Early screening and grading management system for senile heart failure based on metabolic abnormalities

Abstract

The present application relates to the field of heart failure risk assessment, in particular to a kind of early screening and hierarchical management system of senile heart failure based on metabolic abnormality.The system first obtains the initial heart failure performance value of the senile patient to be measured according to the blood sodium concentration value of the senile patient to be measured at each time point and the change of blood sodium concentration value in time sequence, according to the difference between the senile patient to be measured and the senile patient to be selected in the drug data information, the difference between the senile patient to be measured and the senile patient to be selected in the blood drug concentration value at each time point and the time difference of the last time taking medicine, the reference senile patient of the senile patient to be measured is screened from all the senile patients to be selected, according to the difference between the senile patient to be measured and the senile patient to be selected in the blood sodium concentration value at each time point, and the initial heart failure performance value of the senile patient to be measured, the heart failure performance of the senile patient to be measured is hierarchically evaluated.The present application can improve the accuracy of screening and hierarchical evaluation of senile heart failure.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of heart failure risk assessment, in particular to an early screening and grading management system for senile heart failure based on metabolic abnormalities. BACKGROUND

[0002] With the acceleration of global aging process, the incidence and mortality of heart failure in the elderly population are increasing year by year. The early symptoms of heart failure are occult and highly heterogeneous. The elderly patients often have multiple chronic underlying diseases (such as hypertension, diabetes, metabolic syndrome), leading to delayed diagnosis and missed best intervention period. Therefore, early screening and grading management of heart failure in the elderly is particularly important.

[0003] With the development of senile heart failure, the content of metabolic substances in the human body, such as the content of sodium ions or potassium ions in the blood, will gradually decrease. Therefore, in the related art, the content of metabolic substances such as sodium ions or potassium ions in the blood of the elderly patients is detected to realize the screening, grading and evaluation of heart failure in the elderly patients. However, since the elderly patient population is usually in a sub-healthy state and suffers from some underlying diseases such as heart disease or diabetes, the elderly patients need to take various drugs to maintain a healthy state. Taking drugs will change the content of metabolic substances in the body, masking the change characteristics of the content of metabolic substances caused by heart failure, thereby reducing the accuracy of heart failure screening and grading evaluation of the elderly patients. SUMMARY

[0004] In order to solve the technical problem that taking drugs by the elderly patients will change the content of metabolic substances in the body, mask the change characteristics of the content of metabolic substances caused by heart failure, and reduce the accuracy of heart failure screening and grading evaluation of the elderly patients, the purpose of the present application is to provide an early screening and grading management system for senile heart failure based on metabolic abnormalities, and the technical solution adopted is as follows:

[0005] The present application also provides an early screening and grading management system for senile heart failure based on metabolic abnormalities, which comprises:

[0006] A data acquisition module is configured to obtain the blood sodium concentration value of the to-be-tested elderly patient and each to-be-selected elderly patient without heart failure at each time point, and the medication data information of the to-be-tested elderly patient and each to-be-selected elderly patient in different drug dimensions, and simultaneously obtain the blood drug concentration value of the to-be-tested elderly patient and each to-be-selected elderly patient at each time point after taking the drug last time.

[0007] A preliminary evaluation module is configured to obtain the initial heart failure performance value of the to-be-tested elderly patient according to the blood sodium concentration value of the to-be-tested elderly patient at each time point and the change of the blood sodium concentration value over time.

[0008] The similarity analysis module is configured to: take any one of the candidate elderly patients as a target candidate elderly patient, and obtain a similarity of medication information between the target elderly patient and the target candidate elderly patient according to a difference of the medication data information of the same drug between the target elderly patient and the target candidate elderly patient; obtain a similarity of drug efficacy between the target elderly patient and the target candidate elderly patient according to a difference of the blood drug concentration values of the same drug at each time point and a time difference of the last time of taking the drug between the target elderly patient and the target candidate elderly patient; and combine the similarity of medication information and the similarity of drug efficacy between the target elderly patient and each candidate elderly patient to screen a reference elderly patient of the target elderly patient from all the candidate elderly patients.

[0009] The grading evaluation module is configured to grade and evaluate the heart failure performance of the target elderly patient according to a difference of the blood sodium concentration values at each time point between the target elderly patient and the reference elderly patient and the initial heart failure performance value of the target elderly patient.

[0010] Further, the obtaining of the initial heart failure performance value of the target elderly patient comprises:

[0011] performing negative correlation mapping on the average value of the blood sodium concentration values of the target elderly patient at all time points to obtain a first heart failure performance value of the target elderly patient;

[0012] performing curve fitting on the blood sodium concentration values of the target elderly patient at each time point to obtain a blood sodium concentration fitting curve of the target elderly patient;

[0013] obtaining a second heart failure performance value of the target elderly patient according to a difference of the slopes of the blood sodium concentration fitting curve at adjacent time points;

[0014] performing comprehensive processing on the first heart failure performance value and the second heart failure performance value and performing normalization processing to obtain the initial heart failure performance value of the target elderly patient.

[0015] Further, the obtaining of the second heart failure performance value of the target elderly patient comprises:

[0016] taking any two adjacent time points as an adjacent time point group, and taking an absolute value of a difference of the slopes of the blood sodium concentration fitting curve at the two time points in each adjacent time point group as a blood sodium decline trend value of each adjacent time point group;

[0017] taking an average value of the blood sodium decline trend values of all the adjacent time point groups as the second heart failure performance value of the target elderly patient.

[0018] Further, the obtaining of the similarity of medication information between the target elderly patient and the target candidate elderly patient comprises:

[0019] Quantitatively process the medication data information of each drug dimension of the to-be-tested elderly patient and the target candidate elderly patient, respectively, to obtain the word vector of each drug dimension of the to-be-tested elderly patient and the target candidate elderly patient;

[0020] The cosine similarity of the word vectors of the same drug dimension between the to-be-tested elderly patient and the target candidate elderly patient is taken as the initial information similarity of each drug dimension between the to-be-tested elderly patient and the target candidate elderly patient;

[0021] The average value of the initial information similarity of all drug dimensions between the to-be-tested elderly patient and the target candidate elderly patient is normalized to obtain the medication information similarity between the to-be-tested elderly patient and the target candidate elderly patient.

[0022] Further, the obtaining of the efficacy similarity between the to-be-tested elderly patient and the target candidate elderly patient comprises:

[0023] The absolute value of the difference between the time of the last drug taking of the to-be-tested elderly patient and the target candidate elderly patient is negatively correlated to obtain the drug taking time closeness between the to-be-tested elderly patient and the target candidate elderly patient;

[0024] In chronological order, the blood drug concentration values of each drug taken by the to-be-tested elderly patient and the target candidate elderly patient at all time points are sorted to obtain the blood drug concentration sequence of each drug of the to-be-tested elderly patient and the target candidate elderly patient;

[0025] The blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient is obtained according to the difference between the blood drug concentration sequences of the same drug between the to-be-tested elderly patient and the target candidate elderly patient;

[0026] The drug taking time closeness and the blood drug concentration similarity are integrated and normalized to obtain the efficacy similarity between the to-be-tested elderly patient and the target candidate elderly patient.

[0027] Further, the obtaining of the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient comprises:

[0028] The blood drug concentration sequences of the same drug of the to-be-tested elderly patient and the target candidate elderly patient are input into a dynamic time warping algorithm, and the minimum cumulative distance output is negatively correlated to obtain the blood drug concentration similarity coefficient between the to-be-tested elderly patient and the target candidate elderly patient for each same drug;

[0029] The average value of the blood drug concentration similarity coefficients of all same drugs between the to-be-tested elderly patient and the target candidate elderly patient is taken as the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient.

[0030] Further, the reference elderly patient selected from all the candidate elderly patients for the to-be-tested elderly patient comprises:

[0031] The medication information similarity and the drug efficacy similarity between the to-be-tested elderly patient and each of the candidate elderly patients are integrated and normalized to obtain a disease similarity between the to-be-tested elderly patient and each of the candidate elderly patients.

[0032] The reference elderly patient of the to-be-tested elderly patient is selected from all the candidate elderly patients based on the disease similarity.

[0033] Further, the reference elderly patient of the to-be-tested elderly patient is selected from all the candidate elderly patients based on the disease similarity, and the method comprises:

[0034] The candidate elderly patient with the disease similarity greater than a preset similarity threshold is taken as the reference elderly patient of the to-be-tested elderly patient.

[0035] Further, the grading assessment of the heart failure performance of the to-be-tested elderly patient comprises:

[0036] The blood sodium concentration values of the to-be-tested elderly patient and each of the reference elderly patients at all time points are sorted in chronological order to obtain a blood sodium concentration sequence of the to-be-tested elderly patient and each of the reference elderly patients.

[0037] The heart failure performance assessment value of the to-be-tested elderly patient is obtained according to the difference between the blood sodium concentration sequences of the to-be-tested elderly patient and each of the reference elderly patients and the initial heart failure performance value of the to-be-tested elderly patient.

[0038] If the heart failure performance assessment value is not greater than a preset first assessment threshold, the to-be-tested elderly patient is classified as a primary heart failure patient; if the heart failure performance assessment value is greater than the preset first assessment threshold and not greater than a preset second threshold, the to-be-tested elderly patient is classified as an intermediate heart failure patient; and if the heart failure performance assessment value is greater than the preset second assessment threshold, the to-be-tested elderly patient is classified as a high-level heart failure patient, wherein the heart failure severity of the primary heart failure patient, the intermediate heart failure patient and the high-level heart failure patient increases in turn.

[0039] Further, the heart failure performance assessment value of the to-be-tested elderly patient is obtained, and the method comprises:

[0040] The blood sodium concentration sequences of the to-be-tested elderly patient and each of the reference elderly patients are input into a dynamic time warping algorithm, and the output minimum cumulative distance is taken as the blood sodium concentration difference between the to-be-tested elderly patient and each of the reference elderly patients.

[0041] The average of the difference degree of the blood sodium concentration between the to-be-tested elderly patient and all reference elderly patients is taken as the overall difference degree of the to-be-tested elderly patient;

[0042] The overall difference degree and the initial heart failure performance value of the to-be-tested elderly patient are comprehensively combined and normalized to obtain a heart failure performance evaluation value of the to-be-tested elderly patient.

[0043] The present application has the following advantages:

[0044] The present application has the following advantages: BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description only represent some embodiments of the present application, and for those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0046] Figure 1 A block diagram of an elderly heart failure early screening and hierarchical management system based on metabolic abnormalities is provided in one embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object, the following describes in detail the specific implementation, structure, features and effects of an elderly heart failure early screening and hierarchical management system based on metabolic abnormalities according to the present application, with reference to the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0048] Unless otherwise defined, 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 application belongs.

[0049] The specific scheme of the early screening and grading management system for elderly heart failure based on metabolic abnormalities is specifically described below in combination with the drawings.

[0050] Please refer to Figure 1 which shows a block diagram of an early screening and grading management system for elderly heart failure based on metabolic abnormalities according to an embodiment of the present application, which comprises a data acquisition module 101, a preliminary evaluation module 102, a similarity analysis module 103 and a grading evaluation module 104.

[0051] The data acquisition module 101 is configured to acquire the blood sodium concentration value of the to-be-tested elderly patient and each to-be-selected elderly patient without heart failure at each time point, and the medication data information of the to-be-tested elderly patient and each to-be-selected elderly patient in different drug dimensions, and simultaneously acquire the blood drug concentration value of the to-be-tested elderly patient and each to-be-selected elderly patient at each time point after taking the drug last time.

[0052] When the elderly patient shows signs of heart failure, the heart failure leads to a decrease in the heart pumping function, the effective circulating blood volume in the body is reduced, the renin-angiotensin-aldosterone system (RAAS) is activated, after the activation of the RAAS system, aldosterone promotes the reabsorption of sodium ions in the renal tubule, at the same time, it will aggravate water retention, leading to the dilution of the blood sodium concentration (i.e. the concentration of sodium ions in the blood), at the same time, when the patient shows signs of heart failure, because the heart pumping function is decreased, the effective circulating blood volume in the patient's body is reduced, the baroreceptor in the body is activated, stimulating the hypothalamus to secrete non-osmotic arginine vasopressin, leading to an increase in water reabsorption in the glomerulus, the blood is diluted, and the blood sodium concentration is decreased.

[0053] Therefore, the medical device such as the microneedle sensor is used in the embodiment of the present application to collect the blood sodium concentration value of the to-be-tested elderly patient and each to-be-selected elderly patient without heart failure at each time point, wherein the to-be-tested elderly patient may have underlying diseases such as hypertension, diabetes, etc., but whether the to-be-tested elderly patient has signs of heart failure is unknown, and the signs of heart failure of the to-be-tested elderly patient need to be evaluated subsequently, while the to-be-selected elderly patient does not have signs of heart failure, but may also have various underlying diseases, in an embodiment of the present application, each day is regarded as a time point, and the data of each patient in 30 days is collected at the same time, the specific setting mode of the time point and the total data collection time can be set by the implementer according to the specific implementation scene, which is not limited herein.

[0054] Since the old patient group usually has underlying diseases, the old patients need to take various drugs to maintain a healthy state, and taking drugs can change the blood sodium concentration, mask the characteristics of heart failure, so the embodiment of the application also needs to record the medication data information of different drug dimensions of the to-be-tested old patient and each candidate old patient, wherein the drug dimensions include, for example, drug type, medication frequency, medication dose and drug efficacy, etc. Since the effect of the last taken drug on the patient is the most significant, the embodiment of the application also needs to use a microneedle sensor or a chromatographic device to collect the blood drug concentration values of different drugs of the to-be-tested old patient and each candidate old patient at each time point after the last taken drug, and record the time of the last taken drug of the to-be-tested old patient and each candidate old patient.

[0055] The preliminary evaluation module 102 is configured to obtain an initial heart failure performance value of the to-be-tested old patient according to the blood sodium concentration values of the to-be-tested old patient at each time point and the change of the blood sodium concentration values over time.

[0056] When the to-be-tested old patient shows signs of heart failure, the blood sodium concentration in the to-be-tested old patient shows a downward trend, and as the degree of heart failure increases, the blood sodium concentration in the to-be-tested old patient will also decrease rapidly, showing a parabolic change trend. Therefore, the blood sodium concentration values of the to-be-tested old patient at each time point and the change of the blood sodium concentration values over time can be analyzed, and the possibility of the to-be-tested old patient having heart failure can be preliminarily evaluated by the obtained initial heart failure performance value.

[0057] Preferably, in an embodiment of the application, the method for obtaining the initial heart failure performance value of the to-be-tested old patient specifically comprises:

[0058] Firstly, the lower the overall level of the blood sodium concentration of the to-be-tested old patient, the more likely it is that there are signs of heart failure. Therefore, the average value of the blood sodium concentration values of the to-be-tested old patient at all time points can be negatively correlated and mapped to obtain a first heart failure performance value of the to-be-tested old patient.

[0059] Then, the blood sodium concentration values of the to-be-tested old patient at each time point are curve-fitted to obtain a blood sodium concentration fitting curve of the to-be-tested old patient. The existence of an accelerated change in the blood sodium concentration of the to-be-tested old patient over time indicates that there are more likely signs of heart failure. Therefore, the difference in the slope of the blood sodium concentration fitting curve at adjacent time points can be used to obtain a second heart failure performance value of the to-be-tested old patient.

[0060] Preferably, in an embodiment of the application, the method for obtaining the second heart failure performance value of the to-be-tested old patient specifically comprises:

[0061] The absolute value of the difference between the slopes of the fitting curve of the blood sodium concentration at the two time points in each adjacent time point group is taken as the blood sodium decline trend value of each adjacent time point group, and the greater the blood sodium decline trend value is, the greater the difference between the decline rates of the blood sodium concentration at the two adjacent time points is, and thus the blood sodium concentration presents an accelerated decline trend in time sequence. Therefore, the average value of the blood sodium decline trend values of all the adjacent time point groups can be taken as the second heart failure performance value of the to-be-tested elderly patient.

[0062] Further, the first heart failure performance value and the second heart failure performance value are comprehensively processed and normalized, and the calculation result is limited in the range of 0 to 1, so as to obtain the initial heart failure performance value of the to-be-tested elderly patient.

[0063] In the embodiments of the present application, the comprehensive processing of the first heart failure performance value and the second heart failure performance value can be realized by calculating the sum or product value of the two values, which is not limited herein, and the comprehensive processing of two or more data in the subsequent steps can also be realized by using the same method.

[0064] In one embodiment of the present application, the normalization processing can be realized by using an activation function and a hyperbolic tangent function, which is not limited herein, and the data normalization processing in the subsequent steps can also be realized by using the same method.

[0065] As an example, in one embodiment of the present application, the expression of the initial heart failure performance value of the to-be-tested elderly patient can be specifically as follows:

[0066]

[0067] wherein, represents the initial heart failure performance value of the to-be-tested elderly patient; represents the blood sodium concentration value of the to-be-tested elderly patient at the i th time point, ; i = 1, 2, 3, …, n ; and represents the number of time points, and n represents the number of adjacent time point groups. represents the number of adjacent time point groups; represents the first heart failure performance value of the to-be-tested elderly patient; represents the blood sodium decline trend value of the i th adjacent time point group; represents the second heart failure performance value of the to-be-tested elderly patient; represents the hyperbolic tangent function, which is used for normalization processing. It should be noted that the negative correlation mapping can also be realized by other basic mathematical operations in other embodiments of the present application, which is not described herein.

[0068]

[0069] ​​Thus, the analysis of the heart failure signs of the to-be-tested elderly patient is preliminarily realized.

[0070] The similarity analysis module 103 is configured to take any one of the candidate elderly patients as a target candidate elderly patient, obtain the similarity of the medication information between the to-be-tested elderly patient and the target candidate elderly patient according to the difference in the medication data information of the same drug dimension between the to-be-tested elderly patient and the target candidate elderly patient, obtain the similarity of the drug efficacy between the to-be-tested elderly patient and the target candidate elderly patient according to the difference in the blood drug concentration value of the same drug at each time point and the time difference of the last time of taking the drug between the to-be-tested elderly patient and the target candidate elderly patient, and combine the similarity of the medication information and the similarity of the drug efficacy between the to-be-tested elderly patient and each candidate elderly patient to screen the reference elderly patient of the to-be-tested elderly patient from all the candidate elderly patients.

[0071] The early heart failure will cause the blood sodium concentration in the patient to decrease, but because the elderly patient has a decline in body function, the patient often has a plurality of underlying diseases, such as diabetes, hypertension, etc. The patient usually takes corresponding drugs to control the disease in order to inhibit the development of the disease, but the drug efficacy of some drugs will affect the change of the electrolyte concentration content of the human body. For example, the diuretic which has a good effect on the treatment of hypertension, long-term use of the diuretic will increase the osmotic pressure of the cell, increase the content of sodium ions in the blood, and at the same time, the diuretic promotes the excretion of sodium ions in the body, reduces the total amount of sodium ions in the body, and causes the decrease of the blood sodium concentration. Therefore, the change of the blood sodium concentration caused by the drugs taken by the elderly patient will mask the heart failure characteristics, and it is difficult to accurately evaluate the heart failure performance according to the change of the blood sodium ion concentration of the elderly patient. Therefore, it is necessary to find a patient group similar to the underlying disease of the to-be-tested elderly patient in the candidate elderly patients, take the patient group similar to the underlying disease as a control group, compare the characteristic differences of the blood sodium concentration between the to-be-tested elderly patient and the candidate elderly patient similar to the underlying disease, and more accurately perform early screening and grading evaluation on the heart failure signs of the to-be-tested elderly patient.

[0072] Before selecting the to-be-selected elderly patient similar to the to-be-tested elderly patient in the basic disease, the similarity of the basic diseases between the to-be-tested elderly patient and the to-be-selected elderly patient needs to be evaluated and analyzed. When a certain to-be-selected elderly patient and the to-be-tested elderly patient have similar basic diseases, the drug conditions of the two are similar, and the effects of the drugs in the two bodies are also similar. Therefore, the embodiments of the present application first take any to-be-selected elderly patient as a target to-be-selected elderly patient, and analyze the differences in the medication data information of the same drug dimension between the to-be-tested elderly patient and the target to-be-selected elderly patient. The similarity of the medication information reflects the similarity of the medication conditions between the to-be-tested elderly patient and the target to-be-selected elderly patient. The greater the medication information similarity, the more likely it is that the two have similar basic diseases. Subsequently, the to-be-selected elderly patient similar to the to-be-tested elderly patient in the basic disease can be selected based on the medication information similarity.

[0073] Preferably, in an embodiment of the present application, the method for obtaining the medication information similarity between the to-be-tested elderly patient and the target to-be-selected elderly patient specifically comprises:

[0074] Since the medication data information of each drug dimension is usually text information, the medication data information of each drug dimension of the to-be-tested elderly patient and the target to-be-selected elderly patient is first quantitatively processed to obtain the word vector of each drug dimension of the to-be-tested elderly patient and the target to-be-selected elderly patient. In an embodiment of the present application, the existing Doc2Vec algorithm or SBERT algorithm can be used to realize the quantitative processing, which is not limited or elaborated here.

[0075] The cosine similarity of the word vectors between the to-be-tested elderly patient and the target to-be-selected elderly patient in the same drug dimension is taken as the initial information similarity of each drug dimension between the to-be-tested elderly patient and the target to-be-selected elderly patient. The greater the initial information similarity, the more similar the information of the same drug dimension between the to-be-tested elderly patient and the target to-be-selected elderly patient. Furthermore, the average value of the initial information similarity of all drug dimensions between the to-be-tested elderly patient and the target to-be-selected elderly patient can be normalized to limit the calculation result in the range of 0 to 1, thereby obtaining the medication information similarity between the to-be-tested elderly patient and the target to-be-selected elderly patient.

[0076] As an example, in an embodiment of the present application, the expression of the medication information similarity between the to-be-tested elderly patient and the target to-be-selected elderly patient can be specifically, for example:

[0077]

[0078] Wherein, represents the medication information similarity between the to-be-tested elderly patient and the target to-be-selected elderly patient; ​This indicates the first [number] of the elderly patients to be tested. Word vectors for each drug dimension; This indicates the first target elderly patient. Word vectors for each drug dimension; Indicates the number of drug dimensions; This indicates the first [number] elderly patient between the target elderly patient and the intended target elderly patient. Initial information similarity across drug dimensions; This represents the hyperbolic tangent function, used for normalization.

[0079] The medication information similarity between the elderly patients to be tested and each of the selected elderly patients can be obtained by using the same method described above.

[0080] Considering that the same drug, when administered to different patients, will have varying absorption rates due to individual differences, leading to differences in its efficacy, the effects of the same drug will also differ among patients. When a candidate elderly patient and a test elderly patient have similar underlying diseases, the effects of the same drug taken by both will be relatively similar. Since the drug's effect on a patient can be reflected by changes in the drug concentration in the patient's blood, the similarity of drug effects when taken by the two elderly patients can be analyzed by examining the differences in blood drug concentrations of the same drug at various time points between the test elderly patient and the target candidate elderly patient. Considering that the closer the time of the last medication use between the elderly patients to be tested and the target elderly patients, the greater the reference value of the drug efficacy similarity analysis between the elderly patients to be tested and the target elderly patients, the drug efficacy similarity can be obtained by combining the time difference of the last medication use between the elderly patients to be tested and the target elderly patients to reflect the similarity of the drug efficacy exerted by the same drugs taken by the elderly patients to be tested and the target elderly patients. The greater the drug efficacy similarity, the more likely that the two are to have similar underlying diseases. Subsequently, the similarity of medication information and drug efficacy similarity can be combined to screen out the elderly patients with similar underlying diseases to the elderly patients to be tested.

[0081] Preferably, in one embodiment of the present invention, the method for obtaining the pharmacodynamic similarity between the elderly patient to be tested and the target elderly patient specifically includes:

[0082] By performing a negative correlation mapping on the absolute value of the difference in the time of the most recent medication administration between the elderly patients to be tested and the target elderly patients, the similarity in medication administration time between the elderly patients to be tested and the target elderly patients can be obtained.

[0083] In time sequence, the blood drug concentration values of each drug taken by the to-be-tested elderly patient and the target candidate elderly patient at all time points are sorted respectively to obtain the blood drug concentration sequence of each drug of the to-be-tested elderly patient and the target candidate elderly patient.

[0084] According to the difference between the blood drug concentration sequences of the same drugs of the to-be-tested elderly patient and the target candidate elderly patient, the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient is obtained. The greater the blood drug concentration similarity is, the more similar the concentration changes of the same drugs in the blood of the to-be-tested elderly patient and the target candidate elderly patient are, and further, the more consistent the drug effects of the same drugs on the two are.

[0085] Preferably, in an embodiment of the present application, the method for obtaining the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient specifically comprises:

[0086] The blood drug concentration sequences of the same drugs of the to-be-tested elderly patient and the target candidate elderly patient are input into a dynamic time warping algorithm, and the minimum cumulative distance output is negatively correlated to obtain the blood drug concentration similarity coefficient between the to-be-tested elderly patient and the target candidate elderly patient for each same drug. In other embodiments of the present application, the Euclidean distance or Manhattan distance between the blood drug concentration sequences of the same drugs of the two can also be negatively correlated to obtain the blood drug concentration similarity coefficient between the two for each same drug, which is not limited herein.

[0087] Further, the average value of the blood drug concentration similarity coefficients between the to-be-tested elderly patient and the target candidate elderly patient for all same drugs is taken as the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient.

[0088] The greater the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient is, and the closer the drug taking time proximity between the two is, the more similar the drug effects of taking the same drugs between the two are. Therefore, the drug taking time proximity and the blood drug concentration similarity can be comprehensively processed and normalized, and the calculation result is limited in the range of 0 to 1, so as to obtain the drug effect similarity between the to-be-tested elderly patient and the target candidate elderly patient.

[0089] As an example, in an embodiment of the present application, the expression of the drug effect similarity between the to-be-tested elderly patient and the target candidate elderly patient can be specifically, for example:

[0090]

[0091] Wherein, represents the drug effect similarity between the to-be-tested elderly patient and the target candidate elderly patient; ​indicates the time when the drug was last taken by the to-be-tested elderly patient; indicates the time when the drug was last taken by the target to-be-tested elderly patient; indicates the closeness of the drug taking time between the to-be-tested elderly patient and the target to-be-tested elderly patient; indicates the minimum cumulative distance of the blood drug concentration sequence of the same drug between the to-be-tested elderly patient and the target to-be-tested elderly patient output by the dynamic time warping algorithm; indicates the similarity coefficient of the blood drug concentration between the to-be-tested elderly patient and the target to-be-tested elderly patient with respect to the same drug; indicates the closeness of the blood drug concentration between the to-be-tested elderly patient and the target to-be-tested elderly patient; indicates the closeness of the blood drug concentration between the to-be-tested elderly patient and the target to-be-tested elderly patient; indicates the closeness of the blood drug concentration between the to-be-tested elderly patient and the target to-be-tested elderly patient; indicates the number of the same drugs last taken by the to-be-tested elderly patient and the target to-be-tested elderly patient; indicates the hyperbolic tangent function for normalization processing; indicates a preset first adjustment parameter, indicates a preset second adjustment parameter, both of which are used to prevent the denominator from being 0, and the value range of In an embodiment of the present application, and are both set to 0.01, and The specific numerical values of

[0092] The pharmacodynamic similarity between the to-be-tested elderly patient and each to-be-tested elderly patient can be obtained by the same method.

[0093] The greater the medication information similarity and the greater the pharmacodynamic similarity between the to-be-tested elderly patient and a certain to-be-tested elderly patient, the more similar the underlying diseases of the to-be-tested elderly patient and the to-be-tested elderly patient. In turn, the reference elderly patient with similar underlying diseases to the to-be-tested elderly patient can be selected from all to-be-tested elderly patients by combining the medication information similarity and the pharmacodynamic similarity between the to-be-tested elderly patient and each to-be-tested elderly patient.

[0094] Preferably, in an embodiment of the present application, the method for obtaining the reference elderly patient of the to-be-tested elderly patient specifically comprises:

[0095] After the medication information similarity and the pharmacodynamic similarity between the to-be-tested elderly patient and each to-be-tested elderly patient are comprehensively processed and normalized, the calculation result is limited within , so as to obtain the disease similarity between the to-be-tested elderly patient and each to-be-tested elderly patient.

[0096] As an example, in an embodiment of the present application, the expression of the illness similarity between the to-be-tested elderly patient and each of the candidate elderly patients can be specifically, for example:

[0097]

[0098] wherein, represents the illness similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the medication information similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the drug efficacy similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the drug efficacy similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the drug efficacy similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the drug efficacy similarity between the to-be-tested elderly patient and the i-th candidate elderly patient; represents the hyperbolic tangent function, used for normalization processing.

[0099] The greater the illness similarity, the more similar the underlying diseases suffered by the to-be-tested elderly patient and each of the candidate elderly patients, and thus the reference elderly patient of the to-be-tested elderly patient can be selected from all the candidate elderly patients based on the illness similarity.

[0100] Preferably, in an embodiment of the present application, the candidate elderly patient with an illness similarity greater than a preset similarity threshold value can be taken as the reference elderly patient of the to-be-tested elderly patient, wherein the preset similarity threshold value has a value range of 0<similarity threshold value<1. In an embodiment of the present application, the preset similarity threshold value is set to 0.6, and the specific value of the preset similarity threshold value can also be set by the implementer according to the specific implementation scene, which is not limited herein.

[0101] Up to now, the reference elderly patient with similar underlying diseases to the to-be-tested elderly patient has been selected from the candidate elderly patient population.

[0102] The grading evaluation module 104 is configured to grade the heart failure performance of the to-be-tested elderly patient according to the difference between the blood sodium concentration values of the to-be-tested elderly patient and the reference elderly patient at each time point, and the initial heart failure performance value of the to-be-tested elderly patient.

[0103] It can be known from the above analysis that the drug taken by the patient also affects the change of the blood sodium concentration, thereby masking the heart failure characteristics of the patient, so that the initial heart failure performance value cannot be used to accurately evaluate the heart failure signs of the to-be-tested elderly patient. The reference elderly patient does not have heart failure signs, and has similar underlying diseases, drugs taken and similar effects of the drugs to the to-be-tested elderly patient. The greater the difference between the blood sodium concentration values of the to-be-tested elderly patient and the reference elderly patient at each time point, the more likely the to-be-tested elderly patient has heart failure. Therefore, the difference between the blood sodium concentration values of the to-be-tested elderly patient and the reference elderly patient at each time point can be analyzed to eliminate the interference of the drugs for the underlying diseases on the blood sodium concentration, and the initial heart failure performance value of the to-be-tested elderly patient is combined to evaluate the heart failure performance of the to-be-tested elderly patient, thereby improving the accuracy of the heart failure screening and grading evaluation of the elderly patient.

[0104] Preferably, in one embodiment of the present application, the method for grading the heart failure performance of the to-be-tested elderly patient specifically comprises:

[0105] The blood sodium concentration values of the to-be-tested elderly patient and each reference elderly patient at all time points are sorted in chronological order to obtain the blood sodium concentration sequence of the to-be-tested elderly patient and each reference elderly patient.

[0106] According to the difference between the blood sodium concentration sequences of the to-be-tested elderly patient and each reference elderly patient, and the initial heart failure performance value of the to-be-tested elderly patient, the heart failure performance evaluation value of the to-be-tested elderly patient is obtained. The greater the heart failure performance evaluation value, the more likely the to-be-tested elderly patient has heart failure signs and the more severe the heart failure.

[0107] Preferably, in one embodiment of the present application, the method for grading the heart failure performance of the to-be-tested elderly patient specifically comprises:

[0108] The blood sodium concentration sequences of the to-be-tested elderly patient and each reference elderly patient are input into the dynamic time warping algorithm, and the output minimum cumulative distance is used as the blood sodium concentration difference between the to-be-tested elderly patient and each reference elderly patient. In other embodiments of the present application, the Euclidean distance or Manhattan distance between the blood sodium concentration sequences of the to-be-tested elderly patient and each reference elderly patient can also be used as the blood sodium concentration difference between the to-be-tested elderly patient and each reference elderly patient, which is not limited herein.

[0109] The average difference in serum sodium concentration between the elderly patients being tested and all reference elderly patients was used as the overall variability of the elderly patients being tested. A larger overall variability indicates a greater difference in serum sodium concentration between the elderly patients being tested and the reference elderly patient group, thus suggesting a higher likelihood of heart failure signs in the elderly patients being tested. Therefore, the overall variability of the elderly patients being tested and the initial heart failure manifestation values ​​were combined and normalized to limit the calculation results to a range of values. Within the range, the assessment values ​​of heart failure manifestations in the elderly patients to be tested can be obtained.

[0110] As an example, in one embodiment of the present invention, the expression for the assessment value of heart failure performance in the elderly patient to be tested can be specifically as follows:

[0111]

[0112] in, This represents the assessment value of heart failure in the elderly patient being tested; This represents the initial heart failure performance value of the elderly patient being tested; Indicates the elderly patient to be tested and the first Differences in serum sodium concentration among elderly reference patients; This indicates the overall variability among the elderly patients being tested; This represents the hyperbolic tangent function, used for normalization.

[0113] The higher the heart failure assessment value of the elderly patient being tested, the more likely the elderly patient is to have signs of heart failure and the more severe the condition. Therefore, the heart failure assessment value can be used for grading. If the heart failure assessment value is not greater than the preset first assessment threshold, the elderly patient is classified as a primary heart failure patient. If the heart failure assessment value is greater than the preset first assessment threshold but not greater than the preset second threshold, the elderly patient is classified as a moderate heart failure patient. If the heart failure assessment value is greater than the preset second assessment threshold, the elderly patient is classified as a high-grade heart failure patient. The severity of heart failure increases from primary heart failure patient to moderate heart failure patient to high-grade heart failure patient.

[0114] The preset range of the first evaluation threshold is as follows: The preset range of the second evaluation threshold is: In one embodiment of the present invention, a preset first evaluation threshold is set to 0.3 and a preset second evaluation threshold is set to 0.7. The specific values ​​of the preset first evaluation threshold and the preset second evaluation threshold can also be set by the implementer according to the specific real-time scenario, and are not limited here.

[0115] It is to be noted that the sequential order of the above-described embodiments of the present application only for the purpose of description, but not the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0116] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments.

Claims

1. A system for early screening and grading management of elderly heart failure based on metabolic abnormalities, characterized in that, The system comprises: a data acquisition module, configured to acquire blood sodium concentration values of a to-be-tested elderly patient and each candidate elderly patient without heart failure at each time point, and medication data information of different drug dimensions of the to-be-tested elderly patient and each candidate elderly patient, and simultaneously acquire blood drug concentration values of different drugs of the to-be-tested elderly patient and each candidate elderly patient at each time point after the drugs are taken last time; a preliminary evaluation module, configured to obtain an initial heart failure performance value of the to-be-tested elderly patient according to the blood sodium concentration values of the to-be-tested elderly patient at each time point and the change of the blood sodium concentration values over time; a similarity analysis module, configured to take any one candidate elderly patient as a target candidate elderly patient, obtain a medication information similarity between the to-be-tested elderly patient and the target candidate elderly patient according to the difference of the medication data information of the same drug dimension between the to-be-tested elderly patient and the target candidate elderly patient, obtain a drug efficacy similarity between the to-be-tested elderly patient and the target candidate elderly patient according to the difference of the blood drug concentration values of the same drug between the to-be-tested elderly patient and the target candidate elderly patient at each time point and the time difference after the drugs are taken last time, and combine the medication information similarity and the drug efficacy similarity between the to-be-tested elderly patient and each candidate elderly patient to screen a reference elderly patient of the to-be-tested elderly patient from all the candidate elderly patients; a hierarchical evaluation module, configured to perform hierarchical evaluation on the heart failure performance of the to-be-tested elderly patient according to the difference of the blood sodium concentration values between the to-be-tested elderly patient and the reference elderly patient at each time point and the initial heart failure performance value of the to-be-tested elderly patient.

2. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 1, characterized in that, The obtaining of the initial heart failure performance value of the to-be-tested elderly patient comprises: performing negative correlation mapping on the average value of the blood sodium concentration values of the to-be-tested elderly patient at all time points to obtain a first heart failure performance value of the to-be-tested elderly patient; performing curve fitting on the blood sodium concentration values of the to-be-tested elderly patient at each time point to obtain a blood sodium concentration fitting curve of the to-be-tested elderly patient; obtaining a second heart failure performance value of the to-be-tested elderly patient according to the difference of the slopes of the blood sodium concentration fitting curve at adjacent time points; performing comprehensive processing on the first heart failure performance value and the second heart failure performance value and performing normalization processing to obtain the initial heart failure performance value of the to-be-tested elderly patient.

3. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 2, characterized in that, The obtaining of the second heart failure performance value of the to-be-tested elderly patient comprises: taking any two adjacent time points as an adjacent time point group, and taking the absolute value of the difference of the slopes of the blood sodium concentration fitting curve at the two time points in each adjacent time point group as a blood sodium decline trend value of each adjacent time point group; taking the average value of the blood sodium decline trend values of all adjacent time point groups as the second heart failure performance value of the to-be-tested elderly patient.

4. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 1, characterized in that, The obtaining of the medication information similarity between the to-be-tested elderly patient and the target candidate elderly patient comprises: performing quantitative processing on the medication data information of each drug dimension of the to-be-tested elderly patient and the target candidate elderly patient respectively to obtain a word vector of each drug dimension of the to-be-tested elderly patient and the target candidate elderly patient; The cosine similarity of the word vectors of the same drug dimensions between the to-be-tested elderly patient and the target candidate elderly patient is taken as the initial information similarity of each drug dimension between the to-be-tested elderly patient and the target candidate elderly patient; The average of the initial information similarities of all drug dimensions between the to-be-tested elderly patient and the target candidate elderly patient is normalized to obtain the medication information similarity between the to-be-tested elderly patient and the target candidate elderly patient.

5. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 1, characterized in that, The obtaining of the drug efficacy similarity between the to-be-tested elderly patient and the target candidate elderly patient includes: The absolute value of the difference between the time of the last drug taking of the to-be-tested elderly patient and the target candidate elderly patient is negatively correlated to obtain the drug taking time closeness between the to-be-tested elderly patient and the target candidate elderly patient; The blood drug concentration values of each drug taken by the to-be-tested elderly patient and the target candidate elderly patient at all time points are sorted in time sequence to obtain the blood drug concentration sequence of each drug of the to-be-tested elderly patient and the target candidate elderly patient; The blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient is obtained according to the difference between the blood drug concentration sequences of the same drug between the to-be-tested elderly patient and the target candidate elderly patient; The drug efficacy similarity between the to-be-tested elderly patient and the target candidate elderly patient is obtained by comprehensively combining the drug taking time closeness and the blood drug concentration similarity and then performing normalization processing.

6. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 5, characterized in that, The obtaining of the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient includes: The blood drug concentration sequences of the same drug of the to-be-tested elderly patient and the target candidate elderly patient are input into a dynamic time warping algorithm, and the minimum cumulative distance output is negatively correlated to obtain the blood drug concentration similarity coefficient of each same drug between the to-be-tested elderly patient and the target candidate elderly patient; The average of the blood drug concentration similarity coefficients of all same drugs between the to-be-tested elderly patient and the target candidate elderly patient is taken as the blood drug concentration similarity between the to-be-tested elderly patient and the target candidate elderly patient.

7. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 1, characterized in that, The reference elderly patient of the to-be-tested elderly patient is selected from all candidate elderly patients, and the selecting includes: The disease similarity between the to-be-tested elderly patient and each candidate elderly patient is obtained by comprehensively combining the medication information similarity and the drug efficacy similarity between the to-be-tested elderly patient and each candidate elderly patient and then performing normalization processing; The reference elderly patient of the to-be-tested elderly patient is selected from all candidate elderly patients based on the disease similarity.

8. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 7, characterized in that, The selecting of the reference elderly patient of the to-be-tested elderly patient from all candidate elderly patients based on the disease similarity includes: The candidate elderly patient with the disease similarity greater than a preset similarity threshold is taken as the reference elderly patient of the to-be-tested elderly patient.

9. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 1, characterized in that, The grading assessment of the heart failure performance of the to-be-tested elderly patient includes: The blood sodium concentration values of the to-be-tested elderly patient and each reference elderly patient at all time points are sorted in time sequence to obtain the blood sodium concentration sequence of the to-be-tested elderly patient and each reference elderly patient; According to the difference between the sequence of the blood sodium concentration of the to-be-tested elderly patient and each reference elderly patient, and the initial heart failure performance value of the to-be-tested elderly patient, a heart failure performance evaluation value of the to-be-tested elderly patient is obtained; If the heart failure performance evaluation value is not greater than a preset first evaluation threshold, the to-be-tested elderly patient is classified as a primary heart failure patient; if the heart failure performance evaluation value is greater than the preset first evaluation threshold and not greater than a preset second threshold, the to-be-tested elderly patient is classified as an intermediate heart failure patient; if the heart failure performance evaluation value is greater than the preset second evaluation threshold, the to-be-tested elderly patient is classified as a high-level heart failure patient, wherein the severity of heart failure of the primary heart failure patient, the intermediate heart failure patient and the high-level heart failure patient increases in turn.

10. The system for early screening and grading management of senile heart failure based on metabolic abnormalities according to claim 9, characterized in that, The obtaining of the heart failure performance evaluation value of the to-be-tested elderly patient comprises: The sequence of the blood sodium concentration of the to-be-tested elderly patient and each reference elderly patient is input into a dynamic time warping algorithm, and the minimum cumulative distance output is taken as the blood sodium concentration difference between the to-be-tested elderly patient and each reference elderly patient; The average of the blood sodium concentration difference between the to-be-tested elderly patient and all reference elderly patients is taken as the overall difference of the to-be-tested elderly patient; The overall difference of the to-be-tested elderly patient and the initial heart failure performance value are integrated and normalized to obtain the heart failure performance evaluation value of the to-be-tested elderly patient.

Images