Personalized health maintenance dynamic scheme generation method based on spiral circulation mechanism

Abstract

The invention discloses a personalized health-preserving dynamic scheme generation method based on a spiral circulation mechanism, and the method comprises the steps: obtaining the multi-dimensional health assessment data of a user, and carrying out the classification screening to obtain a preliminary health assessment report; constructing a symptom feature matrix according to the preliminary health assessment report and the user chief complaint symptom; constructing a decision tree model containing a preset recuperation principle based on the theory of traditional Chinese medicine, and determining candidate recuperation schemes in combination with the symptom feature matrix and a symptom-recuperation mode mapping algorithm; establishing a multi-dimensional adaptation matrix according to user location information and user basic information, performing adaptation screening on the candidate nursing schemes, and determining a personalized nursing scheme of the user; and regularly monitoring the health data of the user after the personalized adaptive nursing scheme is implemented, and reconstructing and updating the nursing scheme. According to the method, a traditional Chinese medicine theory and modern data analysis are deeply fused, and comprehensive evaluation and accurate nursing are combined, so that an efficient, accurate and high-adaptability personalized health-preserving nursing scheme can be generated.

Description

Technical Field

[0001] The present invention relates to the field of health management technology, and in particular to a method for generating a personalized health maintenance dynamic plan based on a spiral circulation mechanism. Background Art

[0002] With the increasing demand for health management, modern health management requires a comprehensive and dynamic profile of individual health. This requires the integration of multimodal data to tailor personalized treatment plans for patients. However, in current Traditional Chinese Medicine health testing and treatment practices, existing testing systems typically recommend solutions to improve specific health issues based solely on the user's current health indicators. Due to the incomplete reference health data, this approach fails to comprehensively consider the user's long-term health status and, therefore, is unable to develop a sustainable and effective treatment plan. Therefore, it is necessary to monitor the user's long-term health indicators and continuously iterate and adjust the treatment plan to better provide users with personalized health management.

[0003] Existing automated TCM treatment plan generation systems typically rely on pre-storing various treatment methods in a computer and processing them based on information such as the patient's "primary symptoms, brief examination, important examinations, physical classification, current goals, specific countermeasures, prescription name, and prescription lineup." This approach relies on the patient's subjective symptoms and judgment, rather than a comprehensive analysis of objective health data and long-term test records. Due to the lack of comprehensive integration of a patient's historical health data, this approach cannot provide the optimal treatment plan based on individual differences, and therefore cannot achieve truly personalized health management.

[0004] Therefore, the present invention proposes a method for generating a personalized health care and conditioning plan based on spiral circulation, which utilizes hierarchical screening of multidimensional health assessment data, a decision tree model driven by traditional Chinese medicine theory, and a symptom feature matrix and conditioning method mapping algorithm to generate a more accurate personalized conditioning plan. Summary of the Invention

[0005] In view of this, the present invention provides a personalized health care dynamic plan generation method based on a spiral cycle mechanism to solve the technical problem that the existing automatic health care plan generation method lacks comprehensive integration of user health data and cannot realize personalized health management of users.

[0006] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism, comprising:

[0008] Obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report;

[0009] Construct a symptom feature matrix based on the preliminary health assessment report and user-reported symptoms;

[0010] Based on Traditional Chinese Medicine theory, a decision tree model with preset conditioning principles was constructed, and candidate conditioning plans were determined by combining the symptom feature matrix and the symptom-conditioning mapping algorithm.

[0011] Establish a multi-dimensional adaptation matrix based on the user's location information and basic user information, perform adaptation screening on the candidate conditioning plans, and determine the user's personalized conditioning plan;

[0012] The user's health data after the implementation of the personalized adaptive conditioning plan is regularly monitored to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, the conditioning plan is reconstructed and updated.

[0013] Furthermore, the multi-dimensional health assessment data includes: a user's pulse diagnosis, tongue diagnosis, constitution identification and lung function sub-item assessment report within a preset time period;

[0014] The health assessment data is screened in a hierarchical manner to obtain a preliminary health assessment report, including:

[0015] The scores of each assessment report are divided into health levels based on the preset assessment rules;

[0016] Eliminate reports with health levels higher than the health threshold, sort the remaining reports according to the health range, and then sort them from small to large based on the scores to select up to two health assessment reports as preliminary health assessment reports.

[0017] Furthermore, the symptom feature matrix is ​​constructed based on the preliminary health assessment report and the user's main symptoms, including:

[0018] Analyze initial symptoms based on the preliminary health assessment report;

[0019] Generate a comprehensive symptom entry library based on the initial judgment symptoms and the user's main complaint symptoms;

[0020] All symptoms in the comprehensive symptom entry library are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed.

[0021] Furthermore, all symptoms in the comprehensive symptom database are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed, including:

[0022] Conduct an independence test on each symptom and health score, and retain symptoms with a probability greater than the first preset probability range;

[0023] Analyze the mutual relationship between symptoms, and save mutually exclusive symptom pairs whose conditional probability is less than a second preset probability as a symptom co-occurrence matrix;

[0024] The symptom data were reduced in dimension using principal component analysis to obtain the symptom feature matrix.

[0025] Furthermore, the symptom data is reduced in dimension by principal component analysis to obtain a symptom feature matrix, and further includes:

[0026] Based on the dynamic allocation formula, the weight coefficient of each symptom is dynamically allocated according to the symptom duration and attack frequency, and the weight coefficient is applied to the symptom feature matrix;

[0027] The dynamic allocation formula of the weight coefficient is:

[0028] w i =(1+ln(1+T c ))·e -0.1t

[0029] Among them, w i represents the dynamic weight of the ith symptom, T c Indicates the duration of symptoms, e -0.1t Represents the time decay factor.

[0030] Furthermore, a decision tree model including preset conditioning principles is constructed based on TCM theory. The candidate conditioning plans are determined by combining the symptom feature matrix and the symptom-conditioning method mapping algorithm, including:

[0031] Construct a decision tree model with the initial symptom feature matrix as the root node, TCM syndrome classification as the internal node, and preset conditioning principles as the leaf nodes;

[0032] The dual criteria of Gini index and information gain were used to select split nodes, and a two-layer graph mapping was constructed; the first layer was the matching path between symptoms and syndrome types, and the second layer was the matching path between syndrome types and conditioning principles.

[0033] Furthermore, a multi-dimensional adaptation matrix is ​​established based on the user's location information and basic user information, and the candidate conditioning plans are adapted and screened to determine the user's personalized conditioning plan, including:

[0034] The candidate maintenance plans are vectorized to obtain a plan feature vector, and the user location information and basic information are vectorized to obtain a user feature vector;

[0035] Based on TCM theory and practical experience, the influence of each dimension on the adaptability of the conditioning plan is comprehensively considered, and the dimensional weights of user location information and user basic information are set separately;

[0036] Calculate the cosine similarity between the user's feature vector and the feature vectors of each solution, sort the candidate maintenance solutions according to the cosine similarity, and select the personalized maintenance solution that matches the user.

[0037] Furthermore, the user's health data after the implementation of the personalized adaptation and conditioning plan is regularly monitored to obtain health monitoring data, including:

[0038] The sliding window average method is used to calculate the deviation between the user's health data and the reference standard data;

[0039] A deviation early warning mechanism is established based on the size of the deviation and the changing trend. When the monitoring deviation exceeds the corresponding threshold or a new user complaint is received, the user's maintenance plan is updated and sent to the user in real time.

[0040] In a second aspect, the present invention further provides a personalized health dynamic plan generation system based on a spiral cycle mechanism, comprising:

[0041] The health data collection module is used to obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report;

[0042] A preliminary analysis module is used to construct a symptom feature matrix based on the preliminary health assessment report and the user's main symptoms;

[0043] The plan formulation module is used to construct a decision tree model containing preset conditioning principles based on Traditional Chinese Medicine theory, and determine candidate conditioning plans by combining the symptom feature matrix and the symptom-conditioning method mapping algorithm;

[0044] An adaptation optimization module is used to establish a multi-dimensional adaptation matrix based on the user's location information and basic user information, perform adaptation screening on the candidate conditioning plans, and determine the user's personalized conditioning plan;

[0045] The dynamic monitoring module is used to regularly monitor the user's health data after the implementation of the personalized adaptive conditioning plan to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, the conditioning plan is reconstructed and updated.

[0046] In a third aspect, the present invention also provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method for generating a personalized health dynamic plan based on a spiral cycle mechanism as described in the above technical solution is implemented.

[0047] Compared with the existing technology, the personalized health dynamic plan generation method and system based on the spiral cycle mechanism proposed in the present invention have the following advantages:

[0048] The present invention uses hierarchical screening of multidimensional health assessment data, combines the user's main symptoms to construct a symptom feature matrix, and constructs a decision tree model based on traditional Chinese medicine theory. It integrates the overall concept and principle of traditional Chinese medicine for treatment based on syndrome differentiation with modern data analysis, which not only improves the scientific nature and reliability of the conditioning plan, but also better adapts to the needs of modern medical care and health management. It establishes a multidimensional adaptation matrix based on user basic information and location information, and can accurately generate personalized conditioning plans for each user. By regularly monitoring health data and dynamically adjusting the conditioning plan according to the deviation threshold, it ensures that the conditioning plan always matches the user's health status. The present invention infers conditioning principles from detected symptoms, and then derives conditioning methods from the conditioning principles combined with objective factors, and finally generates a comprehensive conditioning plan process mechanism. According to user detection and main complaints, it uses an iterative cycle to always maintain the timeliness and effectiveness of the user's conditioning plan, providing users with health conditioning plans that are more in line with their own needs and objective conditions, which helps users achieve better conditioning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic flow chart of the method for generating a personalized health regimen dynamic plan based on a spiral circulation mechanism provided by the present invention;

[0050] Figure 2 A flow chart of a method for generating a conditioning program provided by the present invention in practical application;

[0051] Figure 3 This is a schematic diagram of the structure of the personalized health dynamic plan generation system based on the spiral circulation mechanism provided by the present invention;

[0052] Figure 4 A schematic diagram of the mechanism for generating a conditioning plan provided by the present invention;

[0053] Figure 5 This is a structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0054] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0055] See Figure 1 This embodiment provides a method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism, including:

[0056] Step S101: Obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report;

[0057] Step S102: constructing a symptom feature matrix based on the preliminary health assessment report and the user's main symptoms;

[0058] Step S103: Constructing a decision tree model including preset nursing principles based on TCM theory, and determining candidate nursing plans by combining the symptom feature matrix and the symptom-nursing method mapping algorithm;

[0059] Step S104: establishing a multi-dimensional adaptation matrix based on the user's location information and basic user information, performing adaptation screening on the candidate conditioning plans, and determining a personalized conditioning plan for the user;

[0060] Step S105: Regularly monitor the user's health data after implementing the personalized adaptive conditioning plan to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, reconstruct and update the conditioning plan.

[0061] The method of this embodiment uses hierarchical screening of multi-dimensional health assessment data, combines the user's main symptoms to construct a symptom feature matrix, and builds a decision tree model based on traditional Chinese medicine theory. It integrates the overall concept and syndrome differentiation and treatment principle of traditional Chinese medicine with modern data analysis, which not only improves the scientific nature and reliability of the conditioning plan, but also better adapts to the needs of modern medical care and health management. It establishes a multi-dimensional adaptation matrix based on the user's basic information and location information, which can accurately generate a personalized conditioning plan for each user. By regularly monitoring health data and dynamically adjusting the conditioning plan according to the deviation threshold, it ensures that the conditioning plan always matches the user's health status. According to the user's detection and main complaint, the user's conditioning plan is always kept timely and effective in an iterative cycle.

[0062] As a preferred embodiment, in step S101, the multi-dimensional health assessment data includes: a user's pulse diagnosis, tongue diagnosis, constitution identification and lung function sub-item assessment report within a preset time period;

[0063] The health assessment data is screened in a hierarchical manner to obtain a preliminary health assessment report, including:

[0064] The scores of each assessment report are divided into health levels based on the preset assessment rules;

[0065] Eliminate reports with health levels higher than the health threshold, sort the remaining reports according to the health range, and then sort them from small to large based on the scores to select up to two health assessment reports as preliminary health assessment reports.

[0066] As a specific embodiment, in actual operation, the user's pulse diagnosis, tongue diagnosis, constitution identification, lung function assessment and other test records for the past month are first scanned, and each report is divided into three health levels: red light, yellow light and green light according to the score range of each item (the red light interval is the first score range with a lower score, which means that the user is in poor condition in this item; the second is the yellow light score range, and the best state is the green light score range, which means that the user's health level is higher than the health threshold). After excluding the green light reports, all reports are first sorted by red light and yellow light, and then sorted by the test score from small to large, and at most two health assessment reports are screened out.

[0067] As a preferred embodiment, in step S102, the symptom feature matrix is ​​constructed based on the preliminary health assessment report and the user's main symptoms, including:

[0068] Analyze initial symptoms based on the preliminary health assessment report;

[0069] Generate a comprehensive symptom entry library based on the initial judgment symptoms and the user's main complaint symptoms;

[0070] All symptoms in the comprehensive symptom entry library are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed.

[0071] As a specific embodiment, the initial judgment symptoms are analyzed according to the preliminary health assessment report, specifically: first, the pulse signals in the pulse assessment report are analyzed by wavelet packet decomposition method, and the energy proportions of the δ (0.5-4Hz), θ (4-8Hz), and α (8-12Hz) frequency bands are extracted, and the energy proportions of the δ (0.5-4Hz), θ (4-8Hz), and α (8-12Hz) frequency bands are extracted according to the formula Determine the TCM signs of weak pulse, string pulse, etc., among which E band Indicates the vibration intensity of the pulse in different frequency bands to quantify the "pulse strength" characteristics; WPT band (k) represents the decomposition of the radial artery pulse wave in a specific frequency domain, and WPT(k) represents the sum of the wavelet packet coefficients across the entire frequency band, representing the overall energy level of the pulse signal. The δ, θ, and α bands are used to characterize ultra-low-frequency vasomotor fluctuations, cardiac pumping rhythmic fluctuations, and arterial elastic vibrations, respectively. It should be noted that a low energy percentage in the δ band may be associated with a "deep and thready pulse" (e.g., Qi and Blood Deficiency Syndrome); abnormal energy in the θ band indicates a "stuttering pulse," and decreased energy in the α band may indicate a "stringy and firm pulse."

[0072] Tongue image analysis involves converting tongue images into the HSV color space, analyzing tongue texture and coating, segmenting the tongue contour using a U-Net model, and quantifying tongue color features using the CIELab color space. Tongue image feature vectors are then associated with tongue texture, coating, and color.

[0073] Finally, a random forest model was constructed, and the pulse frequency domain parameters, tongue HSV parameters and constitution questionnaire scores were input into the model for analysis, and the initial judgment symptoms were finally output, such as the constitution probability distribution: phlegm-damp constitution, associated signs of obesity and phlegm, etc.

[0074] As a preferred embodiment, all symptoms in the comprehensive symptom database are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed, including:

[0075] Conduct an independence test on each symptom and health score, and retain symptoms with a probability greater than the first preset probability range;

[0076] Analyze the mutual relationship between symptoms, and save mutually exclusive symptom pairs whose conditional probability is less than a second preset probability as a symptom co-occurrence matrix;

[0077] The symptom data were reduced in dimension using principal component analysis to obtain the symptom feature matrix.

[0078] As a specific embodiment, an independence test is performed on each symptom and the health score. Specifically, the independence of each symptom and the health score is calculated by a chi-square test, which is expressed as follows:

[0079]

[0080] Among them, χ 2 represents the chi-square statistic, which is used to quantify the degree of deviation between the observed data and the expected value under the independence assumption; i represents the observation frequency, E i Represents the expected frequency. Symptoms with a probability lower than the first preset probability are discarded.

[0081] As a specific embodiment, to analyze the relationship between symptoms, it is necessary to calculate the probability of symptom co-occurrence, and the calculation formula is:

[0082]

[0083] Among them, P(s j ∣s i ) indicates that the symptoms i When symptoms occur, j Conditional probability of simultaneous occurrence, count(s i ∩s j ) indicates symptoms i and s j Frequency of co-occurrence; count(s i ) indicates symptoms i The total frequency of individual occurrences.

[0084] When judging whether they are mutually exclusive symptoms, a two-way conditional probability threshold is used, that is, assuming that the second preset probability is set to 0.1, P(s j ∣s i )<0.1 and P(s j ∣s i )<0.1.

[0085] The medical significance of this is that if two symptoms rarely occur together, they are considered "mutually exclusive symptoms" in traditional Chinese medicine theory. For example, the symptoms of "fear of cold" and "hot flashes" are mutually exclusive symptoms.

[0086] As a specific embodiment, the symptom data is reduced in dimension through principal component analysis. The high-dimensional symptom features can be projected into a two-dimensional space using the t-SNE dimensionality reduction algorithm, and the principal components with a cumulative variance contribution rate greater than 85% are retained to obtain a symptom feature matrix.

[0087] In order to include the duration and frequency of symptoms in the quantitative range and better personalize the user, as a preferred embodiment, the symptom data is reduced in dimension by principal component analysis to obtain a symptom feature matrix, which also includes:

[0088] Based on the dynamic allocation formula, the weight coefficient of each symptom is dynamically allocated according to the symptom duration and attack frequency, and the weight coefficient is applied to the symptom feature matrix;

[0089] The dynamic allocation formula of the weight coefficient is:

[0090] w i =(1+ln(1+T c ))·e -0.1t

[0091] Among them, w i represents the dynamic weight of the ith symptom, T c Indicates the duration of symptoms, e -0.1t Represents the time decay factor.

[0092] Due to individual differences, the manifestation, duration, and frequency of symptoms may vary. Dynamically assigning weighting coefficients allows for a tailored health assessment based on each individual's symptom characteristics. This personalized approach better reflects each individual's health status, providing more targeted solutions for health management or intervention. Frequent and prolonged symptoms generally have a greater impact on health and can be assigned a higher weight. Some symptoms, even if less frequent, may have a serious impact on health if they persist for a long time, and their corresponding weights can also be increased.

[0093] As a preferred embodiment, in step S103, a decision tree model including preset conditioning principles is constructed based on traditional Chinese medicine theory, and candidate conditioning plans are determined by combining the symptom feature matrix and the symptom-conditioning method mapping algorithm, including:

[0094] Construct a decision tree model with the initial symptom feature matrix as the root node, TCM syndrome classification as the internal node, and preset conditioning principles as the leaf nodes;

[0095] The dual criteria of Gini index and information gain were used to select split nodes, and a two-layer graph mapping was constructed; the first layer was the matching path between symptoms and syndrome types, and the second layer was the matching path between syndrome types and conditioning principles.

[0096] As a specific embodiment, a comprehensive TCM syndrome classification set is used to classify all symptoms into 23 conditioning principles according to the corresponding conditioning methods according to TCM theory (respectively: regulating and replenishing Yang Qi, nourishing Yin and promoting fluid production, harmonizing Qi and blood, soothing the liver and regulating Qi, strengthening the spleen and nourishing the stomach, removing dampness and resolving phlegm, nourishing the kidneys and strengthening bones, harmonizing the internal organs, replenishing Qi and Yin, eliminating accumulation and guiding stagnation, removing blood stasis and dredging the meridians, removing wind and dispersing cold, strengthening the foundation and nourishing the essence, moistening the intestines and relieving constipation, clearing heat and purging fire, moistening the lungs and relieving cough, promoting Qi and blood circulation, nourishing the kidneys and replenishing essence, removing blood stasis and resolving stagnation, warming the meridians and dredging the shoulders and neck, regulating the thoracic meridians, and dredging the waist and back). According to the user's health score and main complaint, the conditioning principles suitable for the corresponding symptoms are screened out, and then according to the mutual exclusion and order relationship between the principles, the principle list is deduplicated, sorted and unreasonable items are excluded to obtain the final applicable conditioning principles.

[0097] Taking the common cold as an example, the root node contains various symptom features that a cold patient may experience, such as fever, cough, nasal congestion, runny nose, sore throat, fatigue, and headache. Common cold syndromes in Traditional Chinese Medicine include wind-cold cold, wind-heat cold, and summer-damp cold. During decision tree construction, symptom features that effectively distinguish these syndromes are used as split nodes. For example, to distinguish between wind-cold and wind-heat colds, symptoms such as the severity of fever, the color of sputum (white or yellow), and whether the throat is red, swollen, or painful may be key splitting features. For wind-cold colds, the default treatment principles may include relieving the exterior with pungent and warm herbs and dispersing wind-cold. Common treatment methods include taking Jingfang Baidu Powder or scallion and ginger water. For wind-heat colds, the default treatment principles are relieving the exterior with pungent and cool herbs and clearing heat and detoxifying. Corresponding treatment methods include Yinqiao Powder and chrysanthemum tea. These default treatment principles are leaf nodes. Next, we use the dual criteria of the Gini index and information gain to select split nodes and construct a two-layer graph mapping.

[0098] Specifically, if the symptom of "fever" is used as the splitting node, the Gini index and information gain are calculated to evaluate the effectiveness of this node. First, the data is split according to the presence or absence of fever (or grouped by the degree of fever), which can distinguish between wind-cold colds and wind-heat colds. Once the TCM syndrome type is determined (such as the patient is classified as a wind-cold cold syndrome), it can be mapped to a specific conditioning plan based on the pre-set conditioning principles. At the leaf node of the wind-cold cold syndrome, the corresponding conditioning plan is a specific conditioning method guided by the previously preset principles of pungent and warm diaphoresis, such as recommending taking Jingfang Baidu Powder, keeping warm, and getting adequate rest.

[0099] As a preferred embodiment, in step S104, a multi-dimensional adaptation matrix is ​​established based on the user location information and the user basic information, and the candidate conditioning plans are adapted and screened to determine the user's personalized conditioning plan, including:

[0100] The candidate maintenance plans are vectorized to obtain a plan feature vector, and the user location information and basic information are vectorized to obtain a user feature vector;

[0101] Based on TCM theory and practical experience, the influence of each dimension on the adaptability of the conditioning plan is comprehensively considered, and the dimensional weights of user location information and user basic information are set separately;

[0102] Calculate the cosine similarity between the user's feature vector and the feature vectors of each solution, sort the candidate maintenance solutions according to the cosine similarity, and select the personalized maintenance solution that matches the user.

[0103] Specifically, each conditioning principle has multiple different conditioning plans and methods associated with different conditions such as gender, age, and solar term. Based on the user's current gender, age, and solar term conditions, as well as the conditioning principle obtained in the previous step, a list of all conditioning methods that meet the conditioning principle and the user's health status is screened, thereby generating a final conditioning plan report.

[0104] As a preferred embodiment, in step S105, the user's health data after implementing the personalized adaptive conditioning plan is regularly monitored to obtain health monitoring data, including:

[0105] Regularly monitor the user's health data after implementing the personalized adaptation and conditioning plan to obtain health monitoring data, including:

[0106] The sliding window average method is used to calculate the deviation between the user's health data and the reference standard data;

[0107] A deviation early warning mechanism is established based on the size of the deviation and the changing trend. When the monitoring deviation exceeds the corresponding threshold or a new user complaint is received, the user's maintenance plan is updated and sent to the user in real time.

[0108] As a specific embodiment, in specific applications, a fixed time window can be set (for example, 7 days as a window), and the health data within the window can be averaged to obtain the average health data. At the same time, the reference standard data (standard data set according to the user's ideal health status or disease treatment goals) is also averaged within the same time window. According to the size of the deviation and the trend of change, the user's health status is analyzed and evaluated. A deviation warning mechanism is established, and different deviation thresholds are set. When the deviation exceeds the corresponding threshold, warning signals of different degrees are issued. For example, for blood glucose monitoring indicators, when the deviation of the blood glucose value from the reference standard exceeds 20%, a yellow warning is issued; when the deviation exceeds 30%, a red warning is issued.

[0109] In some embodiments, to avoid the impact of a single window length on the results, a sliding window averaging method with multiple window lengths can be used for calculation, and the deviations under different windows can be comprehensively considered to more accurately reflect the short-term and long-term trends of the user's health data. The user's health care plan is updated in an iterative manner. When new test results or user complaints are detected, the above process is re-executed to provide the user with the latest health care principles and methods, and the user is notified in real time of the updated plan, always providing the user with the most timely and effective health care plan.

[0110] See Figure 2 , Figure 2 A flow chart showing the method of generating a conditioning plan in practical application is presented.

[0111] like Figure 3 As shown, the embodiment of the present invention further provides a personalized health dynamic plan generation system 300 based on a spiral cycle mechanism, comprising:

[0112] The health data collection module 301 is used to obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report;

[0113] A preliminary analysis module 302 is used to construct a symptom feature matrix based on the preliminary health assessment report and the user's main symptoms;

[0114] The plan formulation module 303 is used to construct a decision tree model containing preset nursing principles based on traditional Chinese medicine theory, and determine candidate nursing plans by combining the symptom feature matrix and the symptom-nursing method mapping algorithm;

[0115] Adaptation optimization module 304, used to establish a multi-dimensional adaptation matrix based on the user's location information and basic user information, perform adaptation screening on the candidate conditioning plans, and determine a personalized conditioning plan for the user;

[0116] The dynamic monitoring module 305 is used to regularly monitor the user's health data after the personalized adaptive conditioning plan is implemented to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, the conditioning plan is reconstructed and updated.

[0117] As a specific embodiment, in actual application, this system is based on Figure 4 The shown conditioning plan generation mechanism operates, taking the multi-dimensional health assessment results as symptom metadata, managing them through the symptom library, managing candidate prescriptions (including basic information, efficacy, etc.) through the prescription library, generating a conditioning principle management database for TCM conditioning principles, deeply integrating TCM theory with modern data analysis methods, and ultimately generating efficient, accurate and adaptable personalized health conditioning plans.

[0118] like Figure 5 The present invention also provides an electronic device 500 for generating a personalized dynamic health regimen based on a spiral cycle mechanism. The electronic device can be a computing device such as a mobile terminal, desktop computer, notebook, PDA, or server. The electronic device includes a processor 501, a memory 502, and a display 503.

[0119] In some embodiments, the memory 502 may be an internal storage unit of the computer device, such as a hard drive or memory of the computer device. In other embodiments, the memory 502 may also be an external storage device of the computer device, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the memory 502 may include both an internal storage unit of the computer device and an external storage device. The memory 502 is used to store application software installed on the computer device and various types of data, such as program code installed on the computer device. The memory 502 may also be used to temporarily store data that has been output or is about to be output. In one embodiment, the memory 502 stores a program 504 for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism. The program 504 for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism can be executed by the processor 501, thereby implementing a personalized health regimen dynamic plan generation method based on a spiral cycle mechanism according to various embodiments of the present invention.

[0120] In some embodiments, the processor 501 can be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 502, such as executing a personalized health dynamic program generation method program based on a spiral cycle mechanism.

[0121] In some embodiments, the display 503 can be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display 503 is used to display information on the computer device and to display a visual user interface. The components 501-503 of the computer device communicate with each other via a system bus.

[0122] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism, characterized in that: include: Obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report; Construct a symptom feature matrix based on the preliminary health assessment report and user-reported symptoms; Based on Traditional Chinese Medicine theory, a decision tree model with preset conditioning principles was constructed, and candidate conditioning plans were determined by combining the symptom feature matrix and the symptom-conditioning mapping algorithm. Establish a multi-dimensional adaptation matrix based on the user's location information and basic user information, perform adaptation screening on the candidate conditioning plans, and determine the user's personalized conditioning plan; The user's health data after the implementation of the personalized adaptive conditioning plan is regularly monitored to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, the conditioning plan is reconstructed and updated.

2. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 1, characterized in that: The multi-dimensional health assessment data includes: pulse diagnosis, tongue diagnosis, constitution identification and lung function sub-item assessment reports within a preset time period; The health assessment data is screened in a hierarchical manner to obtain a preliminary health assessment report, including: The scores of each assessment report are divided into health levels based on the preset assessment rules; Eliminate reports with health levels higher than the health threshold, sort the remaining reports according to the health range, and then sort them from small to large based on the scores to select up to two health assessment reports as preliminary health assessment reports.

3. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 1, characterized in that: The symptom feature matrix is ​​constructed based on the preliminary health assessment report and the user's main symptoms, including: Analyze initial symptoms based on the preliminary health assessment report; Generate a comprehensive symptom entry library based on the initial judgment symptoms and the user's main complaint symptoms; All symptoms in the comprehensive symptom entry library are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed.

4. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 3, characterized in that: All symptoms in the comprehensive symptom database are denoised, mutually exclusive symptoms are eliminated, and a symptom feature matrix is ​​constructed, including: Conduct an independence test on each symptom and health score, and retain symptoms with a probability greater than the first preset probability range; Analyze the mutual relationship between symptoms, and save mutually exclusive symptom pairs whose conditional probability is less than a second preset probability as a symptom co-occurrence matrix; The symptom data were reduced in dimension using principal component analysis to obtain the symptom feature matrix.

5. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 4, characterized in that: The symptom data is reduced in dimension by principal component analysis to obtain a symptom feature matrix, and further includes: Based on the dynamic allocation formula, the weight coefficient of each symptom is dynamically allocated according to the symptom duration and attack frequency, and the weight coefficient is applied to the symptom feature matrix; The dynamic allocation formula of the weight coefficient is: w i =(1+ln(1+T c ))·have been -0.1t Among them, w i represents the dynamic weight of the ith symptom, T c Indicates the duration of symptoms, e -0.1t Represents the time decay factor.

6. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 1, characterized in that: Based on Traditional Chinese Medicine theory, a decision tree model with preset conditioning principles was constructed. The candidate conditioning plans were determined by combining the symptom feature matrix and the symptom-conditioning mapping algorithm, including: Construct a decision tree model with the initial symptom feature matrix as the root node, TCM syndrome classification as the internal node, and preset conditioning principles as the leaf nodes; The dual criteria of Gini index and information gain were used to select split nodes, and a two-layer graph mapping was constructed; the first layer was the matching path between symptoms and syndrome types, and the second layer was the matching path between syndrome types and conditioning principles.

7. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 1, characterized in that: A multi-dimensional adaptation matrix is ​​established based on the user's location information and basic user information, and the candidate conditioning plans are adapted and screened to determine the user's personalized conditioning plan, including: The candidate maintenance plans are vectorized to obtain a plan feature vector, and the user location information and basic information are vectorized to obtain a user feature vector; Based on TCM theory and practical experience, the influence of each dimension on the adaptability of the conditioning plan is comprehensively considered, and the dimensional weights of user location information and user basic information are set separately; Calculate the cosine similarity between the user's feature vector and the feature vectors of each solution, sort the candidate maintenance solutions according to the cosine similarity, and select the personalized maintenance solution that matches the user.

8. The method for generating a personalized health regimen dynamic plan based on a spiral cycle mechanism according to claim 1, characterized in that: Regularly monitor the user's health data after implementing the personalized adaptation and conditioning plan to obtain health monitoring data, including: The sliding window average method is used to calculate the deviation between the user's health data and the reference standard data; A deviation early warning mechanism is established based on the size of the deviation and the changing trend. When the monitoring deviation exceeds the corresponding threshold or a new user complaint is received, the user's maintenance plan is updated and sent to the user in real time.

9. A personalized health regimen dynamic plan generation system based on a spiral cycle mechanism, characterized by: include: The health data collection module is used to obtain the user's multi-dimensional health assessment data, perform hierarchical screening on the health assessment data, and obtain a preliminary health assessment report; A preliminary analysis module is used to construct a symptom feature matrix based on the preliminary health assessment report and the user's main symptoms; The plan formulation module is used to construct a decision tree model containing preset conditioning principles based on Traditional Chinese Medicine theory, and determine candidate conditioning plans by combining the symptom feature matrix and the symptom-conditioning method mapping algorithm; An adaptation optimization module is used to establish a multi-dimensional adaptation matrix based on the user's location information and basic user information, perform adaptation screening on the candidate conditioning plans, and determine the user's personalized conditioning plan; The dynamic monitoring module is used to regularly monitor the user's health data after the implementation of the personalized adaptive conditioning plan to obtain health monitoring data. When the health monitoring data reaches a preset deviation threshold, the conditioning plan is reconstructed and updated.

10. An electronic device, characterized in that: The invention comprises a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the method for generating a personalized health dynamic plan based on a spiral cycle mechanism as described in any one of claims 1 to 8 is implemented.

Images