Coronary heart disease risk prediction method based on sleep apnea syndrome

Abstract

The present invention relates to the field of coronary heart disease risk prediction, and specifically to a method for predicting the risk of coronary heart disease based on sleep apnea syndrome. The method first obtains signal data of different signal types and index data of different body index types of each patient during each monitoring process, and obtains the age of each patient, analyzes the difference in signal data of the same signal type, the difference in index data of the same body index type, and the difference in age between each patient, constructs a spatial scatter plot of the target monitoring process, analyzes the density of patient distribution in the neighborhood of the coronary heart disease patient in the spatial scatter plot, and simultaneously combines the distance between the patient to be tested and the coronary heart disease patient to obtain the disease possibility of the patient to be tested, and based on the disease possibility, predicts the risk of the patient to be tested suffering from coronary heart disease. The present invention can accurately predict the risk of coronary heart disease in patients with sleep apnea syndrome.

Description

Technical Field

[0001] The present invention relates to the field of coronary heart disease risk prediction, and in particular to a coronary heart disease risk prediction method based on sleep apnea syndrome. Background Art

[0002] Sleep apnea syndrome is a common sleep disorder characterized by repeated partial or complete obstruction of the upper airway during sleep, leading to apnea or hypoventilation symptoms, which in turn lead to intermittent hypoxemia and hypercapnia. This repeated hypoxia can cause overactivation of the sympathetic nervous system, leading to blood pressure fluctuations and increased heart rate, which may damage vascular endothelial function and induce coronary heart disease. Therefore, early detection and treatment of sleep apnea syndrome is of great significance for the prevention and management of coronary heart disease.

[0003] In related technologies, the idea of ​​supervised learning algorithms such as the K-nearest neighbor algorithm is usually used, and a variety of relevant data of the sleep apnea syndrome patient group are combined to predict the patient's risk of coronary heart disease. However, in the risk prediction process, there are many types of data related to coronary heart disease, and if the K-nearest neighbor algorithm is used directly for prediction, unreasonable parameter selection and construction of the model will affect the accuracy of the prediction results, resulting in the inability to accurately predict the coronary heart disease risk of patients with sleep apnea syndrome through existing methods. Summary of the invention

[0004] In order to solve the technical problem that the existing methods cannot accurately predict the risk of coronary heart disease in patients with sleep apnea syndrome, the purpose of the present invention is to provide a method for predicting the risk of coronary heart disease based on sleep apnea syndrome. The technical solution adopted is as follows:

[0005] The present invention proposes a method for predicting the risk of coronary heart disease based on sleep apnea syndrome, the method comprising:

[0006] Acquire signal data of different signal types and index data of different body index types of each patient during each monitoring process, and acquire the age of each patient, the patients including patients with coronary heart disease and patients to be tested;

[0007] Take any monitoring process as a target monitoring process, take the youngest patient as a reference patient, take any patient as a target patient, and in the target monitoring process, obtain the first relative characteristic value, the second relative characteristic value, and the third relative characteristic value of the target patient in the target monitoring process according to the difference in the signal data of the same signal type, the difference in the indicator data of the same body indicator type, and the difference in age between the target patient and the reference patient; in the target monitoring process, map each patient to a spatial coordinate system based on the first relative characteristic value, the second relative characteristic value, and the third relative characteristic value of each patient to obtain a spatial scatter plot of the target monitoring process;

[0008] In the spatial scatter plot, any one of the patients with coronary heart disease is taken as the target patient with coronary heart disease, and the local neighborhoods of different radii of the target patient with coronary heart disease are constructed with the target patient with coronary heart disease as the center of the sphere; multiple neighboring patients of the target patient with coronary heart disease are obtained according to the number of patients in each local neighborhood and the radius of the local neighborhood; the confidence level of the target patient with coronary heart disease is obtained according to the distance between the target patient with coronary heart disease and each neighboring patient, and the number of patients with coronary heart disease among all the neighboring patients of the target patient with coronary heart disease; any one of the patients to be tested is taken as the target patient to be tested, and the disease coefficient of the target patient to be tested in the target monitoring process is obtained according to the distance between the target patient to be tested and each patient with coronary heart disease, and the confidence level of each patient with coronary heart disease; the disease possibility of the target patient to be tested is obtained according to the change of the disease coefficient of the target patient to be tested in each monitoring process;

[0009] Based on the disease possibility, the risk of the target patient suffering from coronary heart disease is predicted.

[0010] Further, obtaining the first relative characteristic value, the second relative characteristic value, and the third relative characteristic value of the target patient in the target monitoring process includes:

[0011] Taking any signal type as a target signal type, performing Fourier transform processing on the signal data of the target signal type of the target patient and the signal data of the target signal type of the reference patient, respectively, to obtain a frequency spectrum diagram of the target patient with respect to the target signal type and a frequency spectrum diagram of the reference patient with respect to the target signal type;

[0012] Taking the difference in amplitude values ​​of the same frequency value between the target patient's spectrum graph regarding the target signal type and the reference patient's spectrum graph regarding the target signal type as the relative amplitude difference value of each frequency value, and taking the average of the relative amplitude difference values ​​of all frequency values ​​as the relative signal difference value of the target patient regarding the target signal type;

[0013] Taking any one body indicator type as a target body indicator type, and taking the difference between the indicator data of the target body indicator type of the target patient and the indicator data of the target body indicator type of the reference patient as the relative indicator difference value of the target patient with respect to the target body indicator type;

[0014] Taking the set consisting of the relative signal difference values ​​of all signal types and the relative index difference values ​​of all body index types of the target patient as a reference set of the target patient;

[0015] Taking the average value of all elements in the reference set as the first relative characteristic value of the target patient, and taking the average value of the absolute values ​​of all elements in the reference set as the second relative characteristic value of the target patient;

[0016] The difference between the age of the target patient and the age of the reference patient is used as the third relative characteristic value of the target patient.

[0017] Furthermore, obtaining a spatial scatter plot of the target monitoring process includes:

[0018] Constructing a spatial coordinate system, wherein the horizontal axis of the spatial coordinate system is the third relative eigenvalue, the vertical axis of the spatial coordinate system is the first relative eigenvalue, and the vertical axis of the spatial coordinate system is the second relative eigenvalue;

[0019] The three-dimensional data points formed by the third relative eigenvalue, the first relative eigenvalue and the second relative eigenvalue of each patient in the target monitoring process are mapped to the spatial coordinate system to obtain a spatial scatter plot of the target monitoring process, wherein each three-dimensional data point represents a patient.

[0020] Furthermore, taking the target coronary heart disease patient as the sphere center, constructing local neighborhoods of different radii of the target coronary heart disease patient includes:

[0021] The patient closest to the target coronary heart disease patient is used as the nearest patient of the target coronary heart disease patient, and the Euclidean distance between the target coronary heart disease patient and the nearest patient of the target coronary heart disease patient is used as the initial radius of the target coronary heart disease patient;

[0022] Taking the target coronary heart disease patient as the sphere center and the spherical area with the initial radius as the radius, as the initial local neighborhood of the target coronary heart disease patient;

[0023] The radius of the initial local neighborhood is expanded a preset first number of times with a preset expansion step length to obtain local neighborhoods of different radii for the target coronary heart disease patient.

[0024] Furthermore, the step of obtaining a plurality of neighboring patients of the target coronary heart disease patient includes:

[0025] The number of patients in each local neighborhood is used as the numerator, the cube of the radius of each local neighborhood is used as the denominator, the ratio is used as the characteristic coefficient of each local neighborhood, the variance of the characteristic coefficients of all local neighborhoods is negatively correlated and mapped, and the local distribution consistency of the target coronary heart disease patients is obtained;

[0026] The number of patients in the initial local neighborhood of the target coronary heart disease patient is used as the numerator, the initial radius of the target coronary heart disease patient is used as the denominator, and the ratio is used as the local density of the target coronary heart disease patient;

[0027] The local distribution consistency and the local density are integrated and normalized to obtain a weight parameter of a target coronary heart disease patient;

[0028] Rounding up the product of the weight parameter and the preset second number to obtain a reference number of target coronary heart disease patients;

[0029] The reference number of patients closest to the target coronary heart disease patient are regarded as the neighboring patients of the target coronary heart disease patient.

[0030] Furthermore, obtaining the confidence level of the target coronary heart disease patient includes:

[0031] Performing negative correlation mapping on the average values ​​of the Euclidean distances between the target coronary heart disease patient and all the neighboring patients to obtain a local density value of the target coronary heart disease patient;

[0032] The number of patients with coronary heart disease among all the neighboring patients of the target patient with coronary heart disease is used as the numerator, the local density value is used as the denominator, and the ratio is used as the confidence level of the target patient with coronary heart disease.

[0033] Furthermore, the disease coefficient of the target patient to be tested in the target monitoring process includes:

[0034] Negative correlation mapping is performed on the Euclidean distance between the target patient to be tested and each coronary heart disease patient to obtain the initial proximity between the target patient to be tested and each coronary heart disease patient;

[0035] The product value of the proximity degree between the target patient to be tested and each coronary heart disease patient and the confidence degree of each coronary heart disease patient is used as the adjusted proximity degree between the target patient to be tested and each coronary heart disease patient;

[0036] The average value of the adjusted proximity between the target patient to be tested and all patients with coronary heart disease is used as the disease coefficient of the target patient to be tested in the target monitoring process.

[0037] Furthermore, obtaining the disease probability of the target patient to be tested includes:

[0038] Obtaining a first possibility of the target patient to be tested according to a difference in the disease coefficient of the target patient to be tested between each monitoring process and the next adjacent monitoring process;

[0039] The average of the disease coefficients of the target patient to be tested during all monitoring processes is used as the second possibility of the target patient to be tested;

[0040] The first possibility and the second possibility are integrated and normalized to obtain the disease possibility of the target patient to be tested.

[0041] Furthermore, the first possibility of obtaining the target patient to be tested includes:

[0042] The disease coefficient of the target patient to be tested in the next monitoring process adjacent to each monitoring process is used as the numerator, the disease coefficient of the target patient to be tested in each monitoring process is used as the denominator, and the ratio is used as the change trend parameter of the target patient to be tested in each monitoring process;

[0043] The average value of the change trend parameter of the target patient to be tested during all monitoring processes is used as the first possibility of the target patient to be tested.

[0044] Furthermore, the prediction of the risk of the target patient suffering from coronary heart disease includes:

[0045] If the disease probability of the target patient to be tested is greater than a preset probability threshold, the target patient to be tested is at risk of suffering from coronary heart disease, and an early warning is issued to the target patient to be tested.

[0046] The present invention has the following beneficial effects:

[0047] The present invention takes into account that the existing methods cannot accurately predict the risk of coronary heart disease in patients with sleep apnea syndrome. Therefore, the signal data of different signal types and different indicator data of each patient in each monitoring process are first obtained, and the age of each patient is obtained at the same time. The patients include patients with coronary heart disease and patients to be tested. Since the signal data, indicator data and age of each patient are different, the degree of risk of coronary heart disease in different patients is different, and there are also differences in the signal data and indicator data in different monitoring processes. Therefore, any monitoring process is first analyzed, and the youngest patient is taken as the reference patient. The first relative eigenvalue, the second relative eigenvalue and the third relative eigenvalue obtained reflect the approximation of various data of the target patient relative to the reference patient, and then a spatial scatter plot of the target monitoring process is constructed for subsequent detection of coronary heart disease in the patient to be tested. In order to accurately predict the risk, it is considered that the closer other patients around the target coronary heart disease patient are to the target coronary heart disease patient, the higher the data similarity between the other patients around and the target coronary heart disease patient. At the same time, considering that the number distribution of other patients in different neighborhoods around the target coronary heart disease patient may be inconsistent, in order to ensure the accuracy of subsequent analysis, multiple neighboring patients of the target coronary heart disease patient are first obtained, and the obtained confidence level reflects the reference value of the target coronary heart disease patient in the subsequent assessment of the risk of coronary heart disease in the patient to be tested, and then the morbidity coefficient is used to reflect the possibility of the target patient to be tested suffering from coronary heart disease during the target monitoring process. Considering that the morbidity coefficient of the target patient to be tested shows an upward trend in each monitoring process, it means that the risk of coronary heart disease is greater. Therefore, the risk of coronary heart disease in the target patient to be tested can be further accurately predicted through the possibility of illness. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A flow chart of a method for predicting the risk of coronary heart disease based on sleep apnea syndrome provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a method for predicting the risk of coronary heart disease based on sleep apnea syndrome proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0051] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0052] The specific scheme of the method for predicting the risk of coronary heart disease based on sleep apnea syndrome provided by the present invention is described in detail below with reference to the accompanying drawings.

[0053] See also Figure 1 , which shows a flow chart of a method for predicting the risk of coronary heart disease based on sleep apnea syndrome provided by an embodiment of the present invention, the method comprising:

[0054] Step S1: acquiring signal data of different signal types and index data of different body index types of each patient during each monitoring process, and acquiring the age of each patient, including patients with coronary heart disease and patients to be tested.

[0055] There is a close connection between sleep apnea syndrome and coronary heart disease. The main mechanism of sleep apnea syndrome is intermittent hypoxemia and hypercapnia due to repeated apnea and hypoventilation during sleep. This repeated hypoxia can cause overactivation of the sympathetic nervous system, leading to blood pressure fluctuations and increased heart rate, which may damage organs such as blood vessels and the heart, and promote the induction of cardiovascular and cerebrovascular diseases. Therefore, sleep apnea syndrome is considered to be one of the important risk factors for coronary heart disease.

[0056] The embodiment of the present invention first uses relevant medical equipment in the hospital to collect signal data of different signal types and index data of different body index types for each sleep apnea syndrome patient during each monitoring process, and obtains the age of each patient at the same time. The patients include patients with coronary heart disease and patients to be tested, wherein the signal data of different signal types include, for example, breathing sound signals and electrocardiogram signals, etc. The breathing sound signal data can be collected and obtained using medical equipment such as electronic stethoscopes, and the electrocardiogram signal data can be collected and obtained using medical equipment such as electrocardiographs. The index data of different body index types include, for example, the levels of various hormones such as glucocorticoids, estrogen, adrenaline, and thyroid hormones, as well as the patient's BMI data, etc. These data can be collected and obtained using relevant medical equipment or methods. At the same time, age is also an important factor related to coronary heart disease, so the embodiment of the present invention also needs to record the age of each patient.

[0057] It should be noted that among all patients, including patients with coronary heart disease and patients to be tested, patients with coronary heart disease are patients known to have coronary heart disease, and it is unknown whether the patients to be tested have coronary heart disease. Subsequently, a prediction model can be constructed by combining the signal data, indicator data and age of the patient group, and the data similarity between the patients to be tested and the patients with coronary heart disease can be analyzed to achieve accurate prediction of the coronary heart disease risk of the patients to be tested.

[0058] Step S2: Take any monitoring process as the target monitoring process, take the youngest patient as the reference patient, and take any patient as the target patient. In the target monitoring process, according to the difference in signal data of the same signal type between the target patient and the reference patient, the difference in indicator data of the same body indicator type, and the difference in age, obtain the first relative eigenvalue, the second relative eigenvalue, and the third relative eigenvalue of the target patient in the target monitoring process; in the target monitoring process, based on the first relative eigenvalue, the second relative eigenvalue, and the third relative eigenvalue of each patient, map each patient to the spatial coordinate system to obtain a spatial scatter plot of the target monitoring process.

[0059] Since the signal data, index data and age of each patient are different, the degree of coronary heart disease risk of different patients is different, and there are also differences in the signal data and index data in different monitoring processes. Therefore, firstly, any monitoring process is analyzed, and any monitoring process is taken as the target monitoring process. At the same time, the patient with the youngest age is taken as the reference patient. Subsequently, various data of other patients can be compared and analyzed with various data of the reference patient to establish the difference characteristics between each patient. Then, any patient is analyzed, and any patient is taken as the target patient. When the difference of the same type of data between the target patient and the reference patient is smaller, it means that the physical state between the target patient and the reference patient is more similar. Therefore, the difference of signal data of the same signal type, the difference of index data of the same physical index type, and the difference of age between the target patient and the reference patient can be analyzed. The first relative characteristic value, the second relative characteristic value and the third relative characteristic value obtained reflect the approximation of various data of the target patient relative to the reference patient in the target monitoring process. Then, the spatial scatter plot of the target monitoring process can be constructed based on the first relative characteristic value, the second relative characteristic value and the third relative characteristic value of each patient, so as to achieve accurate prediction of the coronary heart disease risk of the patient to be tested.

[0060] Preferably, in one embodiment of the present invention, the method for obtaining the first relative characteristic value, the second relative characteristic value and the third relative characteristic value of the target patient during the target monitoring process specifically includes:

[0061] First, any signal type is taken as the target signal type, and Fourier transform is performed on the signal data of the target signal type of the target patient and the signal data of the target signal type of the reference patient, respectively, to obtain the spectrum diagram of the target signal type of the target patient and the spectrum diagram of the target signal type of the reference patient, where the horizontal axis of the spectrum diagram represents the frequency value, and the vertical axis represents the amplitude value, wherein Fourier transform is a technical means well known to those skilled in the art and will not be elaborated here.

[0062] The difference in amplitude values ​​of the same frequency value between the target patient's spectrum graph regarding the target signal type and the reference patient's spectrum graph regarding the target signal type is taken as the relative amplitude difference value of each frequency value. The farther the relative amplitude difference value is from the value 0, the greater the relative difference in amplitude values ​​of the same frequency value in the spectrum graph regarding the target signal type between the target patient and the reference patient. The average of the relative amplitude difference values ​​of all frequency values ​​is taken as the relative signal difference value of the target patient regarding the target signal type. The farther the relative signal difference value is from the value 0, the greater the relative difference in signal data between the target patient and the reference patient regarding the target signal type.

[0063] As an example, in one embodiment of the present invention, the expression of the relative signal difference value of the target patient with respect to the target signal type may be specifically, for example, as follows:

[0064] .

[0065] in, represents the relative signal difference value of the target patient with respect to the target signal type; The first The amplitude value of each frequency value; The first The amplitude value of each frequency value; Indicates The relative amplitude difference of the frequency values; Represents the number of frequency values ​​in the union of the frequency values ​​in the target patient's spectrum graph regarding the target signal type and the frequency values ​​in the reference patient's spectrum graph regarding the target signal type. For example, the frequency values ​​in the target patient's spectrum graph regarding the target signal type are 1, 2, 3, 4, and the frequency values ​​in the reference patient's spectrum graph regarding the target signal type are 2, 3, 4, 5, then the union of the two is 1, 2, 3, 4, 5. It should be noted that if the frequency values ​​in the target patient's or reference patient's spectrum graph regarding the target signal type do not include a certain frequency value in the union, then the amplitude value of the frequency value in the target patient's or reference patient's spectrum graph regarding the target signal type is considered to be 0. For example, if the target patient's spectrum graph regarding the target signal type does not include frequency value 5, then the amplitude value of frequency value 5 in the target patient's spectrum graph regarding the target signal type is 0.

[0066] The relative signal difference value of each signal type of the target patient can be obtained by the same method as above.

[0067] Then, any body indicator type is taken as the target body indicator type, and the difference between the indicator data of the target body indicator type of the target patient and the indicator data of the target body indicator type of the reference patient is taken as the relative indicator difference value of the target patient regarding the target body indicator type. The farther the relative indicator difference value is from the value 0, the greater the relative difference in the indicator data between the target patient and the reference patient regarding the target body indicator type.

[0068] As an example, in one embodiment of the present invention, the expression of the relative index difference value of the target patient with respect to the target body index type may be specifically, for example, as follows:

[0069] .

[0070] in, Relative index difference value representing the target patient with respect to the target body index type; Indicator data representing a target physical indicator type of a target patient; Indicator data representing a target physical indicator type for a reference patient.

[0071] The relative index difference value of each signal type for the target patient can be obtained by the same method as above.

[0072] The further away from 0 the relative signal difference value of each signal type and the relative index difference value of each signal type of the target patient are, the greater the difference in physical condition between the target patient and the reference patient. Therefore, the set consisting of the relative signal difference values ​​of all signal types and the relative index difference values ​​of all physical index types of the target patient can be used as the reference set of the target patient, and the average value of all elements in the reference set can be used as the first relative characteristic value of the target patient, and the average value of the absolute values ​​of all elements in the reference set can be used as the second relative characteristic value of the target patient. If the first relative characteristic value and the second relative characteristic value between any two patients are close, it means that the signal data and the index data between the two patients are relatively close. At the same time, the difference between the age of the target patient and the age of the reference patient can be used as the third relative characteristic value of the target patient.

[0073] The first relative characteristic value, the second relative characteristic value and the third relative characteristic value of each patient in the target monitoring process can be obtained by the same method as above.

[0074] Then, based on the first relative eigenvalue, the second relative eigenvalue and the third relative eigenvalue of each patient in the target monitoring process, each patient is mapped to the spatial coordinate system to obtain a spatial scatter plot of the target monitoring process. Subsequently, the distance between each patient can be analyzed in the spatial scatter plot to achieve accurate prediction of the coronary heart disease risk of the patient to be tested.

[0075] Preferably, in one embodiment of the present invention, the method for acquiring the spatial scatter diagram of the target monitoring process specifically includes:

[0076] A spatial coordinate system is constructed, wherein the horizontal axis of the spatial coordinate system is the third relative eigenvalue, the vertical axis of the spatial coordinate system is the first relative eigenvalue, and the vertical axis of the spatial coordinate system is the second relative eigenvalue.

[0077] The three-dimensional data points formed by the third relative eigenvalue and the first relative eigenvalue and the second relative eigenvalue of each patient in the target monitoring process are mapped to the spatial coordinate system to obtain a spatial scatter plot of the target monitoring process, wherein each three-dimensional data point represents a patient, that is, each patient is represented as a three-dimensional data point in the spatial scatter plot.

[0078] At this point, a spatial scatter plot of the target monitoring process is constructed.

[0079] Step S3: In the spatial scatter plot, any coronary heart disease patient is taken as the target coronary heart disease patient, and the local neighborhoods of different radii of the target coronary heart disease patient are constructed with the target coronary heart disease patient as the center of the sphere; multiple neighboring patients of the target coronary heart disease patient are obtained according to the number of patients in each local neighborhood and the radius of the local neighborhood; the confidence level of the target coronary heart disease patient is obtained according to the distance between the target coronary heart disease patient and each neighboring patient, and the number of coronary heart disease patients among all neighboring patients of the target coronary heart disease patient; any patient to be tested is taken as the target patient to be tested, and the morbidity coefficient of the target patient to be tested during the target monitoring process is obtained according to the distance between the target patient to be tested and each coronary heart disease patient, and the confidence level of each coronary heart disease patient; the morbidity possibility of the target patient to be tested is obtained according to the change of the morbidity coefficient of the target patient to be tested during each monitoring process.

[0080] In the spatial scatter plot, the higher the local density of a certain coronary heart disease patient, that is, the closer the coronary heart disease patient is to the surrounding patients, the greater the possibility that other patients have coronary heart disease. At this time, the number of patients who may have coronary heart disease around the coronary heart disease patient is greater. Therefore, any coronary heart disease patient is first taken as the target coronary heart disease patient. At the same time, since coronary heart disease can also be caused by non-sleep apnea syndrome factors such as genetics, it has a certain degree of randomness. Therefore, in order to be able to predict the risk of coronary heart disease in the patient to be tested based on the characteristics of patients who already have coronary heart disease, so as to make the prediction of their coronary heart disease risk more accurate, first, with the target coronary heart disease patient as the center of the sphere, local neighborhoods of different radii of the target coronary heart disease patient are constructed. Subsequently, the number of patients in different local neighborhoods and the radius of the local neighborhoods can be analyzed, and the neighboring patients of the target coronary heart disease patient can be selected to improve the subsequent calculation accuracy of the possibility of the patient to be tested suffering from the disease.

[0081] Preferably, in one embodiment of the present invention, the method for acquiring local neighborhoods of different radii of the target coronary heart disease patient specifically includes:

[0082] The patient closest to the target coronary heart disease patient is taken as the nearest patient of the target coronary heart disease patient, and the Euclidean distance between the target coronary heart disease patient and the nearest patient of the target coronary heart disease patient is taken as the initial radius of the target coronary heart disease patient. It should be noted that, in this embodiment of the present invention, the distance between each patient refers to the Euclidean distance between each patient in the spatial scatter plot.

[0083] The spherical area with the target CHD patient as the sphere center and the initial radius as the radius is taken as the initial local neighborhood of the target CHD patient.

[0084] The radius of the initial local neighborhood is expanded for a preset first number of times with a preset expansion step length to obtain local neighborhoods of different radii of the target coronary heart disease patient, wherein, in one embodiment of the present invention, the preset expansion step length is set to the initial radius of the target coronary heart disease patient, and the preset first number is set to 5. At this time, each local neighborhood is respectively the target coronary heart disease patient as the sphere center, , , , , is the spherical area with radius The initial radius representing the initial local neighborhood, the specific values ​​of the preset expansion step and the preset first number can also be set by the implementer according to the specific implementation scenario and are not limited here.

[0085] Since the patients around the target coronary heart disease patient may be unevenly distributed, in order to ensure the accuracy of the subsequent local density analysis of the target coronary heart disease patient, it is also necessary to obtain multiple neighboring patients of the target coronary heart disease patient based on the number of patients in each local neighborhood and the radius of the local neighborhood. Subsequently, the local density of the target coronary heart disease patient can be accurately analyzed based on the selected neighboring patients.

[0086] Preferably, in one embodiment of the present invention, the method for acquiring multiple neighboring patients of a target coronary heart disease patient specifically includes:

[0087] First, the number of patients in each local neighborhood is taken as the numerator, the cube of the radius of each local neighborhood is taken as the denominator, and the ratio is taken as the characteristic coefficient of each local neighborhood. Since the local neighborhood is a spherical space, if the distribution of patients around the target coronary heart disease patient is relatively uniform, then when the local neighborhood of the target coronary heart disease patient gradually expands, the number of patients in each local neighborhood should also increase accordingly. Therefore, when the characteristic coefficients of each local neighborhood of the target coronary heart disease patient are relatively consistent, it means that the distribution of patients around the target coronary heart disease patient is relatively uniform. At this time, more patients can be selected around the target coronary heart disease patient for subsequent local density analysis. Conversely, if the distribution of patients around the target coronary heart disease patient is uneven, fewer patients should be selected around the target coronary heart disease patient to improve the accuracy of subsequent local density analysis. Therefore, the variance of the characteristic coefficients of all local neighborhoods can be negatively correlated to obtain the local distribution consistency of the target coronary heart disease patient. The greater the local distribution consistency, the more uniform the distribution of patients around the target coronary heart disease patient, and more patients can be selected around him for subsequent density analysis.

[0088] Then, the number of patients in the initial local neighborhood of the target coronary heart disease patient is taken as the numerator, the initial radius of the target coronary heart disease patient is taken as the denominator, and the ratio is taken as the local density of the target coronary heart disease patient. The greater the local density, the denser the distribution of patients around the target coronary heart disease patient, and more patients can be selected around him for subsequent density analysis.

[0089] Furthermore, the local distribution consistency and the local density are integrated and normalized to obtain the weight parameter of the target coronary heart disease patient, and the product of the weight parameter and the preset second number is rounded up to obtain the reference number of target coronary heart disease patients, and the reference number of patients closest to the target coronary heart disease patient are used as the neighboring patients of the target coronary heart disease patient, wherein the preset second number is set to 10, and the specific value of the preset second number can also be set by the implementer according to the specific implementation scenario, which is not limited here.

[0090] In the embodiment of the present invention, the local distribution consistency and the local density can be integrated by calculating the sum or product of the two, which is not limited here.

[0091] In one embodiment of the present invention, the normalization processing can be specifically, for example, maximum and minimum value normalization processing, and the normalization in subsequent steps can all adopt maximum and minimum value normalization processing. In other embodiments of the present invention, other normalization methods can be selected according to the specific range of numerical values, which will not be described in detail.

[0092] As an example, in one embodiment of the present invention, the expression for the reference number of target coronary heart disease patients may be specifically, for example, as follows:

[0093] .

[0094] .

[0095] in, represents the reference number of target CHD patients; represents the weight parameter of the target CHD patients; Indicates a preset second quantity; represents the variance of the characteristic coefficients of all local neighbors of the target CHD patient; Indicates the consistency of local distribution of target CHD patients; represents the number of patients in the initial local neighborhood of the target CHD patient; represents the initial radius of the target CHD patient, ; Indicates the local density of target CHD patients; represents the normalization function; Indicates the rounding up symbol; Indicates the preset adjustment parameter, which is used to prevent the denominator from being 0. The value range is , in one embodiment of the present invention, Set to 0.01, The specific value of can also be set by the implementer according to the specific implementation scenario and is not limited here.

[0096] It should be noted that, in other embodiments of the present invention, negative correlation mapping may also be achieved through other basic mathematical operations, which are not limited here.

[0097] Then, based on the distance between the target coronary heart disease patient and each neighboring patient, the local density of the target coronary heart disease patient is analyzed. At the same time, combined with the number of coronary heart disease patients among all the neighboring patients of the target coronary heart disease patient, the reference value of the target coronary heart disease patient in the subsequent assessment of the risk of coronary heart disease in the patient to be tested is obtained. Subsequently, based on the confidence level, weighted adjustments can be made when analyzing the possibility of the patient to be tested being ill, thereby improving the accuracy of the calculation and analysis.

[0098] Preferably, in one embodiment of the present invention, the method for obtaining the confidence level of the target coronary heart disease patient specifically includes:

[0099] First, a negative correlation mapping is performed on the average value of the Euclidean distance between the target CHD patient and all neighboring patients to obtain the local density value of the target CHD patient. The larger the local density value, the closer the neighboring patient is to the target CHD patient.

[0100] The more coronary heart disease patients there are among all the neighboring patients of the target coronary heart disease patient and the smaller the local density value of the target coronary heart disease patient is, the greater the reference value of the target coronary heart disease patient is. Therefore, the number of coronary heart disease patients among all the neighboring patients of the target coronary heart disease patient can be taken as the numerator, the local density value as the denominator, and the ratio can be taken as the confidence level of the target coronary heart disease patient.

[0101] As an example, in one embodiment of the present invention, the expression of the confidence level of the target coronary heart disease patient may be specifically, for example:

[0102] .

[0103] .

[0104] in, Indicates the confidence level of the target CHD patient; represents the number of patients with coronary heart disease among all the close neighbors of the target coronary heart disease patient; represents the local density value of the target coronary heart disease patient, ; Indicates the target coronary heart disease patients and The Euclidean distance between the nearest patients; represents the reference number of target CHD patients, that is, the number of close neighbors of the target CHD patients; represents the average value of the Euclidean distance between the target CHD patient and all neighboring patients, .

[0105] It should be noted that, in other embodiments of the present invention, negative correlation mapping may also be achieved through other basic mathematical operations, which are not limited here.

[0106] The same method as above can be used to obtain the confidence level of each coronary heart disease patient in the target monitoring process, and then the coronary heart disease risk of any patient to be tested can be analyzed. Any patient to be tested can be taken as the target patient to be tested. The closer the target patient to be tested is to the coronary heart disease patient, the higher the possibility that the target patient to be tested suffers from coronary heart disease. Therefore, based on the distance between the target patient to be tested and each coronary heart disease patient, and combined with the confidence level of each coronary heart disease patient, the morbidity coefficient obtained can reflect the possibility that the target patient to be tested suffers from coronary heart disease during the target monitoring process. Subsequently, the trend changes of the morbidity coefficient of the target patient to be tested during multiple monitoring processes can be analyzed, so as to accurately predict the coronary heart disease risk of the target patient to be tested.

[0107] Preferably, in one embodiment of the present invention, the method for obtaining the disease coefficient of the target patient to be tested during the target monitoring process specifically includes:

[0108] The Euclidean distance between the target patient to be tested and each coronary heart disease patient is negatively mapped to obtain the initial proximity between the target patient to be tested and each coronary heart disease patient. The greater the initial proximity, the closer the target patient to be tested is to each coronary heart disease patient.

[0109] The greater the confidence level of each coronary heart disease patient, the greater the reference value of the coronary heart disease patient. Therefore, the product of the proximity between the target patient to be tested and each coronary heart disease patient and the confidence level of each coronary heart disease patient can be used as the adjusted proximity between the target patient to be tested and each coronary heart disease patient, thereby improving the accuracy of the proximity analysis between the target patient to be tested and each coronary heart disease patient.

[0110] Then the average value of the adjusted proximity between the target patient to be tested and all patients with coronary heart disease is used as the morbidity coefficient of the target patient to be tested in the target monitoring process.

[0111] As an example, in one embodiment of the present invention, the expression of the disease coefficient of the target patient to be tested during the target monitoring process can be specifically, for example, as follows:

[0112] .

[0113] in, It indicates the disease coefficient of the target patient to be tested during the target monitoring process; Indicates The confidence level of each patient with coronary heart disease; Indicates the target patient and the The Euclidean distance between patients with coronary heart disease; Indicates the target patient and the Initial proximity between patients with coronary heart disease; Indicates the target patient and the Adjusted proximity between patients with coronary heart disease; Represents the number of patients with coronary heart disease.

[0114] It should be noted that, in other embodiments of the present invention, negative correlation mapping may also be achieved through other basic mathematical operations, which are not limited here.

[0115] The same method as above can be used to obtain the disease coefficient of the target patient in each monitoring process. During multiple monitoring processes, if the disease coefficient of the target patient shows an upward trend, it means that the target patient is more likely to have a risk of coronary heart disease. Therefore, according to the changes in the disease coefficient of the target patient in each monitoring process, the disease possibility of the target patient can be obtained. Subsequently, based on the disease possibility, the coronary heart disease risk of the target patient can be accurately predicted.

[0116] Preferably, in one embodiment of the present invention, the method for obtaining the disease probability of the target patient to be tested specifically includes:

[0117] First, it is necessary to analyze the changing trend of the disease coefficient of the target patient to be tested during multiple monitoring processes, and obtain the first possibility of the target patient to be tested based on the difference in the disease coefficient of the target patient to be tested between each monitoring process and the next adjacent monitoring process.

[0118] Preferably, in one embodiment of the present invention, the method for obtaining the first possibility of the target patient to be tested specifically includes:

[0119] The disease coefficient of the target patient to be tested in the next adjacent monitoring process of each monitoring process is taken as the numerator, and the disease coefficient of the target patient to be tested in each monitoring process is taken as the denominator, and the ratio is taken as the change trend parameter of the target patient to be tested in each monitoring process. The larger the change trend parameter is, the more it means that the disease coefficient of the target patient to be tested shows an upward trend during multiple monitoring processes, and the target patient to be tested is more likely to be at risk of coronary heart disease.

[0120] It should be noted that there is no adjacent next monitoring process to the last monitoring process. At this time, the average value of the change trend parameters of the target patient to be tested in all monitoring processes before the last monitoring process can be used as the change trend parameter of the target patient to be tested in the last monitoring process, so as to facilitate the smooth progress of subsequent calculations.

[0121] The average value of the change trend parameters of the target patient to be tested during all monitoring processes is taken as the first possibility of the target patient to be tested.

[0122] At the same time, the greater the overall level of the target patient's morbidity coefficient during all monitoring processes, the greater the risk of the target patient suffering from coronary heart disease. Therefore, the average value of the target patient's morbidity coefficient during all monitoring processes can be used as the second possibility of the target patient.

[0123] Finally, the first possibility and the second possibility are integrated and normalized to obtain the disease probability of the target patient.

[0124] In the embodiment of the present invention, the integration of the first possibility and the second possibility can be achieved by calculating the sum or product of the two possibilities, which is not limited here.

[0125] As an example, in one embodiment of the present invention, the expression of the disease probability of the target patient to be tested can be specifically, for example, as follows:

[0126] .

[0127] in, Indicates the probability of disease in the target patient to be tested; Indicates the first possibility of the target patient to be tested; Indicates the second possibility of the target patient to be tested; Represents the normalization function.

[0128] The disease probability of each patient to be tested can be obtained by the same method as above.

[0129] Step S4: Based on the probability of disease, the risk of the target patient suffering from coronary heart disease is predicted.

[0130] The greater the probability of the target patient suffering from the disease, the higher the risk of the target patient suffering from coronary heart disease. Therefore, based on the probability of suffering from the disease, the risk of the target patient suffering from coronary heart disease can be accurately predicted, thereby improving the accuracy of coronary heart disease risk prediction.

[0131] Preferably, in one embodiment of the present invention, the method for predicting the risk of a target patient suffering from coronary heart disease specifically includes:

[0132] If the probability of the target patient to be tested having the disease is greater than the preset probability threshold, the target patient to be tested is at risk of suffering from coronary heart disease, and an early warning is issued to the target patient to remind the patient to take timely prevention and treatment. The early warning can be achieved through the prompt lights in the hospital and other means.

[0133] The preset possibility threshold is set to 0.5, and the specific value of the preset possibility threshold can also be set by the implementer according to the specific implementation scenario, which is not limited here.

[0134] Then, the same method as above can be used to predict the risk of coronary heart disease for each patient.

[0135] It should be noted that the sequence of the above embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0136] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for predicting the risk of coronary heart disease based on sleep apnea syndrome, characterized in that: The method comprises: Acquire signal data of different signal types and index data of different body index types of each patient during each monitoring process, and acquire the age of each patient, the patients including patients with coronary heart disease and patients to be tested; Take any monitoring process as a target monitoring process, take the youngest patient as a reference patient, take any patient as a target patient, and in the target monitoring process, obtain the first relative characteristic value, the second relative characteristic value, and the third relative characteristic value of the target patient in the target monitoring process according to the difference in the signal data of the same signal type, the difference in the indicator data of the same body indicator type, and the difference in age between the target patient and the reference patient; in the target monitoring process, map each patient to a spatial coordinate system based on the first relative characteristic value, the second relative characteristic value, and the third relative characteristic value of each patient to obtain a spatial scatter plot of the target monitoring process; In the spatial scatter plot, any one of the patients with coronary heart disease is taken as the target patient with coronary heart disease, and the local neighborhoods of different radii of the target patient with coronary heart disease are constructed with the target patient with coronary heart disease as the center of the sphere; multiple neighboring patients of the target patient with coronary heart disease are obtained according to the number of patients in each local neighborhood and the radius of the local neighborhood; the confidence level of the target patient with coronary heart disease is obtained according to the distance between the target patient with coronary heart disease and each neighboring patient, and the number of patients with coronary heart disease among all the neighboring patients of the target patient with coronary heart disease; any one of the patients to be tested is taken as the target patient to be tested, and the disease coefficient of the target patient to be tested in the target monitoring process is obtained according to the distance between the target patient to be tested and each patient with coronary heart disease, and the confidence level of each patient with coronary heart disease; the disease possibility of the target patient to be tested is obtained according to the change of the disease coefficient of the target patient to be tested in each monitoring process; Based on the disease possibility, the risk of the target patient suffering from coronary heart disease is predicted.

2. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The obtaining of the first relative characteristic value, the second relative characteristic value and the third relative characteristic value of the target patient in the target monitoring process comprises: Taking any signal type as a target signal type, performing Fourier transform processing on the signal data of the target signal type of the target patient and the signal data of the target signal type of the reference patient, respectively, to obtain a frequency spectrum diagram of the target patient with respect to the target signal type and a frequency spectrum diagram of the reference patient with respect to the target signal type; Taking the difference in amplitude values ​​of the same frequency value between the target patient's spectrum graph regarding the target signal type and the reference patient's spectrum graph regarding the target signal type as the relative amplitude difference value of each frequency value, and taking the average of the relative amplitude difference values ​​of all frequency values ​​as the relative signal difference value of the target patient regarding the target signal type; Taking any one body indicator type as a target body indicator type, and taking the difference between the indicator data of the target body indicator type of the target patient and the indicator data of the target body indicator type of the reference patient as the relative indicator difference value of the target patient with respect to the target body indicator type; Taking the set consisting of the relative signal difference values ​​of all signal types and the relative index difference values ​​of all body index types of the target patient as a reference set of the target patient; Taking the average value of all elements in the reference set as the first relative characteristic value of the target patient, and taking the average value of the absolute values ​​of all elements in the reference set as the second relative characteristic value of the target patient; The difference between the age of the target patient and the age of the reference patient is used as the third relative characteristic value of the target patient.

3. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The spatial scatter plot of the target monitoring process is obtained as follows: Constructing a spatial coordinate system, wherein the horizontal axis of the spatial coordinate system is the third relative eigenvalue, the vertical axis of the spatial coordinate system is the first relative eigenvalue, and the vertical axis of the spatial coordinate system is the second relative eigenvalue; The three-dimensional data points formed by the third relative eigenvalue, the first relative eigenvalue and the second relative eigenvalue of each patient in the target monitoring process are mapped to the spatial coordinate system to obtain a spatial scatter plot of the target monitoring process, wherein each three-dimensional data point represents a patient.

4. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The method of constructing local neighborhoods of different radii of the target coronary heart disease patient with the target coronary heart disease patient as the sphere center includes: The patient closest to the target coronary heart disease patient is used as the nearest patient of the target coronary heart disease patient, and the Euclidean distance between the target coronary heart disease patient and the nearest patient of the target coronary heart disease patient is used as the initial radius of the target coronary heart disease patient; Taking the target coronary heart disease patient as the sphere center and the spherical area with the initial radius as the radius, as the initial local neighborhood of the target coronary heart disease patient; The radius of the initial local neighborhood is expanded a preset first number of times with a preset expansion step length to obtain local neighborhoods of different radii for the target coronary heart disease patient.

5. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 4, characterized in that: The method of obtaining a plurality of neighboring patients of the target coronary heart disease patient comprises: The number of patients in each local neighborhood is used as the numerator, the cube of the radius of each local neighborhood is used as the denominator, the ratio is used as the characteristic coefficient of each local neighborhood, the variance of the characteristic coefficients of all local neighborhoods is negatively correlated and mapped, and the local distribution consistency of the target coronary heart disease patients is obtained; The number of patients in the initial local neighborhood of the target coronary heart disease patient is used as the numerator, the initial radius of the target coronary heart disease patient is used as the denominator, and the ratio is used as the local density of the target coronary heart disease patient; The local distribution consistency and the local density are integrated and normalized to obtain a weight parameter of a target coronary heart disease patient; Rounding up the product of the weight parameter and the preset second number to obtain a reference number of target coronary heart disease patients; The reference number of patients closest to the target coronary heart disease patient are regarded as the neighboring patients of the target coronary heart disease patient.

6. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The confidence level of obtaining the target coronary heart disease patient includes: Performing negative correlation mapping on the average values ​​of the Euclidean distances between the target coronary heart disease patient and all the neighboring patients to obtain a local density value of the target coronary heart disease patient; The number of patients with coronary heart disease among all the neighboring patients of the target patient with coronary heart disease is used as the numerator, the local density value is used as the denominator, and the ratio is used as the confidence level of the target patient with coronary heart disease.

7. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The disease coefficient of the target patient to be tested in the target monitoring process includes: Negative correlation mapping is performed on the Euclidean distance between the target patient to be tested and each coronary heart disease patient to obtain the initial proximity between the target patient to be tested and each coronary heart disease patient; The product value of the proximity degree between the target patient to be tested and each coronary heart disease patient and the confidence degree of each coronary heart disease patient is used as the adjusted proximity degree between the target patient to be tested and each coronary heart disease patient; The average value of the adjusted proximity between the target patient to be tested and all patients with coronary heart disease is used as the disease coefficient of the target patient to be tested in the target monitoring process.

8. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The method of obtaining the disease probability of the target patient to be tested includes: Obtaining a first possibility of the target patient to be tested according to a difference in the disease coefficient of the target patient to be tested between each monitoring process and the next adjacent monitoring process; The average of the disease coefficients of the target patient to be tested during all monitoring processes is used as the second possibility of the target patient to be tested; The first possibility and the second possibility are integrated and normalized to obtain the disease possibility of the target patient to be tested.

9. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 8, characterized in that: The first possibility of obtaining the target patient to be tested includes: The disease coefficient of the target patient to be tested in the next monitoring process adjacent to each monitoring process is used as the numerator, the disease coefficient of the target patient to be tested in each monitoring process is used as the denominator, and the ratio is used as the change trend parameter of the target patient to be tested in each monitoring process; The average value of the change trend parameter of the target patient to be tested during all monitoring processes is used as the first possibility of the target patient to be tested.

10. The method for predicting the risk of coronary heart disease based on sleep apnea syndrome according to claim 1, characterized in that: The prediction of the risk of the target patient suffering from coronary heart disease comprises: If the disease probability of the target patient to be tested is greater than a preset probability threshold, the target patient to be tested is at risk of suffering from coronary heart disease, and an early warning is issued to the target patient to be tested.

Images