Coronary heart disease dynamic risk early warning system based on traditional Chinese medicine urine diagnosis and application of coronary heart disease dynamic risk early warning system

Abstract

According to the coronary heart disease dynamic risk early warning system based on Chinese traditional medicine urine diagnosis and the application thereof, the input and output modules are integrated, key data such as urine color values and hematocrit can be received in real time, the risk probability that a current individual suffers from the coronary heart disease is accurately calculated by applying an advanced algorithm, and the risk of the coronary heart disease is accurately detected through a non-invasive detection means. Before symptoms appear, the high risk of the coronary heart disease is accurately captured. A user does not need to install software, can easily access a web version system through a browser, and can conveniently use without a medical background. The system is suitable for prevention and monitoring of healthy people. And performing real-time evaluation and long-term tracking on coronary heart disease risk crowds and patients. Through clinical verification, the system keeps high prediction precision and practicability in different situations, and shows wide application prospects and reliability.

Description

Technical Field

[0001] The present invention relates to the fields of medicine, computer programming technology and bioinformatics, and in particular to a dynamic risk early warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application. Background Art

[0002] The incidence and prevalence of coronary heart disease continue to rise, and the age group of patients is gradually getting younger. The number of deaths from coronary heart disease in China ranks second in the world, causing 2 million deaths each year. People with undiagnosed coronary heart disease or asymptomatic people are at a very high risk of sudden cardiac death, accounting for 45% of the total number of sudden deaths. Therefore, accurately identifying and intervening in people with potential high risk of coronary heart disease before symptoms appear is of vital importance to reducing the incidence and mortality of coronary heart disease.

[0003] Currently, coronary angiography is the most commonly used diagnostic standard for coronary heart disease. Although it has high accuracy, it is invasive, costly, and difficult to implement. More importantly, this technology cannot warn of the high risk of coronary heart disease before symptoms appear, thus missing the golden opportunity for early intervention. Therefore, developing a simpler, more efficient, and non-invasive coronary heart disease early warning system is an urgent problem to be solved.

[0004] As the end product of human metabolism, the formation and composition of urine are closely linked to the blood circulation system, especially closely related to heart function. The color, composition and concentration of urine can all be used as an indirect reflection of heart health. When the heart is damaged or there is a potential risk, changes in its physiological function will directly affect blood circulation, which will then be reflected in urine. We innovatively proposed a dynamic risk warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application, aiming to provide patients and medical professionals with timely and accurate warning information before the symptoms of coronary heart disease appear, so as to achieve early intervention and reduce morbidity and mortality. Summary of the invention

[0005] The purpose of the present invention is to provide a dynamic risk warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application. The system can receive and analyze key physiological indicators such as urine color value and hematocrit by integrating input modules and output modules, and use advanced algorithms to calculate the risk probability of an individual suffering from coronary heart disease, so as to achieve accurate and real-time assessment of the risk of coronary heart disease. The system intuitively displays the dynamic risk value through a web-based nomogram to help users clearly understand the current coronary heart disease risk level and corresponding medical advice. The system is not only for preventive monitoring of healthy people, but also suitable for real-time assessment and long-term tracking of people and patients at risk of coronary heart disease, aiming to help people more comprehensively understand their own cardiovascular health status, so as to take scientific preventive measures and effectively reduce the risk of coronary heart disease.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a dynamic risk early warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application. The system consists of an input module and an output module.

[0008] The input module is used to input the independent warning indicators of the object to be predicted, and the independent warning indicators including urine color L value and hematocrit HCT are screened out according to the minimum absolute shrinkage and selection operator LASSO regression algorithm.

[0009] The output module uses the modeling algorithm and the discriminant function algorithm to process and analyze the indicators received by the input module and calculate the risk probability of the current individual suffering from coronary heart disease. The output module has a built-in dynamic nomogram, and the calculation process is as follows:

[0010] Coronary heart disease risk probability value Y = -1.13e-06 × X 3 +0.000368655×X 2 -0.029226938×X+0.77222565

[0011] X=3.077795449×L-2.5×HCT+39.199363851

[0012] The urine color L value indicates the brightness change of the black and white urine color, and the hematocrit HCT is a blood cell analysis indicator.

[0013] The output module is a web-based system. By entering the value of the indicator on the web page, the system will automatically calculate and output the probability of coronary heart disease risk.

[0014] The application of the present invention comprises the following steps:

[0015] Prepare the receiver operating characteristic (ROC) curve;

[0016] Prepare the calibration curve of the early warning system;

[0016] Prepare clinical decision curve analysis (DCA) curves for early warning systems;

[0017] Prepare the Bootstrap validation curve of the prediction model.

[0018] Compared with the prior art, the advantages and positive effects of the present invention are:

[0019] The present invention adopts non-invasive detection methods such as urine color value and hematocrit, which are simple to operate and have no special technical requirements for the operator. In addition, urine can be produced naturally every day and will not cause any harm to the body. It can achieve accurate assessment of coronary heart disease risk before the symptoms of coronary heart disease appear, avoiding the traumatic nature and high cost of traditional detection methods.

[0020] The present invention innovatively uses urine color as a core warning indicator for coronary heart disease, while incorporating other coronary heart disease risk factors to establish a urine-based multidimensional coronary heart disease warning system, which has not been reported before at home or abroad.

[0021] The present invention adopts an advanced algorithm and provides a coronary heart disease risk warning system that comprehensively considers multiple independent indicators, screens out the warning indicators most relevant to the risk of coronary heart disease, and assigns different weight coefficients, which can more comprehensively reflect the individual's coronary heart disease risk status, thereby improving the accuracy of risk assessment.

[0022] The present invention intuitively displays the calculated risk to the user through a web version system. The user can clearly understand the current coronary heart disease risk probability through simple operations, so as to take corresponding preventive measures. This web version system allows users to query and predict anytime and anywhere, greatly improving the convenience and practicality of use.

[0023] The present invention is not only suitable for the preventive monitoring of coronary heart disease in healthy people, but also for the real-time assessment and long-term tracking of people at risk of coronary heart disease and patients. Through regular testing and analysis, it helps to take preventive measures in advance and reduce the incidence of coronary heart disease and its complications. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 : Coronary heart disease risk warning nomogram.

[0025] Figure 2 : Receiver operating characteristic (ROC) curve of the training set of the embodiment of the present invention.

[0026] Figure 3 : Calibration curve of the training set according to the embodiment of the present invention.

[0027] Figure 4 : The DCA curve of clinical decision making of the training set in the embodiment of the present invention.

[0028] Figure 5 : Receiver operating characteristic (ROC) curve of the validation set of the embodiment of the present invention.

[0029] Figure 6 : Calibration curve of the verification set of the embodiment of the present invention.

[0030] Figure 7: DCA curve of clinical decision making in the validation set of the embodiment of the present invention.

[0031] Figure 8 : Web version of dynamic risk warning system for coronary heart disease. DETAILED DESCRIPTION

[0027] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. The following embodiments will further explain the relevant technical contents of the present invention in detail, but the disclosed contents are not intended to limit the scope of protection of the present invention. Any product identical or similar to the present invention obtained by anyone under the enlightenment of the present invention or by combining the features of the present invention with other prior arts falls within the scope of protection of the present invention.

[0028] Example 1

[0029] In this example, the main inclusion criteria for the research subjects are as follows: ① Age ≥ 40 years old; ② Meet the diagnostic criteria for coronary heart disease and be in the stable stage of coronary heart disease. The main exclusion criteria are as follows: ① Patients with severe electrolyte disorders, malignant tumors and other serious diseases; ② Pregnant or lactating women; ③ Patients with mental illness or a history of mental illness; ④ Patients with acute or chronic urinary tract infections; ⑤ Patients with infectious diseases.

[0030] According to the above inclusion and exclusion criteria, 500 research subjects were recruited in the China-Japan Friendship Hospital from 2023 to 2024, and all clinical and laboratory indicators were collected. After excluding 241 subjects with missing information on research indicators (general demographic characteristics, lifestyle, laboratory test indicators, urine color space values, and comorbidities), the data of 259 research subjects were finally included for analysis.

[0031] 70% of the 259 subjects, a total of 181 cases, were selected as the training set, and the LASSO regression algorithm was used to screen independent early warning indicators. The remaining 30% of the subjects, a total of 78 cases, were selected as the validation set.

[0032] The baseline characteristics of the 259 subjects are shown in Table 1.

[0033] Table 1 Baseline characteristics of the training set and validation set

[0034] R software (version 4.2.1) was used for statistical analysis. The least absolute shrinkage and selection operator LASSO regression algorithm were used to screen independent early warning indicators of CHD. The lambda value was determined by 10-fold cross validation and error.

[0035] The independent early warning indicators urine color L value and hematocrit HCT were screened out, and multi-factor Logistic regression fitting was performed to obtain the fitted coronary heart disease Logistic early warning regression equation

[0036] The logistic regression equation shows that urine color L value is a risk factor for coronary heart disease, that is, the higher the urine color L value, the more likely it is to develop coronary heart disease. For every unit increase in urine color L value, the risk of coronary heart disease increases by nearly 1.15 times on average. Hematocrit is a protective factor for coronary heart disease. The lower the hematocrit, the more likely it is to develop coronary heart disease. For every unit decrease in hematocrit, the risk of coronary heart disease decreases by 11%.

[0037] Visualizing the coronary heart disease logistic equation into a nomogram makes the prediction results more intuitive and easy to understand. Figure 1 The nomogram for early warning of coronary heart disease risk shown in the figure has all early warning indicators on the left and the range of values ​​and individual scores of the indicators on the right. All individual scores are added together to get the total score. By comparing the position of the total score on the predicted probability scale, the individual's coronary heart disease risk probability can be determined. 0.5 is the threshold probability. When the risk probability is greater than or equal to 0.5, it indicates a high risk state for the onset of coronary heart disease. The higher the risk probability, the greater the possibility of coronary heart disease.

[0038] Based on the nomogram, the receiver operating characteristic (ROC) curve is drawn to measure the discrimination of the warning effect, that is, the ability of the warning indicator to distinguish between individuals at high and low risk of coronary heart disease. The ROC curve is drawn with sensitivity as the vertical axis and 1-specificity as the horizontal axis. Sensitivity represents the proportion of samples predicted to be positive and actually positive to all positive samples, reflecting the model's ability to identify positive samples; while 1-specificity represents the proportion of samples predicted to be positive but actually negative to all negative samples, reflecting the degree of misjudgment of the model for negative samples. The closer the ROC curve is to the upper left corner, the better the discrimination. In addition, the larger the area under the curve (AUC) value, the better the prediction performance. By calculating sensitivity-(1-specificity), the Youden index can be obtained. The point corresponding to the maximum value of the Youden index is the optimal critical point, which can be used to judge high-risk populations.

[0039] like Figure 2 As shown, in the ROC curve of the training set of the embodiment of the present invention, the sensitivity and specificity of the combined curve of urine color L value and hematocrit are 0.702 and 0.625 respectively, and the area under the curve AUC is 0.874 (P=0.000, 95% CI: 0.804-0.945), indicating that it has good warning value for coronary heart disease. Furthermore, the urine color L value is used as an independent warning indicator for coronary heart disease, and its optimal critical point is 52.5, indicating that if the urine color L value is greater than or equal to 52.5, the individual can be diagnosed as or can be diagnosed as coronary heart disease and is considered a high-risk population.

[0040] Based on the nomogram, the calibration curve is drawn to measure the calibration degree of the early warning effect, that is, the consistency between the risk probability predicted by the early warning indicator and the actual risk of coronary heart disease. The horizontal axis of the curve represents the predicted risk probability of coronary heart disease, and the vertical axis represents the actual risk. The diagonal line of y=x represents the ideal situation where the predicted probability is completely consistent with the actual probability. If the curve is closely around the diagonal line, it means that the predicted probability is relatively consistent with the actual probability and the calibration effect is good.

[0041] like Figure 3 As shown, in the Calibration curve of the training set of the embodiment of the present invention, the curve fluctuates around the diagonal line, indicating that the consistency between the predicted risk probability and the actual observation result is high.

[0042] Based on the nomogram, the DCA curve for clinical decision-making is drawn to evaluate the clinical practicality of the warning effect. The horizontal axis of the DCA curve coordinate system represents the decision threshold, and the vertical axis represents the net benefit. Ideally, the DCA curve should be as close to the upper right corner of the coordinate system as possible, which means that a higher net benefit can still be obtained at a higher decision threshold. The horizontal line at 0 on the vertical axis is the None line, indicating a net benefit of 0. If the DCA curve is above the None line, it means that there is a positive net benefit at this decision threshold. The curve that changes with the decision threshold is the All line, which represents the net benefit that can be obtained if all people are warned of coronary heart disease. If the DCA curve is above the All line, it means that the decision threshold has a higher net benefit than all warnings, that is, it can more accurately identify the risk population that needs warnings.

[0043] like Figure 4 As shown, the clinical decision DCA curve of the training set of the embodiment of the present invention is above the None line and the All line, indicating that it can provide higher clinical benefits at different decision thresholds. And as the decision threshold changes, the net benefit of the curve remains at a high level, which reflects the stability and reliability of the early warning effect in different clinical scenarios.

[0044] The validation set was used to further evaluate the performance and generalization ability of the present invention in early warning of coronary heart disease. The validation set used an independent data set and did not participate in the training process to ensure the fairness and objectivity of the evaluation results. The ROC curve, calibration curve and DCA curve were constructed by the Bootstrap self-service sampling method.

[0045] Bring the validation set data into the constructed nomogram, and the ROC curve is as follows Figure 5As shown, the area under the curve (AUC) was 0.938 (P=0.000, 95% CI: 0.898-0.978), the sensitivity was 0.922, and the specificity was 0.873, which was also able to maintain a high degree of prediction accuracy and practicality on the validation set.

[0046] The calibration curve of the validation set is as follows Figure 6 As shown, the curve fits the diagonal line closely. In each prediction probability interval, the predicted positive case ratio is very close to the actually observed positive case ratio, without any obvious deviation or trend away from the diagonal line, indicating that there is a high degree of consistency between the predicted probability and the actual observation results.

[0047] The DCA curve of the validation set is as follows Figure 7 As shown, the net benefits at different thresholds are higher than the None line of the no-warning strategy and the All line of the comprehensive warning strategy. As the decision threshold increases from 0 to 1, the change trend of the net benefit is stable, and it is significantly better than the random guessing line at multiple threshold points, indicating that the present invention has high accuracy in identifying patients who need warnings, has practical application value in clinical decision-making, and can increase benefits while avoiding excessive intervention.

[0048] In order to further improve the convenience and accessibility of coronary heart disease risk warning, this paper designs and implements a web-based coronary heart disease risk dynamic warning system based on the preparation and performance evaluation of the nomogram. The system aims to provide users with an intuitive, efficient and easy-to-operate coronary heart disease risk assessment tool. Web-based coronary heart disease dynamic risk warning system: https: / / ruitin.shinyapps.io / dynnomapp / Figure 8 As shown in the figure, by inputting the corresponding numerical value in the input module, the corresponding coronary heart disease risk probability value can be automatically output and displayed intuitively in digital form on the output module interface. There is no need to manually calculate the score, which greatly simplifies the complex calculation steps in the traditional risk assessment process. Even if the user does not have a professional medical background, they can easily understand and apply the system.

[0049] In summary, the present invention provides a dynamic risk early warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application. The system realizes accurate assessment of the risk of coronary heart disease before the onset of symptoms through non-invasive detection means. The system can receive the data input by the user in real time and return the calculation results to realize rapid early warning. There is no need to download or install any software, and the system can be accessed and used through a browser. The user interface is concise and clear, and the operation steps are simple and easy to understand, and it can be easily used even without a professional medical background. In practical applications, this system has demonstrated good early warning performance and clinical practicality. This system not only achieved satisfactory prediction accuracy on the training set, but also maintained a high degree of stability and reliability on the validation set, indicating that this system has high early warning accuracy and clinical practicality.

[0050] The above are exemplary embodiments, and the exemplary embodiments are selected for description in order to explain the specific principles and practical applications of the present invention, so that relevant personnel can understand and use them easily. Among them, the methods mentioned in the embodiments are not completely limited to such methods, and relevant personnel can make corresponding adjustments according to the principles in actual operations. The scope of protection of the present invention is intended to be defined by the claims and their equivalents.

Claims

1. A dynamic risk early warning system for coronary heart disease based on urine diagnosis in traditional Chinese medicine and its application, characterized in that: The system consists of input modules and output modules: An input module is used to receive the urine color L value and hematocrit HCT of the object to be predicted as independent warning indicators; The output module is used to process the indicators received by the input module using an algorithm, calculate the risk probability of the current individual suffering from coronary heart disease, and intuitively display the risk probability through a web version system.

2. The coronary heart disease dynamic risk early warning system according to claim 1, characterized in that: The urine color L value represents the brightness change of the black and white urine color, and the hematocrit HCT is a blood cell analysis indicator. Both of them serve as key indicators for early warning of coronary heart disease risk.

3. The coronary heart disease dynamic risk early warning system according to claim 1, characterized in that: The independent early warning indicators in the input module are screened from multiple candidate indicators through the minimum absolute shrinkage and selection operator LASSO regression algorithm.

4. The coronary heart disease dynamic risk early warning system according to claim 1, characterized in that: The modeling algorithm in the output module uses the following formula to calculate the coronary heart disease risk probability value Y: Y=-1.13e-06×X 3 +0.000368655×X 2 -0.029226938×X+0.77222565, where X=3.077795449×L-2.5×HCT+39.199363851.

5. The coronary heart disease dynamic risk early warning system according to claim 1, characterized in that: The application of the system comprises the following steps: The receiver operating characteristic (ROC) curve was prepared to measure the discrimination of the warning effect, that is, the ability of the warning index to distinguish between individuals with high and low risk of CHD; Prepare the calibration curve of the early warning system to measure the calibration degree of the early warning effect, that is, the consistency between the risk probability predicted by the early warning indicator and the actual risk of coronary heart disease; Prepare the clinical decision DCA curve of the early warning system to evaluate the clinical practicality of the early warning effect; A Bootstrap self-sampling validation curve of the prediction model was prepared, and the validation set was used to further evaluate the performance and generalization ability of the present invention in early warning of coronary heart disease.

6. The coronary heart disease dynamic risk early warning system according to any one of claims 1 to 5, characterized in that: The system is particularly suitable for real-time assessment and long-term tracking of people at risk of coronary heart disease and patients.

7. The coronary heart disease dynamic risk early warning system according to any one of claims 1 to 5, characterized in that: The system can also be used for preventive monitoring of healthy people.

8. The application method of the coronary heart disease dynamic risk early warning system according to claim 1, characterized in that: This method is implemented through a web-based system. Users do not need to download or install any software. They can access and use the system through a browser. By inputting the value of the indicator, the system will automatically calculate and output the probability of coronary heart disease risk.

Images