the health status. If syndromes such as heart qi deficiency and kidney yin deficiency are detected, it will provide accurate syndrome differentiation in combination with the results of the classification model and give personalized conditioning suggestions. For example, for those with heart yang deficiency, warm tonifying conditioning is recommended, and for those with heart yin deficiency, nourishing yin to soothe the heart is suggested to achieve individualized traditional Chinese medicine health intervention.

[0093] In an optional embodiment, when analyzing the user's health status, the health information feedback module also comprehensively evaluates by combining the user's pulse information, respiratory rate, and body temperature information. The specific evaluation calculation is as follows:

[0094] Q = α0 + α1|F - F′| + α2|B - B′| + α3|T - T′|

[0095] In the formula, Q is the health index, representing the health status value after comprehensive evaluation; α0 is the constant term, representing the basic health index when there is no deviation; α1 represents the weight coefficient of the pulse deviation; F represents the actual pulse value, F' represents the pulse standard value; α2 represents the weight coefficient of the respiratory rate deviation, B represents the actual respiratory rate, B' represents the respiratory rate standard value in the ideal health state; α3 represents the weight coefficient of the body temperature deviation, T represents the current body temperature value, and T' represents the body temperature standard value.

[0096] By quantifying the health status, it further helps to identify and pay attention to health problems and guides the adoption of appropriate health intervention measures; through the combined analysis of multiple vital signs, a comprehensive assessment of the health status is achieved, so as to provide scientific data support for preventive health management and early disease diagnosis in combination with health trend prediction.

[0097] 5. Status Tracking and Trend Analysis Module:

[0098] In the embodiment of the present invention, the status tracking and trend analysis module includes: a time series analysis unit and a health warning unit, where:

[0099] Time Series Analysis Unit: By performing time series analysis on the pulse wave data, the system can draw a health trend graph of the visceral functions changing with time, which can clearly show the changes in the health status of the five zang-organs and six fu-organs in each region of the upper jiao (heart, lung), middle jiao (spleen, stomach), and lower jiao (liver, kidney, bladder). Through the trend graph, users can observe the fluctuations of the visceral functions. For example, in the case of long-term high stress or irregular work and rest, the heart and lung health may show a downward trend, and the system will visually present these changes.

[0100] Health warning unit: Based on the anomalies in the trend chart, such as obvious fluctuations or long-term decline trends in the state of a certain internal organ, the health warning unit generates corresponding reminders. Through the warning mechanism, users can timely pay attention to their own health status and take targeted adjustments, such as strengthening exercise or adjusting diet, to prevent further health problems.

[0101] 6. Personalized baseline measurement module:

[0102] In the embodiments of the present invention, the personalized baseline measurement module includes: a personalized baseline establishment unit and a personalized analysis unit, where:

[0103] Personalized baseline establishment unit: This unit allows users to record their personal baseline pulse characteristics during the first detection and establish an individualized initial reference baseline. This baseline includes the user's pulse waveform characteristics and the state of internal organs, which will serve as the basis for comparison in subsequent detections, enabling the system to more accurately judge changes in the health status. The core of the personalized baseline measurement module is to establish an individualized health baseline for each user to reduce judgment biases caused by ignoring individual differences in routine health detections. In this way, the system can more precisely adapt to the unique pulse characteristics of users and improve the accuracy of health status assessment. During the user's first use, the system will record the user's pulse data and establish a personalized health baseline. The baseline data includes the basic characteristics of the pulse wave in the time domain and frequency domain, which can not only reflect the overall state of the user's internal organs but also adapt to the individual pulse characteristics.

[0104] Personalized analysis unit: Based on the user's initial baseline data for subsequent analysis, the system can more accurately identify the health conditions that match the user's personal characteristics. This unit will give priority to comparing with the personalized baseline when analyzing the results of each user's detection, reducing errors caused by individual differences and improving the accuracy of health assessment. Personalized analysis: In each new detection, the system will compare the pulse characteristics with the baseline. Such a comparison can eliminate the noise caused by the user's own pulse characteristics, enabling the system to more quickly and accurately judge abnormal health changes and generate health suggestions that match the user's characteristics.

[0105] 7. AR display module:

[0106] In the embodiments of the present invention, the AR display module can be a terminal device such as AR glasses, a computer, or a tablet; using augmented reality technology to present information collected by the FMCW radar and sensor devices, health feedback provided by the health information feedback module and generated personalized health guidance information, and health trend charts drawn by the status tracking and trend analysis module, etc., on the real-world view.

[0107] This module, based on AR technology, allows users to maintain contact with the real world while providing additional information or functions, making it very suitable for application scenarios that require real-time information overlay. Users can obtain a more immersive data presentation experience without completely disconnecting from the real environment. AR can integrate virtual elements into the user's actual environment, providing a brand-new way of participation in the field of traditional Chinese medicine monitoring, enhancing the experience in real life without the need to completely break away from the current environment, helping to improve the intelligence of traditional Chinese medicine health monitoring, making it more intuitive and efficient, and facilitating the meeting of modern personalized and intelligent traditional Chinese medicine health monitoring management needs. Additionally, technologies such as VR display modules and conventional liquid crystal display methods can also be used for information presentation, which will not be elaborated here one by one.

[0108] From the description of the above embodiments, those skilled in the art can learn that the present invention proposes a health monitoring system based on multi-domain feature extraction of pulse signals and deep learning. This system overcomes the limitations of the prior art through multi-domain feature extraction and machine learning techniques. The present invention has the following advantages:

[0109] 1. More comprehensive information acquisition: Through the combination of time domain, frequency domain, and short-time transformation, the system can not only capture the intensity information of the pulse wave but also extract the instantaneous energy distribution and phase changes of the internal organs, facilitating a more comprehensive and in-depth health monitoring analysis.

[0110] 2. Innovative multi-domain signal processing method: Using multiple combinations of wavelet, Fourier, and Hilbert transforms, a multi-dimensional and accurate pulse signal analysis system is formed, which is significantly innovative in multi-domain signal processing.

[0111] 3. Accurate correlation between internal organ health: Using frequency decomposition and feature mapping, the pulse signal is accurately correlated with the health status of the five internal organs and six hollow organs and traditional Chinese medicine syndrome elements, facilitating the provision of more targeted health assessment results.

[0112] 4. Automation and personalization: Based on a machine learning model, the system automatically identifies the health status and can be adjusted according to the user's personalized baseline to adapt to the health needs of different users.

[0113] 5. Dynamic monitoring and health feedback: The system long-term tracks the user's health status and issues timely warnings when there are health fluctuations, facilitating the provision of dynamic health monitoring management and feedback support for users.

[0114] 6. Intelligence: The system adopts AR augmented reality, which can directly present information on the real-world view, enhancing the intelligence of traditional Chinese medicine health monitoring, making it more intuitive and efficient, and facilitating the meeting of modern personalized and intelligent traditional Chinese medicine health monitoring management needs.

[0115] Further, referring to Figure 2As shown, an embodiment of the present invention further provides an electronic device, which applies the above-mentioned health monitoring system based on multi-domain feature extraction of pulse signals and deep learning to assist in traditional Chinese medicine health analysis and evaluation; the electronic device may include a processor 10, a memory 11, a communication bus 12, and a communication interface 13, and may also include a computer program stored in the memory 11 and operable on the processor 10.

[0116] Among them, the processor 10 may be composed of integrated circuits in some embodiments. For example, it may be composed of a single packaged integrated circuit, or may be composed of multiple integrated circuits with the same or different functions, including a combination of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control core (Control Unit) of the electronic device, connecting various components of the entire electronic device through various interfaces and lines, and executing various functions of the electronic device and processing data by running or executing programs or modules stored in the memory 11, and calling data stored in the memory 11.

[0117] Those skilled in the art should understand that the embodiments of the present invention can be provided as a software system, an electronic device, or a computer program product, etc. Therefore, the present invention can adopt the form of a complete software embodiment, a complete hardware embodiment, or an embodiment combining software and hardware aspects.

[0118] It should be noted that the word "comprising" does not exclude the existence of components or steps not listed in the claims. The word "a" or "an" before a component does not exclude the existence of multiple such components. The present invention can be implemented by means of hardware including several different components and by means of a properly programmed computer.

[0119] This specification describes in a progressive manner. For the same or similar parts between various embodiments, reference can be made to each other. The unelaborated parts of the embodiments of the present invention can be obtained from the corresponding product manuals or the prior art in this field, which are well-known in this field and will not be elaborated too much.

[0120] The above has introduced the various embodiments of the present invention in detail, and the principles and implementation methods of the present invention have been elaborated. The description of the above embodiments is only used to help understand the method and its core idea of the present invention.

[0121] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A health monitoring system based on multi-domain feature extraction of pulse signals and deep learning, characterized in that, The system includes: a multi-dimensional feature extraction module, a feature correlation mapping module, a state recognition module, and a health information feedback module, where: The multi-dimensional feature extraction module uses an FMCW radar and sensor devices to collect and obtain pulse signals, and performs multi-level processing and feature extraction on the obtained pulse signals to extract features related to the health status. The feature correlation mapping module is used to match and correlate the features of pulse signals in different frequency bands with the health status. The state recognition module automatically recognizes and classifies the features of pulse signals and the health status based on a deep learning model. The health information feedback module is used to analyze the user's health status and provide health feedback according to the health status recognition result of the state recognition module, and generate corresponding personalized health guidance information.

2. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, wherein, The system further includes: a state tracking and trend analysis module, which is used to perform time series analysis on the pulse signal data and draw a health trend graph that changes over time.

3. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, wherein The system further includes: a personalized baseline measurement module, which is used to establish an initial reference baseline for the user's health status, and give priority to comparing with the initial reference baseline when analyzing the results of each user's detection to judge abnormal health changes.

4. A health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 2, characterized in that, The system further includes: an AR display module, which uses augmented reality technology to present the information collected by the FMCW radar and sensor devices, the health feedback provided by the health information feedback module and the generated personalized health guidance information, and the health trend graph drawn by the state tracking and trend analysis module on the real-world view.

5. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, characterized in that, The operating frequency range of the FMCW radar covers 24 GHz to 77 GHz; the sensor devices include: an infrared sensor and a pressure sensor, where: the infrared sensor collects the optical waveform information of the pulse by detecting the change in light absorption caused by blood flow; the pressure sensor directly captures the pressure change generated by the pulse fluctuation through mechanical contact.

6. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, wherein The multi-dimensional feature extraction module performs multi-level processing and feature extraction on the obtained pulse signals, specifically including: time-domain analysis, frequency-domain analysis, and non-linear feature analysis, where: In time-domain analysis, the basic features of the pulse signal are extracted, including the amplitude, wave peak, wave valley of the waveform, and the pulse transmission time. In frequency-domain analysis, the pulse signal is converted from the time domain to the frequency domain, and its frequency decomposition is performed to extract the energy distribution in different frequency bands. In non-linear feature analysis, the entropy value analysis and Lyapunov exponent method are used to evaluate the dynamic complexity of the pulse signal and the stability of the visceral functions.

7. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 6, characterized in that In the frequency-domain analysis, first, the pulse signal is converted into frequency-domain information through Fourier transform to obtain the energy distribution of the signal at different frequencies; then, it is restored to the time-domain signal through inverse Fourier transform, and the Hilbert transform is used to generate an analytic signal; finally, the analytic signal and the original signal are combined to obtain a signal containing amplitude and phase information; the calculation formulas include: Where, X(f) represents the signal spectrum after the Fourier transform of the time-domain signal x(t); f represents the frequency variable; t represents the time variable, which is the time position of the signal in the time domain; j represents the imaginary unit; e represents the complex exponential function; X n (t) represents the frequency-domain information obtained after the Fourier transform of the signal x n (t), that is, the spectral representation of the signal, x n (t) represents the pulse signal obtained after the preprocessing of the signal x(t); denotes the generation of the analytic signal by the Hilbert transform. τ represents the time delay or lag and serves as the virtual integration variable in the Hilbert transform process in this formula; z(t) represents the analytic signal compounded with the signal x(t); A(t) represents the instantaneous amplitude information of the signal; φ(t) represents the instantaneous phase information of the signal.

8. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 7, characterized in that, The feature correlation mapping module includes: a frequency-domain and visceral function correlation mapping unit, an amplitude and fluctuation amplitude analysis unit, and a waveform feature database, where: The frequency domain and zang-fu function correlation mapping unit is used to correlate the pulse signal characteristics of different frequency bands with the zang-fu function states, and analyze the zang-fu states according to the fluctuations of specific frequencies; The amplitude and fluctuation amplitude analysis unit is used to analyze the envelope of the pulse signal after Hilbert transform, obtain the instantaneous amplitude and phase information, judge the strength of the zang-fu functions through the change in the fluctuation amplitude of the amplitude, and observe the flow state of qi and blood through the dynamic change of the phase; The waveform feature database is used to store the pulse diagnosis data and pulse characteristics, providing a pattern reference for the judgment of zang-fu function states; the database contains the standard pulse signal waveform data under various constitutions and health states, and has diverse qi-blood and zang-fu state patterns.

9. The health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, characterized in that, The state recognition module uses the ResNet101 deep learning model to automatically recognize and classify the pulse signal characteristics and health states through layer-by-layer convolution.

10. A health monitoring system based on multi-domain feature extraction of pulse signals and deep learning according to claim 1, characterized in that, When analyzing the user's health state, the health information feedback module also comprehensively evaluates by combining the user's pulse information, breathing frequency, and body temperature information. The evaluation method is as follows: Q = α0 + α1|F - F'| + α2|B - B'| + α3|T - T'| In the formula, Q is the health index, representing the health state value after comprehensive evaluation; α0 is the constant term, representing the basic health index without any deviation; α1 represents the weight coefficient of the pulse deviation; F represents the actual pulse value, F' represents the standard pulse value; α2 represents the weight coefficient of the breathing frequency deviation, B represents the actual breathing frequency, B' represents the standard breathing frequency value in the ideal health state; α3 represents the weight coefficient of the body temperature deviation, T represents the current body temperature value, T' represents the standard body temperature value.

Images