An intelligent follow-up method and system for home heart failure patients

Abstract

The present application relates to the technical field of physiological data processing, in particular to an intelligent follow-up method and system for home heart failure patients. The method maps the physiological data vector of each follow-up into a sample space, clusters in the sample space, compares the distance between the clusters, obtains the recovery degree of the patient under real-time follow-up, adjusts the time interval between the previous follow-up in combination with the psychological state of the patient, and obtains the time interval of the next follow-up. In the next follow-up of the patient, the state of the patient is evaluated, the follow-up priority is determined, and the follow-up plan is set according to the follow-up priority. The present application quantifies the physiological state change of the patient between different follow-up processes in the sample space, obtains the time interval of the next follow-up, and sets a scientific follow-up plan by using the follow-up priority.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of physiological data processing, in particular to an intelligent follow-up method and system for home heart failure patients. BACKGROUND

[0002] Due to the characteristics of home heart failure patients, in order to effectively monitor and treat the patients, it is necessary to arrange a follow-up plan for the patients. The common follow-up plan is usually regular follow-up, that is, a fixed follow-up cycle is set for follow-up. However, this follow-up plan is mainly set according to experience, and the use of the same follow-up plan for different patients may lead to that the changes of the disease state of some patients cannot be found in time, affecting the recovery of the patients. SUMMARY

[0003] In order to solve the technical problem that the regular follow-up follow-up plan is easy to lead to that the disease state of the patient cannot be found in time in the prior art, the purpose of the present application is to provide an intelligent follow-up method and system for home heart failure patients, and the technical scheme adopted is as follows:

[0004] The present application provides an intelligent follow-up method for home heart failure patients, which comprises:

[0005] For each home heart failure patient, the physiological data of each follow-up record is composed into a physiological data vector, the physiological data vector is mapped into a sample space of the patient, and a physiological data sample point of each follow-up is obtained, and the sample space further comprises a preset normal physiological data sample point;

[0006] The sample points in the sample space are clustered to obtain a plurality of clustering clusters; a standard clustering cluster is screened out according to the distribution of the normal physiological data sample point; the clustering cluster in which the physiological data sample point of the real-time follow-up is located is a target clustering cluster, and the clustering cluster in which the physiological data sample point of the previous follow-up of the real-time follow-up is located is a comparison clustering cluster; the recovery degree of the patient under the real-time follow-up is obtained according to the distance distribution between the target clustering cluster, the comparison clustering cluster and the standard clustering cluster;

[0007] The psychological state of the patient under the real-time follow-up is obtained, the time interval between the real-time follow-up and the previous follow-up is adjusted according to the recovery degree and the psychological state, and the time interval of the next follow-up of each patient is obtained;

[0008] If there are a plurality of patients in the next follow-up, the follow-up priority of each patient is obtained according to the time interval of each patient under a preset follow-up number and the distance between the target clustering cluster and the standard clustering cluster; and the priority follow-up is performed according to the follow-up priority.

[0009] Further, the physiological data includes weight, blood pressure, heart rate and renal function data, and after obtaining the physiological data, the physiological data in each dimension is normalized by using a range standardization algorithm.

[0010] Further, the sample points in the sample space are clustered by using a DBSCAN density clustering algorithm to obtain a plurality of clustering clusters.

[0011] Further, the standard clustering cluster is the clustering cluster with the most normal physiological data sample points.

[0012] Further, the recovery degree obtaining method comprises:

[0013] obtaining a clustering center point of each clustering cluster;

[0014] the distance between the clustering center points of the target clustering cluster and the standard clustering cluster as a first distance, the distance between the clustering center points of the comparison clustering cluster and the standard clustering cluster as a second distance, and the distance between the clustering center points of the target clustering cluster and the comparison clustering cluster as a comparison recovery state;

[0015] the difference between the second distance and the first distance as a state difference value, and multiplying the comparison recovery state after normalization processing by the state difference value to obtain the recovery degree.

[0016] Further, the psychological state obtaining method comprises:

[0017] during the follow-up process, the psychological state of the patient is scored by using a PHQ-9 psychological disorder screening scale, and the scoring result is taken as the psychological state.

[0018] Further, the time interval of the next follow-up of each patient is obtained, comprising:

[0019] taking the time interval between the real-time follow-up and the previous follow-up as an initial time interval, normalizing the recovery degree to obtain a first adjustment coefficient, negatively correlating and normalizing the psychological state to obtain a second adjustment coefficient, and taking the product of the first adjustment coefficient and the second adjustment coefficient as a third adjustment coefficient;

[0020] if the third adjustment coefficient is less than a preset coefficient threshold, the difference between the third adjustment coefficient and the preset coefficient threshold is taken as a fourth adjustment coefficient, the product of the fourth adjustment coefficient and the initial time interval is taken as an adjustment amount, and the difference between the initial time interval and the adjustment amount is taken as the time interval of the next follow-up of each patient;

[0021] if the third adjustment coefficient is greater than the preset coefficient threshold, after obtaining the adjustment amount, the sum of the initial time interval and the adjustment amount is taken as the time interval of the next follow-up of each patient.

[0022] Further, the follow-up priority acquisition method comprises:

[0023] For each patient, the reciprocal of the accumulated sum of the time intervals under the preset follow-up times is taken as a time weight; and the first distance corresponding to the real-time follow-up is normalized and multiplied by the time weight to obtain the follow-up priority.

[0024] Further, the cluster center point is a centroid point in the cluster.

[0025] The application further provides an intelligent follow-up system for a home heart failure patient, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the steps of any one of the intelligent follow-up methods for the home heart failure patient when executing the computer program.

[0026] The application has the following beneficial effects:

[0027] In order to analyze the change of the patient's condition state in the follow-up process, the physiological data generated by each follow-up are recorded and mapped in the sample space of the patient, and the sample space contains normal physiological data sample points, so that the sample points generated by the patient follow-up can be compared. Further, by comparing the cluster between different follow-up times and comparing the cluster generated by the follow-up with the standard cluster, the recovery degree of the patient in the real-time follow-up is determined, that is, the higher the recovery degree, the less the high-frequency follow-up is needed. The time interval of the next follow-up can be determined by combining the physiological state of the patient. Further, considering that the patient may appear overlapping in the follow-up plan, after determining the time interval of the next follow-up, the follow-up priority of the patient is further obtained for the follow-up plan of multiple patients, and the follow-up based on the follow-up priority of the patient can realize scientific and effective follow-up plan arrangement, and the condition state of the patient can be found in time. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0029] Figure 1 A flow chart of an intelligent follow-up method for a home heart failure patient provided by an embodiment of the application. DETAILED DESCRIPTION

[0030] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific implementation, structure, features and effects of the intelligent follow-up method and system for home heart failure patients according to the present application are described in detail below in combination with 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.

[0031] 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 the present application belongs.

[0032] The specific scheme of the intelligent follow-up method and system for home heart failure patients provided by the present application is specifically described below in combination with the drawings.

[0033] Please refer to Figure 1 which shows the flow chart of the intelligent follow-up method for home heart failure patients provided by one embodiment of the present application, and the method comprises:

[0034] Step S1: For each home heart failure patient, the physiological data recorded at each follow-up is composed into a physiological data vector, the physiological data vector is mapped into the sample space of the patient, and the physiological data sample point of each follow-up is obtained. The sample space also includes a preset normal physiological data sample point.

[0035] The embodiment of the present application aims to statistically record the physiological data of the patient in the follow-up process and develop a follow-up plan by evaluating the recovery state and other characteristics. Therefore, the multi-dimensional physiological data generated by each patient in each follow-up process needs to be recorded and statistically analyzed.

[0036] Preferably, in one embodiment of the present application, the physiological data includes body weight, blood pressure, heart rate and renal function data. Different dimensions of physiological data have different data ranges and dimensions, so in order to ensure the operation in the subsequent data processing and sample point mapping process, the obtained data needs to be de-dimensioned and normalized. Therefore, after obtaining the physiological data, the physiological data of each dimension is normalized by using the range standardization algorithm. It should be noted that the process of range standardization for normalization is a technical means familiar to those skilled in the art, which is not described here.

[0037] After obtaining the physiological data in each dimension, the physiological data vector can be formed. Because the heart failure patients generate various physiological data in the follow-up process, and because each person has different physical conditions, the threshold cannot be directly used to measure the good or bad of the patient's condition in each follow-up process, therefore, the embodiment of the present application aims to evaluate the recovery ability of the patient by observing the changes of the physiological data vector between different follow-up times. In order to evaluate the changes of the physiological data vector, the physiological data vector generated in each follow-up is mapped into the sample space of the patient, forming the physiological data sample point of each follow-up, so that the distance between the sample points can represent the changes of the physiological data of the patient between different follow-up processes. At the same time, the normal physiological data sample points are also preset in the sample space, which can play a comparison role and highlight the changes of the physiological data of the patient between different follow-up times.

[0038] In the embodiment of the present application, the number of normal physiological data sample points is set to 30, and the method of constructing the sample space is a technical means familiar to those skilled in the art, and will not be described in detail. Each patient corresponds to a sample space.

[0039] Step S2: clustering the sample points in the sample space to obtain a plurality of clustering clusters; screening a standard clustering cluster according to the distribution of the normal physiological data sample points; the clustering cluster in which the physiological data sample point of the real-time follow-up is located is the target clustering cluster, and the clustering cluster in which the physiological data sample point of the previous follow-up of the real-time follow-up is located is the comparison clustering cluster; obtaining the recovery degree of the patient in the real-time follow-up according to the distance distribution between the target clustering cluster, the comparison clustering cluster and the standard clustering cluster.

[0040] Because there are many sample points in the sample space, and the physiological condition of the patient is changing, the physiological data sample point generated in the follow-up will not move in a fixed trend along with the execution of the follow-up, that is, the distribution of the sample points in the sample space is more complex and chaotic, therefore, in order to improve the accuracy of data analysis, the sample points in the sample space are clustered to obtain a plurality of clustering clusters, that is, each clustering cluster represents a physiological condition, and the difference between different physiological conditions can be represented by comparing the distance between the clustering clusters.

[0041] In an embodiment of the present application, the neighborhood radius is set to 5, the minimum point set number is set to 3, and the DBSCAN density clustering algorithm is used to cluster the sample points in the sample space to obtain a plurality of clustering clusters.

[0042] The embodiment of the present application screens a standard clustering cluster according to the distribution of the normal physiological data sample points, that is, the standard clustering cluster can be used as comparison information to represent a normal physiological condition. In an embodiment of the present application, the standard clustering cluster is the clustering cluster in which the normal physiological data sample points are the most.

[0043] Further, the cluster of the physiological data sample point of the real-time follow-up is taken as a target cluster, and the cluster of the physiological data sample point of the previous follow-up is taken as a contrast cluster. That is, the recovery degree of the patient in the real-time follow-up can be obtained according to the distance distribution among the target cluster, the contrast cluster and the standard cluster. For example, the farther the distance between the target cluster and the contrast cluster, the more obvious the change of the physiological state of the patient between the two follow-ups. If the distance between the target cluster and the standard cluster is closer, it indicates that the change of the physiological state of the patient is positive, that is, the recovery degree is greater.

[0044] Preferably, in an embodiment of the present application, the recovery degree obtaining method comprises:

[0045] The cluster center point of each cluster is obtained, that is, the cluster center point is used to represent the position of the cluster.

[0046] The distance between the cluster center points of the target cluster and the standard cluster is taken as a first distance, and the distance between the cluster center points of the contrast cluster and the standard cluster is taken as a second distance. The greater the first distance or the second distance, the closer the physiological state of the patient in the corresponding follow-up process to the normal physiological state, that is, the healthier the patient.

[0047] The distance between the cluster center points of the target cluster and the contrast cluster is taken as a contrast recovery state. The greater the distance between the two, the greater the change of the physiological state of the patient between the two follow-up processes, and the greater the contrast recovery state.

[0048] The difference between the second distance and the first distance is taken as a state difference value, and the contrast recovery state is multiplied by the state difference value after normalization processing to obtain the recovery degree. If the state difference value is positive and greater, it indicates that the physiological state of the patient in the real-time follow-up process is closer to the normal state than in the previous follow-up, and the physiological state of the patient is positively recovered. On the contrary, it indicates that the physiological state of the patient has not changed or has become worse, and the physiological state of the patient has not recovered or even deteriorated. That is, when the state difference value is positive and greater, the greater the contrast recovery state, the more obvious the recovery of the patient, and the greater the recovery degree.

[0049] It should be noted that the normalization method can also be processed by range standardization, and other conventional mathematical means can also be used by those skilled in the art to achieve the purpose of normalization, which will not be described and limited here.

[0050] In an embodiment of the present application, the cluster center point is the centroid of the cluster.

[0051] Step S3: obtaining the psychological state of the patient under real-time follow-up, adjusting the time interval between the real-time follow-up and the previous follow-up according to the recovery degree and the psychological state, and obtaining the time interval of the next follow-up of each patient.

[0052] For home patients, not only the physiological state needs to be concerned, but also the psychological state needs to be concerned because of the physical condition and the environment. The worse the psychological state of the patient is, the more the physiological recovery will be affected by the psychological disease. Therefore, the psychological state of the patient needs to be obtained during the follow-up process. The worse the psychological state is, the more the time interval of the next follow-up process needs to be reduced to avoid high-frequency follow-up. Further combined with the recovery degree, the greater the recovery degree is, the less the high-frequency follow-up is needed, and the time interval of the next follow-up process can be increased.

[0053] Preferably, in an embodiment of the present application, the method for obtaining the psychological state comprises:

[0054] During the follow-up process, the psychological state of the patient is scored by using the PHQ-9 psychological disorder screening scale, and the scoring result is taken as the psychological state. It should be noted that the PHQ-9 psychological disorder screening scale is a depression scoring table, and the psychological state can be quantified by scoring. The higher the score is, the worse the psychological state is.

[0055] Further, in an embodiment of the present application, the time interval of the next follow-up of each patient is obtained, comprising:

[0056] The time interval between the real-time follow-up and the previous follow-up is taken as an initial time interval. The recovery degree is normalized to obtain a first adjustment coefficient. The negative correlation of the psychological state is mapped and normalized to obtain a second adjustment coefficient. The product of the first adjustment coefficient and the second adjustment coefficient is taken as a third adjustment coefficient. It should be noted that because the recovery degree and the follow-up time interval are positively correlated, and the psychological state is negatively correlated, the quantification result of the psychological state needs to be negatively correlated, normalized and combined to obtain the third adjustment coefficient which is also a coefficient with a value range of 0 to 1. In the embodiment of the present application, the negative number of the psychological state is taken as the power of the exponential function with the natural constant as the base to obtain the negative correlation mapping and normalization result.

[0057] If the third adjustment coefficient is less than a preset coefficient threshold, it indicates that the recovery degree of the patient is poor, and the psychological state score is large, and the psychological problem is serious, and the follow-up time interval needs to be shortened. Therefore, the difference between the third adjustment coefficient and the preset coefficient threshold is taken as a fourth adjustment coefficient, the product of the fourth adjustment coefficient and the initial time interval is taken as an adjustment amount, and the difference between the initial time interval and the adjustment amount is taken as the time interval of the next follow-up of each patient.

[0058] Conversely, if the third adjustment coefficient is greater than the preset coefficient threshold, it indicates that the psychological state score of the patient is smaller, the psychological state is better and the recovery degree is larger, and a longer time interval can be set, so that the sum of the adjustment amount and the initial time interval is used as the time interval of the patient in the next follow-up.

[0059] It should be noted that the coefficient threshold can be set according to the analysis of the doctor, and the embodiment of the present application can be set to 0.6, which is not limited and described here.

[0060] Step S4: If there are multiple patients in the next follow-up, the follow-up priority of each patient is obtained according to the time interval of each patient at the preset follow-up times and the distance between the target cluster and the standard cluster. The follow-up priority is used for priority follow-up.

[0061] In step S3, the follow-up time interval of each patient in the next follow-up is obtained, and the time of the next follow-up of each patient can be determined according to the obtained time interval. However, there can be multiple patients for the same follow-up time, and the states of different patients can be different, so if there are multiple patients in the next follow-up, the embodiment of the present application further analyzes the time interval of each patient at the preset follow-up times. If the time interval in several follow-ups is relatively small, it indicates that the state of the patient is relatively urgent and needs to be followed up more preferentially. And the distance between the target cluster and the standard cluster can also be considered as the physiological state of the standard patient, so the follow-up priority of each patient can be obtained based on the two characteristics, that is, the smaller the time interval at the preset follow-up times, and the greater the distance between the target cluster and the standard cluster, the greater the follow-up priority of the patient. Based on the follow-up priority, the patients to be followed up next time can be sorted in order, and the follow-up is performed in order according to the priority, so that the scientific and efficient follow-up plan arrangement is realized.

[0062] Preferably, in an embodiment of the present application, the follow-up priority acquisition method comprises:

[0063] For each patient, the reciprocal of the sum of the time intervals at the preset follow-up times is used as the time weight, and the smaller the sum of the time intervals, the worse the state of the patient, and the greater the time weight. The first distance corresponding to the real-time follow-up is normalized and multiplied by the time weight to obtain the follow-up priority. In the embodiment of the present application, the preset follow-up times are set to 6, that is, the time weight is obtained by counting the time interval between the last 6 follow-ups.

[0064] In summary, the embodiment of the present application maps the physiological data vector of each follow-up into a sample space, clusters in the sample space, compares the distance between the clusters, obtains the recovery degree of the patient under real-time follow-up, adjusts the time interval between the previous follow-up in combination with the psychological state of the patient, obtains the time interval of the next follow-up, evaluates the state of the patient in the next follow-up, determines the follow-up priority, and sets the follow-up plan according to the follow-up priority. The present application quantifies the physiological state change of the patient between different follow-up processes in the sample space, obtains the time interval of the next follow-up, and sets a scientific follow-up plan by using the follow-up priority.

[0065] Based on the same inventive concept, the present application further provides an intelligent follow-up system for a heart failure patient at home, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of any one of the intelligent follow-up methods for a heart failure patient at home when executing the computer program.

[0066] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent 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, multitasking and parallel processing are possible or can be advantageous.

[0067] 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 mainly describes the difference from other embodiments.

Claims

1. An intelligent follow-up method for home heart failure patients, characterized in that, The method comprises: For each home heart failure patient, a physiological data vector is formed by physiological data recorded at each follow-up visit, the physiological data vector is mapped into a sample space of the patient, a physiological data sample point at each follow-up visit is obtained, and the sample space further comprises a preset normal physiological data sample point; The sample points in the sample space are clustered to obtain a plurality of clustering clusters; a standard clustering cluster is screened out according to the distribution of the normal physiological data sample point; a clustering cluster in which a physiological data sample point at a real-time follow-up visit is located is a target clustering cluster, and a clustering cluster in which a physiological data sample point at a previous follow-up visit of the real-time follow-up visit is located is a comparison clustering cluster; a recovery degree of the patient at the real-time follow-up visit is obtained according to a distance distribution between the target clustering cluster, the comparison clustering cluster and the standard clustering cluster; A psychological state of the patient at the real-time follow-up visit is obtained, the recovery degree and the psychological state are used to adjust a time interval between the real-time follow-up visit and the previous follow-up visit, and a time interval of a next follow-up visit of each patient is obtained; If there are a plurality of patients at the next follow-up visit, a follow-up priority of each patient is obtained according to the time interval of each patient at a preset follow-up visit and a distance between the target clustering cluster and the standard clustering cluster; and the follow-up priority is used for priority follow-up; The recovery degree is obtained by the following method: A clustering center point of each clustering cluster is obtained; A distance between the clustering center points of the target clustering cluster and the standard clustering cluster is a first distance, a distance between the clustering center points of the comparison clustering cluster and the standard clustering cluster is a second distance, and a distance between the clustering center points of the target clustering cluster and the comparison clustering cluster is a comparison recovery state; A difference between the second distance and the first distance is used as a state difference, and the comparison recovery state is multiplied by the state difference after normalization processing to obtain the recovery degree; The psychological state is obtained by the following method: During the follow-up visit, a PHQ-9 psychological disorder screening scale is used to score the psychological state of the patient, and the score result is used as the psychological state; The time interval of the next follow-up visit of each patient is obtained by the following method: An initial time interval between the real-time follow-up visit and the previous follow-up visit is obtained, the recovery degree is normalized to obtain a first adjustment coefficient, the psychological state is negatively correlated and normalized to obtain a second adjustment coefficient, and a product of the first adjustment coefficient and the second adjustment coefficient is used as a third adjustment coefficient; If the third adjustment coefficient is less than a preset coefficient threshold, a difference between the third adjustment coefficient and the preset coefficient threshold is used as a fourth adjustment coefficient, a product of the fourth adjustment coefficient and the initial time interval is used as an adjustment amount, and a difference between the initial time interval and the adjustment amount is used as the time interval of the next follow-up visit of each patient; If the third adjustment coefficient is greater than the preset coefficient threshold, a sum of the initial time interval and the adjustment amount is used as the time interval of the next follow-up visit of each patient after the adjustment amount is obtained. 2.The intelligent follow-up method for home heart failure patients according to claim 1, characterized in that, The physiological data comprises body weight, blood pressure, heart rate and renal function data, and a range standardization algorithm is used to normalize the physiological data in each dimension after the physiological data is obtained. 3.The intelligent follow-up method for home heart failure patients according to claim 1, characterized in that, The sample points in the sample space are clustered by using a DBSCAN density clustering algorithm to obtain a plurality of clustering clusters. 4.The intelligent follow-up method for home heart failure patients according to claim 1, characterized in that, The standard clustering cluster is a clustering cluster in which the number of normal physiological data sample points is the largest. 5.The intelligent follow-up method for home heart failure patients according to claim 1, characterized in that, The method for obtaining the follow-up priority comprises: For each patient, the reciprocal of the accumulated sum of the time intervals under the preset follow-up number is taken as a time weight; and the first distance corresponding to the real-time follow-up is multiplied by the time weight after normalization to obtain the follow-up priority. 6.The intelligent follow-up method for home heart failure patients according to claim 1, characterized in that, The clustering center point is a centroid point in a clustering cluster.

7. An intelligent follow-up system for home heart failure patients, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The processor implements the steps of the intelligent follow-up method for home heart failure patients according to any one of claims 1-6 when executing the computer program.

Images