Female internal secrecy detection report interpretation system based on artificial intelligence

Abstract

The invention discloses a female internal secret detection report interpretation system based on artificial intelligence, and relates to the technical field of artificial intelligence health detection, and the system comprises a detection index module, report generation, health management generation, and disease risk prediction. According to the invention, the system provides a personalized ovarian health degree report and expert suggestions according to the age, AMH value, fertility requirement and menstruation rule condition of the user, so that the user can better know the ovarian health condition of the user; by detecting the AMH value, the system can screen the ovarian disease risk in an early stage, timely treatment suggestions are provided for a user, and disease deterioration is prevented; an artificial intelligence algorithm is adopted by a report generation engine, an overview, interpretation, disease types and expert suggestions of a user report can be intelligently derived according to information input by a user and a detection result, and the accuracy and efficiency of the report are greatly improved; the system also provides a personalized health management scheme to help the user to improve living habits and reduce the risk of ovarian diseases.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence health detection technology, and specifically to a female endocrine detection report interpretation system based on artificial intelligence. Background Art

[0002] In the field of women's health, ovarian health testing has always been an important research direction. As the core organ of the female reproductive system, the health of the ovaries directly affects women's fertility and overall health. However, traditional ovarian health testing methods have many limitations.

[0003] First, the traditional ovarian health test mainly relies on the doctor's clinical experience and subjective judgment. This method is often limited by the doctor's personal knowledge and experience, which makes it difficult to ensure the accuracy and consistency of the test results. At the same time, the test process is cumbersome, requiring a lot of time and human resources, making the test efficiency low. Secondly, the traditional ovarian health test usually only focuses on a single physiological indicator, such as AMH (anti-Mullerian hormone) level, while ignoring other factors closely related to ovarian health, such as age, fertility requirements, menstrual conditions, etc. The impact of these factors on ovarian health cannot be ignored, and traditional methods often fail to fully consider these factors, resulting in incomplete and inaccurate test results. In addition, there are certain problems in the interpretation of traditional ovarian health test reports. On the one hand, since the interpretation of the report depends on the doctor's personal experience and knowledge, the interpretation results are highly subjective, and there may be differences in the interpretation results between different doctors. On the other hand, the traditional report interpretation method often lacks consideration of the personalized needs of users and cannot provide users with personalized health advice and health management solutions.

[0004] In view of this, a female endocrine test report interpretation system based on artificial intelligence is provided to overcome the above problems, aiming to achieve a comprehensive, accurate and personalized interpretation of the ovarian health test report through artificial intelligence technology. By integrating the user's basic information and menstrual conditions, combined with the detection index AMH value, the system can intelligently derive the overview, interpretation, disease type and expert advice of the user's report. This method can not only improve the accuracy and efficiency of the test results, but also provide users with personalized health advice and management plans, help users better understand their ovarian health status, reduce the risk of ovarian diseases, and improve fertility. Therefore, the proposal of the present invention has important practical significance and application value. Summary of the invention

[0005] The purpose of the present invention is to provide a female endocrine test report interpretation system based on artificial intelligence to solve the problems raised in the above-mentioned background technology.

[0006] In order to solve the above technical problems, the present invention provides an artificial intelligence-based female endocrine test report interpretation system, which is used to check the AMH index, grasp the ovarian reserve function status, screen the risk of ovarian diseases early, and detect the ovarian age, including:

[0007] Detection indicator module: detect AMH, namely anti-Mullerian hormone;

[0008] User input module: Users input basic information and menstrual status through the mini program:

[0009] User report module: The user inputs basic information and menstrual status. When the AMH value is detected after blood sampling, a user report is generated through the report generation engine. The user report mainly includes: test results, overview, result interpretation, expert advice and health management;

[0010] Report overview, interpretation, and expert advice knowledge base module: Based on age and test indicators, report overview and report interpretation are derived; based on age, test indicators, fertility requirements, and menstrual patterns, expert advice is derived;

[0011] Report content derivation module: including report overview, interpretation, and expert advice paradigm.

[0012] Furthermore, the system can also be used to interpret the test report of ovarian fertility reserve capacity; that is, in the test index module, AMH and six sex hormones FSH, LH, E2, P, PRL, T are tested to comprehensively evaluate ovarian fertility; the user input module remains unchanged; in the user report module, after sampling, the AMH value and the value SH of the six sex hormones are detected, and a user report is generated through the report generation engine; the report generation engine generates a user report according to R=

[0013] The fn(a,amh,sh,f,m) paradigm derives the overview, interpretation, and disease type of the user report based on parameters a, amh, and sh.

[0014] Furthermore, the system can also evaluate endocrine abnormalities, irregular menstruation, and provide health guidance for pregnancy preparation by checking the six sex hormone indicators of FSH, LH, E2, P, PRL, and T.

[0015] Furthermore, it also includes a health management module; that is, the health management plan included in the six assessment reports of ovarian health, fertility reserve capacity, and sex hormones, the disease types derived according to the medical content deduction paradigms R=fn(a,amh,f,m), R=fn(a,amh,f,m) and R=fn(a,amh,sh,f,m) of the three products, combined with the user's age and BMI value, generate the user's health management content; the health management module includes a health management knowledge base.

[0016] Furthermore, the health management module also includes health management paradigms;

[0017] Normal form: H = fn(a,c,b);

[0018] H: Health management;

[0019] a: parameter age;

[0020] c: parameter disease category;

[0021] b: parameter bmi.

[0022] Furthermore, the health management module also includes a derivation process: based on the disease type (c), the user's basic information: age (a), bmi (b), the user's health management plan is derived according to the H=fn(a,c,b) paradigm; based on the medical content overview, interpretation, expert advice and health management plan, a complete report for the user is finally generated.

[0023] Furthermore, in the health management module, push notification and health reminder functions are introduced; the user information collected through the user's basic information, test indicators, and questionnaires are formatted and provided to the general large model as context to generate personalized health management guidance, including: diet plan, nutrient plan, health check-up, and lifestyle recommendations.

[0024] Furthermore, a user feedback mechanism is established in the system to collect users' opinions and suggestions on the system.

[0025] Furthermore, a binary classification prediction model is trained to quickly and accurately determine abnormal values ​​of indicators.

[0026] Furthermore, a multi-classification decision tree prediction model is used to predict possible diseases based on historical test results and user information.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The system can provide personalized ovarian health reports and expert advice based on the user's age, AMH value, fertility requirements and menstrual regularity, allowing users to better understand their ovarian health.

[0029] By detecting the AMH value, the system can screen for the risk of ovarian disease at an early stage, provide users with timely treatment recommendations, and prevent the disease from worsening.

[0030] The report generation engine uses artificial intelligence algorithms, which can intelligently derive the overview, interpretation, disease type and expert advice of the user report based on the information entered by the user and the test results, greatly improving the accuracy and efficiency of the report.

[0031] The system also provides personalized health management plans to help users improve their lifestyle habits, reduce the risk of ovarian diseases, and improve fertility.

[0032] This system is not only suitable for the interpretation of ovarian health test reports, but can also be extended to the interpretation of other test items related to ovarian health, including ovarian reproductive reserve capacity testing and six sex hormone assessments; by continuously optimizing algorithms and adding functional modules, it can meet the personalized needs of more users.

[0033] The binary classification prediction model learns user information and detection indicator data, and can perform artificial intelligence screening for abnormal values ​​of various indicators. For example, AMH fluctuates greatly in the past two months. Combined with the collected user information, the prediction model can be used to determine whether it is normal.

[0034] The user information collected through the user's basic information, test indicators, and questionnaires is formatted and provided to the general large model as context to generate personalized health management guidance, including: diet plan, nutrient plan, health check-up, and lifestyle recommendations.

[0035] Possible diseases are predicted based on historical test results and user information through a multi-classification decision tree prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flowchart of user input of ovarian health test report in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0037] Figure 2 A user report flow chart of the ovarian health test report in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0038] Figure 3 This is a flow chart of ovarian health assessment in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0039] Figure 4 It is a user report flow chart of the ovarian fertility reserve capacity test report in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0040] Figure 5 It is a flowchart of the fertility reserve capacity assessment in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0041] Figure 6 It is a user report flow chart of the sex hormone six-item assessment test report in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0042] Figure 7A flowchart of the six-item evaluation of sex hormones in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0043] Figure 8 It is a nutrient intake reminder flow chart of the health management module in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0044] Fig. 9 It is a flow chart of the timed push reminder of the health management module in the female endocrine test report interpretation system based on artificial intelligence of the present invention;

[0045] Fig.10 This is a flow chart of the entire report engine in the artificial intelligence-based female endocrine detection report interpretation system of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. 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.

[0047] See also Figure 1-Figure 7 , the present invention provides a technical solution:

[0048] See also Figure 1-Figure 3 As shown in the figure, the AI-based female endocrine test report interpretation system is used to understand the ovarian reserve function status, screen the risk of ovarian diseases early, and detect the ovarian age by checking the AMH index, including:

[0049] Detection indicator module: detect AMH, namely anti-Mullerian hormone;

[0050] User input module: Users input basic information and menstrual status through the mini program:

[0051] Date of birth (age);

[0052] Height and weight were used to calculate BMI;

[0053] Fertility requirements;

[0054] irregular menstrual frequency (none, 1-2 times, more than 3 times, menopause);

[0055] Menstrual cycle (less than 21 days, 21-25 days, 26-30 days, 31-35 days, more than 35 days);

[0056] User report module: The user inputs basic information and menstrual status. When the AMH value is detected after blood sampling, a user report is generated through the report generation engine. The user report mainly includes: test results, overview, result interpretation, expert advice and health management;

[0057] Report overview, interpretation, and expert advice knowledge base module: includes a three-layer structure:

[0058] The first level: ages 18 to 50, with 5 years old in each age group and 50 or older in each age group;

[0059] The second level: detection index AMH, disease category;

[0060] Based on age and test indicators, derive report overview and report interpretation;

[0061] The third layer: user information: whether there is a desire to have children, whether the menstruation is regular;

[0062] Expert recommendations are derived based on age (first level), test indicators (second level), fertility requirements, and menstrual patterns;

[0063] Report content derivation module:

[0064] Includes report overview, interpretation, and expert advice paradigm;

[0065] Paradigm: R = fn(a,amh,f,m);

[0066] R: Overview, Interpretation, Expert Advice;

[0067] a: parameter age;

[0068] amh: parameter detection index AMH value;

[0069] f: parameter fertility requirement;

[0070] m: parameter menstrual regularity;

[0071] It also includes the derivation process:

[0072] The user enters basic information such as age (a), BMI (b), fertility requirements (f), and menstrual regularity (m)) through the mini program, and then collects samples and sends them back to the testing laboratory;

[0073] After the testing laboratory receives the sample and completes the test on the machine, it pushes the test indicator AMH value (AMH) to the report generation engine;

[0074] The report generation engine derives the overview, interpretation and disease type of the user report based on the parameters a and amh according to the R=fn(a,amh,f,m) paradigm;

[0075] Generate expert advice for users based on their fertility requirements and menstrual regularity;

[0076] According to the R=fn(a,amh,f,m) paradigm, produce the medical part of the user report, including overview, interpretation, and expert advice;

[0077] Example description:

[0078] A female user, 32 years old, 165cm tall, 55kg in weight (BMI is 19.9, within the normal range), has not given birth yet but plans to get pregnant in the next two years. She noticed that her menstrual cycle has become irregular recently, occasionally coming a few days early or late, so she decided to use the AI-based ovarian health test report interpretation system to check her ovarian health status;

[0079] Step 1: User input;

[0080] The user entered her basic information and menstrual status through the mini program:

[0081] Date of birth: January 1990;

[0082] Height: 165cm;

[0083] Weight: 55kg;

[0084] Fertility requirements: Yes, planning to get pregnant within the next two years;

[0085] Irregular menstruation frequency: 1-2 times;

[0086] Menstrual cycle: usually 28-30 days, but irregular recently;

[0087] Step 2: Sample testing;

[0088] The user followed the system's instructions, went to the designated testing point, collected a blood sample, and sent it to the testing laboratory. After receiving the sample, the laboratory conducted an AMH test and pushed the test results (for example, the AMH value was 2.5 ng / ml) to the report generation engine.

[0089] Step 3: Report generation;

[0090] The report generation engine generates an ovarian health report based on the user's age (32 years old), AMH value (2.5ng / ml), fertility requirements, and menstrual regularity using the R=fn(a,amh,f,m) paradigm; the report contains the following parts:

[0091] Overview: The user's ovarian reserve function is within the normal range, but due to the irregular menstrual cycle, further observation is recommended;

[0092] Interpretation of results: AMH values ​​are within the normal range, indicating good ovarian reserve function, but irregular menstrual cycles may be related to endocrine changes or other factors;

[0093] Expert advice: Given that Li Hua has plans to have children, it is recommended that she maintain a good lifestyle, reduce stress, and undergo further endocrine examinations if necessary;

[0094] Health management: The system provides Li Hua with personalized health management suggestions based on her age, BMI, and disease risk, including guidance on diet, exercise, sleep, etc.;

[0095] The AI-based ovarian health test report interpretation system has demonstrated the following creative and beneficial effects in practical applications:

[0096] Personalized interpretation: The system can provide personalized ovarian health reports and expert advice based on the user's age, AMH value, fertility requirements and menstrual regularity, so that users can better understand their ovarian health status;

[0097] Early screening: By detecting the AMH value, the system can screen the risk of ovarian disease at an early stage and provide users with timely treatment suggestions to prevent the disease from worsening;

[0098] Intelligent derivation: The report generation engine uses artificial intelligence algorithms to intelligently derive the overview, interpretation, disease type and expert advice of the user report based on the information entered by the user and the test results, greatly improving the accuracy and efficiency of the report;

[0099] The implementation is:

[0100] Data Collection:

[0101] Based on rich and diverse data, including user age, AMH value, fertility requirements, menstrual regularity, living habits, family medical history, past medical history and other multi-dimensional information, for example, for a 35-year-old user with an AMH value of 1.8ng / ml, fertility requirements, irregular menstruation (more than 3 times), less exercise, and a family history of ovarian disease, these detailed data provide comprehensive input for subsequent derivation;

[0102] Data preprocessing:

[0103] Clean and standardize the collected data, such as processing possible missing values ​​(such as users forgetting to fill in the frequency of exercise), outliers (AMH values ​​that are too high or too low), and creating new features through feature engineering (such as the ratio of irregular menstruation times to AMH values);

[0104] Model selection:

[0105] You can choose a random forest model that is suitable for processing multi-feature data. For example, a 28-year-old user with an AMH value of 2.0ng / ml, no fertility needs, a menstrual cycle of 28-30 days, and regular exercise can be used to deduce the results of multiple decision trees.

[0106] Model training:

[0107] During the training process, adjust the model parameters. For example, for the data of the above example user, find the optimal model configuration by adjusting the number of decision trees, conditions for splitting nodes, and other parameters. At the same time, use regularization methods to prevent overfitting, and use early stopping methods to avoid overtraining.

[0108] Model Evaluation:

[0109] Use a variety of evaluation indicators, such as accuracy, mean square error, etc. to evaluate the accuracy of the model for predicting the user's ovarian health status;

[0110] Model optimization:

[0111] If the model does not perform well for a certain type of user (e.g., older users with very irregular menstruation), you may want to increase the amount of data for this type of user, or try transfer learning and fine-tune using a pre-trained model in a related field.

[0112] Model structure:

[0113] For neural network models, such as multi-layer perceptron, if this model is selected, for the above example user, according to data complexity and computing resources, three hidden layers may be set, each with 64 neurons, and ReLU activation function may be selected; for decision tree models, the depth of the tree may be limited to 5 layers to prevent overfitting;

[0114] Example description:

[0115] For a 32-year-old female user with an AMH value of 2.5 ng / ml, who has a fertility plan and has irregular menstruation (1-2 times), the system first obtains the information she inputs, and then the report generation engine intelligently derives her report based on the set paradigm (such as R = fn(a,amh,f,m)) and the trained model;

[0116] The overview section may state: Ovarian reserve is within normal range, but menstrual irregularities require further observation;

[0117] The interpretation section will explain: normal AMH values ​​indicate good ovarian reserve function, and irregular menstruation may be related to endocrine changes;

[0118] The type of disease may not be clear yet, but it will indicate the potential risk of endocrine disorders;

[0119] Experts will recommend that she maintain a good lifestyle based on her birth plan and menstrual pattern, and conduct endocrine examinations if necessary;

[0120] Through such intelligent deduction, users are provided with personalized, accurate and efficient report content;

[0121] In summary, through comprehensive data collection and preprocessing, reasonable model selection and training, accurate evaluation and effective optimization, and appropriate model structure design, the report generation engine can intelligently derive the overview, interpretation, disease type and expert advice of the user report based on user input and test results, greatly improving the accuracy and efficiency of the report;

[0122] Health management: The system also provides personalized health management plans to help users improve their lifestyle habits, reduce the risk of ovarian diseases, and improve fertility.

[0123] For further information, see Figure 4-Figure 5 As shown, this system can also be used to interpret the test report of ovarian fertility reserve capacity;

[0124] By checking AMH, six sex hormones FSH, LH, E2, P, PRL, T indicators, the ovarian fertility is comprehensively evaluated;

[0125] Detection indicators:

[0126] AMH: Anti-Mullerian hormone;

[0127] FSH: follicle stimulating hormone;

[0128] LH: luteinizing hormone;

[0129] E2: estradiol;

[0130] P: progesterone;

[0131] PRL: prolactin;

[0132] T: testosterone;

[0133] User input:

[0134] Users enter basic information, fertility requirements, and menstrual status through the mini program:

[0135] Date of birth (age);

[0136] Height and weight were used to calculate BMI;

[0137] Fertility requirements;

[0138] irregular menstrual frequency (none, 1-2 times, more than 3 times, menopause);

[0139] Menstrual cycle (less than 21 days, 21-25 days, 26-30 days, 31-35 days, more than 35 days);

[0140] Number of days of menstruation;

[0141] User reports:

[0142] The user enters basic information and menstrual status. After sampling, the AMH value and the values ​​of six sex hormones are detected, and a user report is generated through the report generation engine. The user report mainly includes: test results, overview, result interpretation, expert advice and health management;

[0143] Report overview, interpretation, expert advice knowledge base:

[0144] The knowledge base has four main layers:

[0145] The first level: Age 18 to 50, with 5-year-olds in each age group and 50 or older in each age group. There is no age limit.

[0146] The second level: detection index AMH;

[0147] The third level: six sex hormones (FSH, LH, E2, P, PRL, T), disease categories;

[0148] Based on age and test indicators, derive report overview and report interpretation;

[0149] The fourth layer: user information: whether there is a desire to have children, whether the menstruation is regular;

[0150] Expert recommendations are derived based on age (first layer), test indicators (second and third layers), fertility requirements, and menstrual patterns;

[0151] Derivation of medical content of the report:

[0152] Report overview, interpretation, expert advice paradigm:

[0153] Paradigm: R = fn(a,amh,sh,f,m);

[0154] R: Overview, Interpretation, Expert Advice;

[0155] a: parameter age;

[0156] amh: parameter detection index AMH value;

[0157] sh: values ​​of six sex hormone parameters (FSH, LH, E2, P, PRL, T);

[0158] f: parameter fertility requirement;

[0159] m: parameter menstrual regularity;

[0160] Derivation process:

[0161] The user enters basic information such as age (a), BMI (b), fertility requirements (f), and menstrual regularity (m)) through the mini program, and then takes the sample and sends it back to the testing laboratory;

[0162] After the testing laboratory receives the sample and completes the test on the machine, it pushes the test index AMH (AMH), the values ​​of the six sex hormone parameters FSH, LH, E2, P, PRL, and T (sh) to the report generation engine;

[0163] The report generation engine derives the overview, interpretation and disease type of the user report based on the parameters a, amh and sh according to the R = fn(a,amh,sh,f,m) paradigm;

[0164] Derive expert advice for users based on their fertility requirements and menstrual regularity;

[0165] According to the R=fn(a,amh,sh,f,m) paradigm, produce the medical part of the user report, including overview, interpretation, and expert advice;

[0166] Practical application examples:

[0167] A 30-year-old female user is considering having a child in the next two years. The woman has had regular menstruation, but recently she has been feeling a little anxious about whether her ovarian fertility reserve is sufficient. She has taken a test using the AI-based ovarian health test report interpretation system.

[0168] User input:

[0169] Date of birth: March 1993;

[0170] Height: 165cm;

[0171] Weight: 55kg;

[0172] Fertility requirements: Yes, plan to have a baby within two years;

[0173] Irregular menstruation frequency: None;

[0174] Menstrual cycle: 28-30 days;

[0175] Menstrual period: 5 days;

[0176] Testing process:

[0177] The woman completed blood sampling at the designated sampling point and sent the sample to the testing laboratory; the laboratory tested the woman's AMH value and the values ​​of six sex hormones (FSH, LH, E2, P, PRL, T) using high-precision instruments;

[0178] Report Generation:

[0179] After the test is completed, the laboratory pushes the test results to the report generation engine; the report generation engine intelligently derives the user report based on the woman’s age, BMI, fertility requirements, menstrual regularity, and the detected AMH value and the values ​​of the six sex hormones through the R=fn(a,amh,sh,f,m) paradigm;

[0180] Report Contents:

[0181] Test results: Displays the specific values ​​of the woman's AMH, FSH, LH, E2, P, PRL, and T, and compares them with the normal range;

[0182] Overview: Based on the woman’s age and test results, the system assessed that her ovarian reproductive reserve capacity was above average, and she was suitable for childbearing within the next two years;

[0183] Interpretation of results: The physiological significance of AMH values ​​and six sex hormones, as well as their role in the assessment of ovarian reproductive reserve capacity, are explained in detail;

[0184] Expert advice: Based on the woman's fertility requirements and menstrual regularity, the system provides targeted health advice and management plans, such as a proper diet, moderate exercise, and avoiding excessive stress, to improve ovarian fertility reserve capacity;

[0185] Health management: The system provides the woman with a personalized health management plan, including regular testing, healthy diet, exercise plan, etc., to help her better maintain ovarian health;

[0186] Effect:

[0187] Comprehensive assessment of ovarian reproductive reserve capacity: By testing the AMH value and the values ​​of six sex hormones, the system can comprehensively assess the ovarian reproductive reserve capacity and provide users with more accurate assessment results;

[0188] Personalized health advice: The system provides users with personalized health advice and management plans based on factors such as age, BMI, fertility requirements, and menstrual regularity to help users better maintain ovarian health;

[0189] Early screening and prevention: Through regular testing, the system can screen out the risk of ovarian disease at an early stage, provide users with timely treatment suggestions, and prevent the disease from worsening; at the same time, through the implementation of health management plans, the risk of ovarian disease can be reduced and fertility can be improved;

[0190] Improve detection efficiency and accuracy: The use of artificial intelligence algorithms for report generation and interpretation has greatly improved detection efficiency and accuracy. At the same time, through automated processing, the possibility of human intervention and errors is reduced;

[0191] Strong application scalability: The system is not only suitable for the interpretation of ovarian health test reports, but can also be extended to the interpretation of other test items related to ovarian health, including ovarian fertility reserve capacity testing.

[0192] For further information, see Figure 6-Figure 7 As shown, this system can also conduct endocrine abnormality assessment, menstrual irregularity assessment, pregnancy preparation health guidance, and endocrine health management by checking six sex hormone indicators: FSH, LH, E2, P, PRL, and T.

[0193] Detection indicators:

[0194] FSH: follicle stimulating hormone;

[0195] LH: luteinizing hormone;

[0196] E2: estradiol;

[0197] P: progesterone;

[0198] PRL: prolactin;

[0199] T: testosterone;

[0200] User input:

[0201] Users enter basic information, fertility requirements, and menstrual status through the mini program:

[0202] Date of birth (age);

[0203] Height and weight were used to calculate BMI;

[0204] Fertility requirements;

[0205] irregular menstrual frequency (none, 1-2 times, more than 3 times, menopause);

[0206] Menstrual cycle (less than 21 days, 21-25 days, 26-30 days, 31-35 days, more than 35 days);

[0207] Number of days of menstruation;

[0208] User reports:

[0209] The user inputs basic information and menstrual status, and after sampling, the values ​​of six sex hormones are tested, and a user report is generated through the report generation engine. The user report mainly includes: test results, overview, result interpretation, expert advice and health management;

[0210] Report overview, interpretation, expert advice knowledge base:

[0211] The knowledge base has three main layers:

[0212] The first level: ages 18 to 40, with an age group of 40 or above, and no age limit;

[0213] The second level: six sex hormones (FSH, LH, E2, P, PRL, T) and disease categories;

[0214] Based on age and test indicators, derive report overview and report interpretation;

[0215] The third layer: user information: whether there is a desire to have children, whether the menstruation is regular;

[0216] Expert recommendations were derived based on age (first level), six test index hormones (second level), fertility requirements, and menstrual regularity;

[0217] Derivation of medical content of the report:

[0218] Report overview, interpretation, expert advice paradigm:

[0219] Paradigm: R = fn(a, sh, f, m);

[0220] R: Overview, Interpretation, Expert Advice;

[0221] a: parameter age;

[0222] sh: values ​​of six sex hormone parameters (FSH, LH, E2, P, PRL, T);

[0223] f: parameter fertility requirement;

[0224] m: parameter menstrual regularity;

[0225] Derivation process:

[0226] The user enters basic information such as age (a), BMI (b), fertility requirements (f), and menstrual regularity (m)) through the mini program, and then collects samples and sends them back to the testing laboratory;

[0227] After the testing laboratory receives the sample and completes the test on the machine, it pushes the values ​​of the six test hormone parameters FSH, LH, E2, P, PRL, and T(sh) to the report generation engine;

[0228] The report generation engine derives the overview, interpretation and disease type of the user report based on the parameters a and sh according to the R=fn(a,amh,f,m) paradigm;

[0229] Generate expert advice for users based on their fertility requirements and menstrual regularity;

[0230] According to the R=fn(a,sh,f,m) paradigm, produce the medical part of the user report, including overview, interpretation, and expert advice;

[0231] Practical application examples:

[0232] A 35-year-old female user is preparing for pregnancy. Her menstruation has been irregular, and she is worried that this will affect her pregnancy plan. She decided to use our AI-based ovarian health test report interpretation system to test six sex hormones.

[0233] User input:

[0234] Date of birth: May 1988;

[0235] Height: 160cm;

[0236] Weight: 60kg;

[0237] Fertility requirements: Yes, I am trying to get pregnant;

[0238] Irregular menstruation frequency: more than 3 times;

[0239] Menstrual cycle: 32-35 days;

[0240] Menstrual period: 4 days;

[0241] Testing process:

[0242] The woman completed blood sampling at the designated sampling point and sent the sample to the testing laboratory; the laboratory tested the woman's six sex hormone values ​​(FSH, LH, E2, P, PRL, T) using high-precision instruments;

[0243] Report Generation:

[0244] After the test is completed, the laboratory will push the test results to the report generation engine; the report generation engine will intelligently derive the user report based on the woman’s age, fertility requirements, menstrual regularity, and the values ​​of the six sex hormones detected, using the R=fn(a,sh,f,m) paradigm;

[0245] Report Contents:

[0246] Test results: Displays the specific values ​​of the woman's six sex hormones and compares them with the normal range;

[0247] Overview: Based on the woman's age and test results, the system assessed that her sex hormone levels were slightly abnormal, which may cause irregular menstruation, but she was still fertile overall;

[0248] Interpretation of results: The physiological significance of the six sex hormones and their role in assessing menstrual regularity and pregnancy health were explained in detail; the possible causes and effects of the abnormal hormone levels of the woman were particularly pointed out;

[0249] Expert advice: Based on the woman's fertility requirements and irregular menstruation, the system provides targeted health advice and management plans; for example, it recommends that the woman adjust her lifestyle, including a healthy diet, moderate exercise, and reduced stress, to promote endocrine balance and regular menstruation. At the same time, it also provides special treatment advice for abnormal sex hormones;

[0250] Health management: The system developed a personalized health management plan for the woman, including regular testing of sex hormone levels, monitoring of menstruation, and adjustment of lifestyle habits, to help her better maintain ovarian health and prepare for pregnancy;

[0251] Effect:

[0252] Comprehensive assessment of endocrine abnormalities and irregular menstruation: By detecting the values ​​of six sex hormones, the system can comprehensively assess the user's endocrine status and menstrual conditions, providing users with a more comprehensive health assessment;

[0253] Pregnancy preparation health guidance: The system provides targeted pregnancy preparation health guidance based on the user's fertility requirements and menstrual conditions, helping users to better formulate pregnancy preparation plans and improve the success rate of pregnancy;

[0254] Endocrine health management: The system helps users adjust their lifestyle and treatment plans through personalized health management plans, promote endocrine balance and menstrual regularity, and reduce the risk of ovarian diseases;

[0255] Improve detection efficiency and accuracy: The use of artificial intelligence algorithms for report generation and interpretation greatly improves detection efficiency and accuracy; at the same time, the automated processing process reduces the possibility of human intervention and errors;

[0256] Strong application scalability: The system is not only suitable for the interpretation of ovarian health test reports, but can also be extended to the interpretation of other test items related to endocrine and menstrual health, including the six-item evaluation of sex hormones; by continuously optimizing algorithms and adding functional modules, it can meet the personalized needs of more users; for example, the system can be further expanded to the evaluation and management of diseases such as polycystic ovary syndrome (PCOS) and menopausal syndrome.

[0257] Furthermore, it also includes a health management module; that is, the health management plan included in the six assessment reports of ovarian health, fertility reserve capacity, and sex hormones, the disease types derived according to the medical content deduction paradigms R=fn(a,amh,f,m), R=fn(a,amh,f,m) and R=fn(a,amh,sh,f,m) of the three products, combined with the user's age and BMI value, generate the user's health management content; the health management module includes a health management knowledge base.

[0258] Furthermore, the health management module also includes health management paradigms;

[0259] Normal form: H = fn(a,c,b);

[0260] H: Health management;

[0261] a: parameter age;

[0262] c: parameter disease category;

[0263] b: parameter bmi.

[0264] Furthermore, the health management module also includes a derivation process: based on the disease type (c), the user's basic information: age (a), bmi (b), the user's health management plan is derived according to the H=fn(a,c,b) paradigm; based on the medical content overview, interpretation, expert advice and health management plan, a complete report for the user is finally generated.

[0265] For further information, see Figure 8-Figure 9 As shown in the figure, in the health management module, push notification and health reminder functions are introduced; the health manager sets reminders for users to take nutrients, track exercise, etc., and sends them to users: users can see the reminders created by the nutritionist in the mini program, and the push service will push reminders regularly;

[0266] Implementation of technical architecture:

[0267] Backend server: responsible for generating push content and processing user feedback;

[0268] Message queue: used to asynchronously process push tasks to ensure efficient and stable push;

[0269] Push service: send push content to target users;

[0270] Implementation process:

[0271] Trigger push: trigger push tasks based on preset rules (such as user birthday, physical examination date, medication taking time, etc.) or user behavior (such as push related health knowledge after the user views a specific report);

[0272] Generate push content: The backend server generates push content based on trigger conditions and user information;

[0273] Send push: Send push content to the message queue, and the push service will handle the sending task asynchronously.

[0274] The user information collected through the user's basic information, test indicators, and questionnaires is formatted and provided to the general large model as context to generate personalized health management guidance, including: diet plan, nutrient plan, health check-up, and lifestyle suggestions; that is, personalized health management is generated through AI GC; specifically:

[0275] User basic information: age, height, weight;

[0276] Questionnaire collects information:

[0277]

[0278]

[0279]

[0280]

[0281]

[0282] Technical implementation:

[0283] Collect user information, test index values, living habits, symptoms, medical history, etc. through mini-program questionnaires

[0284] Standardize the format of the collected answers to the questions into contextual descriptions that the big model can understand

[0285] for example:

[0286] What is your main health concern or need? Select “Weight Management”;

[0287] Planning to have a baby? Select "Yes";

[0288] Do you smoke? Select "No";

[0289] The answers to these three questions are formatted and annotated, and finally become:

[0290] Users are more concerned about weight management, have plans to get pregnant, and do not smoke.

[0291] Implementing the Promopt project:

[0292] Role definition: You are the health manager of Yishengjiankang R&D;

[0293] Instructions: Health management plans that require big models to answer, including diet plans, nutrient plans, health examinations, lifestyle suggestions, etc.

[0294] Input: context description and health management knowledge base in the standardized format in step 2;

[0295] Output: The format in which the large model is output.

[0296] Output personalized health management plans through universal big models.

[0297] Implementation example:

[0298] Basic information: Age 45;

[0299] Detection index: AMH 0.05;

[0300] Questionnaire information:

[0301]

[0302]

[0303] Based on the above user's basic information, test indicators, and questionnaire information, a prompt project is generated. Through the large model, a personalized health management plan including diet plan, nutrient plan, health check-up, and lifestyle suggestions is generated:

[0304] Low-carb diet:

[0305] Pay attention to increasing the intake of high-quality protein, especially: eggs, milk and dairy products, beans and bean products, meat (fish, shrimp, poultry, lean beef / pork);

[0306] Eat 1 catties of fresh vegetables every day (1 bowl for each meal), and give priority to dark leafy vegetables;

[0307] Eat moderate amounts of fruit, and choose low-sugar fruits, such as peaches, blueberries, citrus fruits, figs, cherry tomatoes, etc.;

[0308] Pay attention to calcium supplementation in your diet, drink milk (300g) every day, and if you are lactose intolerant, you can choose yogurt.

[0309] Other calcium-supplementing foods include: cheese, soy products, water spinach, edamame, red amaranth and other dark vegetables. Or take calcium tablets under guidance.

[0310] Nutrient program:

[0311] The goal is to enhance antioxidant capacity, improve sleep, and strengthen bone nutrition on the basis of diet.

[0312] Suggested nutrients to consider adding:

[0313] Magnesium threonate;

[0314] sleep support (inhibitory neurotransmitters);

[0315] Black cohosh;

[0316] Calcium supplements;

[0317] Health Checkup:

[0318] Health check-ups should be performed regularly and abnormal data should be fed back to adjust health plans:

[0319] Focus: 25-hydroxyvitamin D, bone density, etc.

[0320] Note: The test results of 25-hydroxyvitamin D can indicate the nutritional status of vitamin D. Vitamin D should be supplemented as needed according to the test results. It is also recommended to regularly check gynecological items under the guidance of a doctor.

[0321] Lifestyle recommendations: Try to engage in outdoor sports and increase strength training to exercise skeletal muscles under professional guidance; avoid contact with reproductive toxic substances, especially plastic tableware for food: use less plastic products, including disposable paper cups.

[0322] Furthermore, a user feedback mechanism is established in the system to collect users’ opinions and suggestions on the system;

[0323] Design feedback interface:

[0324] Design an easily accessible feedback interface in the user report or system applets. This interface can be a simple form for users to fill in their opinions and suggestions. The form can include some preset questions or options, such as "What aspects of the system are you satisfied with?", "What aspects of the system do you think can be improved?", etc., to collect more specific feedback. You can also provide a text input box for users to freely enter their opinions and suggestions.

[0325] Implement feedback submission function:

[0326] Write backend code to receive and process feedback data submitted by users; this data can be stored in a database for subsequent analysis and processing; to ensure data integrity and security, the submitted data should be validated and cleaned to prevent malicious input or invalid data;

[0327] Display feedback results:

[0328] A feedback management interface can be set up in the system to allow system administrators or developers to view and process feedback submitted by users; this interface can display detailed information of each feedback, including submission time, user information, feedback content, etc.; system administrators or developers can improve and optimize the system based on these feedbacks;

[0329] Set up feedback processing flow:

[0330] Establish a clear feedback processing process to ensure that each feedback is handled promptly and effectively. For example, you can set different levels of feedback processing priority to give priority to important or urgent feedback. At the same time, you can also set a feedback processing time limit to ensure that each feedback can be responded to within a reasonable time.

[0331] Send feedback response:

[0332] When the system administrator or developer has processed the user's feedback, a feedback response should be sent to the user to inform them that the feedback has been received and processed, and to explain the processing results or improvement plan.

[0333] For further information, see Fig.10 As shown in the figure, a binary classification prediction model is trained to quickly and accurately determine abnormal values ​​of indicators; specifically:

[0334] Artificial intelligence screening of abnormal values ​​of various indicators:

[0335] Solution: The binary classification prediction model learns user information and detection indicator data, and can perform artificial intelligence screening for abnormal values ​​of various indicators. For example, AMH fluctuates greatly in the past two months. Combined with the collected user information, the prediction model can be used to determine whether it is normal.

[0336] The main information collected is as follows:

[0337] Index information:

[0338] Screening indicators include:

[0339] AMH: Anti-Mullerian hormone;

[0340] Six sex hormones: FSH: follicle stimulating hormone, LH: luteinizing hormone, E2: estradiol, P: progesterone, PRL: prolactin, T: testosterone;

[0341] VD: Vitamin D;

[0342] HPV: 23 types;

[0343] Screening index values ​​include:

[0344] Last indicator test value: (1.2);

[0345] The test value of this indicator is: (4.3);

[0346] User Information:

[0347] Basic information: age, height, weight;

[0348] Lifestyle: whether you smoke, stay up late, or drink alcohol;

[0349] Taking medications: birth control pills, vitamin D, metformin, psychiatric drugs, hormone drugs;

[0350] Medical history: polycystic ovary syndrome, ovarian (reserve) dysfunction, hyperandrogenism, premature ovarian insufficiency, hypothyroidism, hyperthyroidism, ovarian cysts;

[0351] Menstrual status:

[0352] Current menstrual status:

[0353] Last menstrual period date (2024-07-11);

[0354] Period: (5 days);

[0355] Cycle: (28 days);

[0356] Menstrual volume: (feeling more, less, moderate);

[0357] Whether menstruation is regular (basically regular, 1-2 times irregular, 3 times or more irregular);

[0358] Menopause: (yes, no);

[0359] Menstrual status in the last three years;

[0360] Technical implementation:

[0361] First, collect user information, test index values, lifestyle habits, symptoms, medical history, etc. through the mini program questionnaire;

[0362] Standardize and normalize user information;

[0363] The processed data is input into the binary classification prediction model for learning and prediction;

[0364] Finally, the probability of abnormal value of the detection index value is obtained;

[0365] The unreasonable indicator values ​​are obtained through probability sorting and fed back to the user.

[0366] Implementation details:

[0367] Data preparation:

[0368] First, normalize the data;

[0369] Age: The collected age data is relatively discrete. In the final diagnosis, the age will be divided into the following categories. According to the distribution of women served by our business and system, it can basically be divided into the following age groups;

[0370] [0,18), [18-25), [25,30), [30,35), [35,40), [40,45), [45,50),

[0371] [50,+∞);

[0372] Height and weight: Height and weight information is processed through BMI. According to the definition of the World Health Organization, BMI is divided into the following levels:

[0373] Underweight: (0,185], Normal weight: [18.5,24.9], Overweight: [25,29.9], Obesity level I: [30-34.9], Obesity level II: [35-39.9], Obesity level III: [40,+∞);

[0374] Menstrual cycle: Medically defined, a normal menstrual cycle is 21-35 days: [21, 35];

[0375] <21 days: frequent menstruation;

[0376] >35 days: infrequent menstruation;

[0377] [21,35]: Normal menstruation;

[0378] Menstrual period: Medically defined, the menstrual period is generally 3 to 7 days;

[0379] >7 days: prolonged menstruation;

[0380] <3 days: short menstruation;

[0381] [3,7] days: normal menstrual period;

[0382] Others: For some symptoms, only yes or no is available, so they are directly defined as 0 and 1;

[0383] Feature Engineering:

[0384] Considering that many medical symptoms are strongly correlated, we focused on doing a lot of tuning work on processing category features and feature interactions.

[0385] Processing categorical features: LightGBM supports directly processing categorical features, but proper encoding and processing can still improve model performance, especially when there are strong relationships between features.

[0386] Feature interactions: Creating new features (such as feature combinations) can help the model learn complex relationships, which is crucial in some application scenarios.

[0387] Common feature intersection methods are:

[0388] Crossover of eigenvalues: that is, new features are formed by adding or multiplying eigenvalues.

[0389] Feature category intersection: combine two or more category features to form a new combined feature.

[0390] In medical scenarios, simple feature crossover may not be appropriate, as they have certain medical logic. We implemented the feature value crossover algorithm based on certain medical logic.

[0391] Menstruation related:

[0392]

[0393] Parameter Description:

[0394] T i : represents the length of the i-th menstrual cycle (in days), and the data is obtained through user records;

[0395] N: indicates the number of menstrual cycles observed, N is 3-6 cycles (i.e. data from the past 3 months to half a year)

[0396] μ: represents the mean and standard deviation of the menstrual cycle length, which can be calculated from the observed data.

[0397] δ: represents the threshold for judging irregular menstruation (can be the number of days or multiples of standard deviation), which is generally 7 days in medicine;

[0398] Binary data:

[0399] The prediction model has the ability to predict whether the test value is normal by learning the data labeled with two categories (1 for normal, 0 for abnormal);

[0400] Examples of normal values ​​for binary classification:

[0401] Examples of abnormal AMH values:

[0402] Detection index: AMH;

[0403] Age: 24

[0404] Weight: 50kg;

[0405] Height: 159cm;

[0406] The last AMH index test value: 1.7; The current AMH index value: 1.5;

[0407] Do you smoke: No;

[0408] Stay up late: No;

[0409] Do you drink alcohol: No;

[0410] Taking medication: No;

[0411] Medical history: No;

[0412] Irregular menstruation: No;

[0413] Last menstrual period: 2024-07-11; Menstrual period: 5 days;

[0414] Cycle: 28 days;

[0415] Is menstruation regular: basically regular; Menopause: no;

[0416] Is it normal value: 0;

[0417] Examples of normal AMH values:

[0418] Detection index: AMH;

[0419] Age: 24

[0420] Weight: 50kg;

[0421] Height: 159cm;

[0422] The last AMH index test value: 1.1; the current AMH index value: 1.2;

[0423] Do you smoke: No;

[0424] Stay up late: No;

[0425] Do you drink alcohol: No;

[0426] Taking medication: No;

[0427] Medical history: ovarian (reserve) dysfunction;

[0428] Irregular menstruation: No;

[0429] Last menstrual period: 2024-07-11;

[0430] Menstrual period: 5 days;

[0431] Cycle: 28 days;

[0432] Is menstruation regular: 1-2 times irregular;

[0433] Menopause: No;

[0434] Is it normal value: 1;

[0435] Examples of abnormal values ​​of six sex hormones:

[0436] Detection indicators: 6 items of sex hormones;

[0437] Age: 24

[0438] Weight: 50kg;

[0439] Height: 159cm;

[0440] The last test values ​​of the six sex hormone indicators were: P = 0.6, FSH = 7, LH = 10, E2 = 10, PRL = 20, T = 0.5;

[0441] The six sex hormone index values ​​were: P = 0.6, FSH = 7, LH = 10, E2 = 30, PRL = 20, T = 0.5;

[0442] Do you smoke: No;

[0443] Stay up late: No;

[0444] Do you drink alcohol: No;

[0445] Taking medication: No;

[0446] Medical history: occult ovarian insufficiency;

[0447] Irregular menstruation: No;

[0448] Last menstrual period: 2024-07-11;

[0449] Menstrual period: 5 days;

[0450] Cycle: 28 days;

[0451] Is menstruation regular: basically regular;

[0452] Menopause: No;

[0453] Is it normal value: 0;

[0454] Examples of normal values ​​for six sex hormones:

[0455] Detection index: AMH;

[0456] Age: 24

[0457] Weight: 50kg;

[0458] Height: 159cm;

[0459] The last AMH index test value: P = 0.6, FSH = 7, LH = 10, E2 = 10, PRL = 20, T = 0.5; the current AMH index value: P = 0.6, FSH = 7, LH = 10, E2 = 12, PRL = 20, T = 0.5;

[0460] Do you smoke: No;

[0461] Stay up late: No;

[0462] Do you drink alcohol: No;

[0463] Taking medication: No;

[0464] Medical history: ovarian (reserve) dysfunction;

[0465] Irregular menstruation: No;

[0466] Last menstrual period: 2024-07-11; Menstrual period: 5 days;

[0467] Cycle: 28 days;

[0468] Is menstruation regular: 1-2 times irregular; Menopause: No;

[0469] Is it normal value: 1;

[0470] Model training:

[0471]

[0472]

[0473] Furthermore, a multi-classification decision tree prediction model is used to predict possible diseases based on historical test results and user information; specifically:

[0474] Through the multi-classification decision tree prediction model, possible diseases are predicted based on historical test results and user information. The early predictions of ovarian-related diseases include:

[0475] Latent diminished ovarian reserve (latent DOR);

[0476] Subclinical ovarian dysfunction;

[0477] Diminished ovarian reserve (DOR);

[0478] Subclinical premature ovarian insufficiency (sub-POI);

[0479] Diminished ovarian reserve;

[0480] Failure of ovarian reserve;

[0481] Be aware of the risk of PCOS;

[0482] PCOS polycystic ovary syndrome risk;

[0483] High risk of PCOS;

[0484] Hyperprolactinemia;

[0485] Low gonadotropin;

[0486] Risk of premature ovarian insufficiency (POI);

[0487] Risk of premature ovarian failure (POF);

[0488] High ovarian reserve;

[0489] Latent ovarian insufficiency;

[0490] Ovarian dysfunction;

[0491] The main information collected is as follows:

[0492] Historical indicator values:

[0493] AMH: Anti-Mullerian hormone;

[0494] Six sex hormones: FSH: follicle stimulating hormone, LH: luteinizing hormone, E2: estradiol, P: progesterone, PRL: prolactin, T: testosterone;

[0495] VD: Vitamin D;

[0496] HPV: 23 types;

[0497] User Information:

[0498] Basic information: age, height, weight;

[0499] lifestyle;

[0500] Whether you smoke;

[0501] Stay up late?

[0502] Whether you drink alcohol;

[0503] Taking medications: birth control pills, vitamin D, metformin, psychiatric drugs, hormone drugs;

[0504] symptom:

[0505] Oily skin and recurring acne;

[0506] Localized hirsutism (short and thick black hair appears on the lips, beard, inner thighs, lower abdomen, areola, etc.);

[0507] Acanthosis nigricans (thickening of the skin with gray-brown pigmentation on the neck, armpits, and groin);

[0508] Frequent gastrointestinal discomfort;

[0509] Severe hair loss;

[0510] Anxiety, irritability, and mood swings;

[0511] Hot flashes, sweating;

[0512] Sleep disturbances;

[0513] Menstrual status:

[0514] Current menstrual status:

[0515] Last menstrual period date (2024-07-11);

[0516] Menstrual period: (5 days);

[0517] Cycle: (28 days);

[0518] Menstrual volume: (feeling more, less, moderate);

[0519] Menstrual changes:

[0520] Changes in the number of days of menstruation: (extended, shortened, no significant changes);

[0521] Changes in cycle days: (extended, shortened, no significant changes);

[0522] Changes in menstrual volume: (increase, decrease, no significant change);

[0523] Whether menstruation is regular (basically regular, 1-2 times irregular, 3 times or more irregular);

[0524] Menopause: (yes, no);

[0525] Menstrual status in the last three years;

[0526] Technical implementation:

[0527] First, collect user information, multiple historical test index values, living habits, symptoms, medical history, etc. through the mini program questionnaire;

[0528] Standardize and normalize user information;

[0529] The processed data is input into the multi-classification decision tree prediction model for prediction;

[0530] Finally, we get the probability of the disease we may have.

[0531] Obtain possible diseases through probability sorting and provide feedback to users;

[0532] The core of this process is to train a multi-classification decision tree prediction model, which can quickly predict possible diseases.

[0533] Multi-class data:

[0534] The prediction model has the ability to predict possible diseases by learning from labeled data;

[0535] Example 1:

[0536] Detection index: AMH;

[0537] Age: 24

[0538] Weight: 50kg;

[0539] Height: 159cm;

[0540] Last AMH index test value: 8;

[0541] AMH index value this time: 7;

[0542] Do you smoke: No;

[0543] Stay up late: No;

[0544] Do you drink alcohol: No;

[0545] Taking medication: No;

[0546] Symptoms: Localized hirsutism;

[0547] Irregular menstruation: No;

[0548] Last menstrual period: 2024-07-11;

[0549] Menstrual period: 5 days;

[0550] Cycle: 28 days;

[0551] Is menstruation regular: 1-2 times irregular;

[0552] Menopause: No;

[0553] Possible diseases: Be aware of the risk of PCOS;

[0554] Example 2:

[0555] Detection indicators: 6 items of sex hormones;

[0556] The last test values ​​of the six sex hormone indicators were: P = 0.6, FSH = 7, LH = 10, E2 = 10, PRL = 20, T = 0.5;

[0557] The six sex hormone index values ​​were: P = 0.6, FSH = 7, LH = 10, E2 = 30, PRL = 20, T = 0.5;

[0558] Age: 24

[0559] Weight: 50kg;

[0560] Height: 159cm;

[0561] Do you smoke: No;

[0562] Stay up late: No;

[0563] Do you drink alcohol: No;

[0564] Taking medication: No;

[0565] Symptoms: severe hair loss, anxiety, irritability, and emotional instability;

[0566] Irregular menstruation: No;

[0567] Last menstrual period: 2024-07-11;

[0568] Menstrual period: 5 days;

[0569] Cycle: 28 days;

[0570] Is menstruation regular: 1-2 times irregular;

[0571] Menopause: No;

[0572] Possible diseases: ovarian (reserve) dysfunction; model training;

[0573]

[0574]

[0575] Summarize:

[0576] Personalized interpretation: The system can provide personalized ovarian health reports and expert advice based on the user's age, AMH value, fertility requirements and menstrual regularity, so that users can better understand their ovarian health status;

[0577] Early screening: By detecting the AMH value, the system can screen the risk of ovarian disease at an early stage and provide users with timely treatment suggestions to prevent the disease from worsening;

[0578] Intelligent derivation: The report generation engine uses artificial intelligence algorithms to intelligently derive the overview, interpretation, disease type and expert advice of the user report based on the information entered by the user and the test results, greatly improving the accuracy and efficiency of the report;

[0579] Health management: The system also provides personalized health management plans to help users improve their lifestyle habits, reduce the risk of ovarian diseases, and improve fertility.

[0580] Strong application scalability: This system is not only suitable for the interpretation of ovarian health test reports, but can also be extended to the interpretation of other test items related to ovarian health, including ovarian reproductive reserve capacity detection and six sex hormone assessments; by continuously optimizing algorithms and adding functional modules, it can meet the personalized needs of more users.

[0581] The binary classification prediction model learns user information and detection indicator data, and can perform artificial intelligence screening for abnormal values ​​of various indicators. For example, AMH fluctuates greatly in the past two months. Combined with the collected user information, the prediction model can be used to determine whether it is normal.

[0582] The user information collected through the user's basic information, test indicators, and questionnaires is formatted and provided to the general large model as context to generate personalized health management guidance, including: diet plan, nutrient plan, health check-up, and lifestyle recommendations.

[0583] Possible diseases are predicted based on historical test results and user information through a multi-classification decision tree prediction model.

Claims

1. The female endocrine test report interpretation system based on artificial intelligence is characterized by: It is used to check the AMH index to understand the status of ovarian reserve function, screen the risk of ovarian disease early, and detect the age of the ovaries, including: Detection indicator module: detect AMH, namely anti-Mullerian hormone; User input module: Users input basic information and menstrual status through the mini program: User report module: The user inputs basic information and menstrual status. When the AMH value is detected after blood sampling, a user report is generated through the report generation engine. The user report mainly includes: test results, overview, result interpretation, expert advice and health management; Report overview, interpretation, and expert advice knowledge base module: Based on age and test indicators, report overview and report interpretation are derived; based on age, test indicators, fertility requirements, and menstrual patterns, expert advice is derived; Report content derivation module: including report overview, interpretation, and expert advice paradigm.

2. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: This system can also be used to interpret the test report of ovarian fertility reserve capacity; that is, in the test index module, AMH, six sex hormones FSH, LH, E2, P, PRL, T indicators are tested to comprehensively evaluate ovarian fertility; the user input module remains unchanged; The user report module generates a user report through the report generation engine when the AMH value and the value SH of the six sex hormones are detected after sampling; the report generation engine derives the overview, interpretation and disease type of the user report based on the parameters a, amh and sh according to the R=fn(a,amh,sh,f,m) paradigm.

3. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: This system can also evaluate endocrine abnormalities, irregular menstruation, and provide health guidance for pregnancy preparation by checking six sex hormone indicators: FSH, LH, E2, P, PRL, and T.

4. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: It also includes a health management module; that is, the health management program included in the six evaluation reports of ovarian health, fertility reserve capacity, and sex hormones generates health management content for users based on the disease types derived from the medical content derivation paradigms R=fn(a,amh,f,m), R=fn(a,amh,f,m), and R=fn(a,amh,sh,f,m) of the three products, combined with the user's age and BMI value; The health management module includes a health management knowledge base.

5. The artificial intelligence-based female endocrine test report interpretation system according to claim 4, characterized in that: The health management module also includes the health management paradigm; Normal form: H = fn(a,c,b); H: Health management; a: parameter age; c: parameter disease category; b: parameter bmi.

6. The artificial intelligence-based female endocrine test report interpretation system according to claim 4, characterized in that: The health management module also includes a derivation process: based on the disease type (c), the user's basic information: age (a), bmi (b), the user's health management plan is derived according to the H = fn (a, c, b) paradigm; Based on the medical content overview, interpretation, expert advice and health management plan, a complete report is ultimately generated for the user.

7. The artificial intelligence-based female endocrine test report interpretation system according to claim 4, characterized in that: In the health management module, push notification and health reminder functions are introduced. Through the user's basic information, test indicators, and user information collected through questionnaires, the questions are formatted and provided to the general large model as context to generate personalized health management guidance, including: diet plan, nutrient plan, health check-up, and lifestyle recommendations.

8. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: Establish a user feedback mechanism in the system to collect users' opinions and suggestions on the system.

9. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: Train a binary classification prediction model to quickly and accurately determine abnormal values ​​of indicators.

10. The artificial intelligence-based female endocrine test report interpretation system according to claim 1, characterized in that: Possible diseases are predicted based on historical test results and user information through a multi-classification decision tree prediction model.

Images