CN114334157A - Cerebral apoplexy risk prediction method and device - Google Patents
Cerebral apoplexy risk prediction method and device Download PDFInfo
- Publication number
- CN114334157A CN114334157A CN202210057344.0A CN202210057344A CN114334157A CN 114334157 A CN114334157 A CN 114334157A CN 202210057344 A CN202210057344 A CN 202210057344A CN 114334157 A CN114334157 A CN 114334157A
- Authority
- CN
- China
- Prior art keywords
- risk
- target user
- risk assessment
- indexes
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 208000006011 Stroke Diseases 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 49
- 206010008190 Cerebrovascular accident Diseases 0.000 title claims abstract description 37
- 230000002490 cerebral effect Effects 0.000 title claims abstract description 37
- 238000012502 risk assessment Methods 0.000 claims abstract description 142
- 238000013058 risk prediction model Methods 0.000 claims abstract description 90
- 238000012545 processing Methods 0.000 claims abstract description 23
- 230000003044 adaptive effect Effects 0.000 claims abstract description 22
- 238000012216 screening Methods 0.000 claims abstract description 17
- 230000036541 health Effects 0.000 claims abstract description 12
- 201000010099 disease Diseases 0.000 claims description 22
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 22
- 230000006399 behavior Effects 0.000 claims description 20
- 238000011156 evaluation Methods 0.000 claims description 19
- 230000035622 drinking Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 8
- 230000000391 smoking effect Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 206010020772 Hypertension Diseases 0.000 claims description 6
- 239000008280 blood Substances 0.000 claims description 6
- 210000004369 blood Anatomy 0.000 claims description 6
- 230000036772 blood pressure Effects 0.000 claims description 6
- 208000026106 cerebrovascular disease Diseases 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 6
- 206010012601 diabetes mellitus Diseases 0.000 claims description 6
- 239000003814 drug Substances 0.000 claims description 6
- 235000006694 eating habits Nutrition 0.000 claims description 6
- 230000036578 sleeping time Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000012353 t test Methods 0.000 claims description 4
- 238000003657 Likelihood-ratio test Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 229940079593 drug Drugs 0.000 claims description 3
- 238000007477 logistic regression Methods 0.000 claims description 3
- 230000001737 promoting effect Effects 0.000 description 5
- 230000006978 adaptation Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 208000019622 heart disease Diseases 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 238000000692 Student's t-test Methods 0.000 description 1
- 235000019504 cigarettes Nutrition 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000037213 diet Effects 0.000 description 1
- 235000005911 diet Nutrition 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000036449 good health Effects 0.000 description 1
- 230000009862 primary prevention Effects 0.000 description 1
- 230000005180 public health Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Abstract
The invention discloses a cerebral apoplexy risk prediction method, which comprises the following steps: according to a risk assessment index system, data information of a target user is collected in a classified mode, and an adaptive risk prediction model is selected according to the data information of the target user; calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user; and comparing the target user risk score with a risk classification threshold value to obtain the risk grade of the target user. By adopting the technical scheme, the method and the device can be used for efficiently and conveniently screening the early-stage risks of the cerebral apoplexy for the target population, and promote the individualized health management.
Description
Technical Field
The invention relates to the technical field of intelligent medical treatment and medical health, in particular to a method and a device for predicting stroke risk.
Background
According to statistics, the morbidity, mortality and disability caused by stroke are steadily increased in the last 20 years. By 2019, allThe globus has 1.01 hundred million cases of cerebral apoplexy, 655 million people die and 1.43 million disabled people with adjusted life-Span (DALYs) caused by cerebral apoplexy, and research shows that the disease burden has a explosive growth situation and presents the rapid growth of low-income groups, obvious gender and regional difference and the trend of youthfulness[2]。
Since 2011, screening and intervention projects for high-risk groups of cerebral apoplexy are brought into national major public health service projects. How to identify the risk factors of the person as the first person in charge of the health is the basis for implementing the autonomous health management. The cerebral apoplexy onset risk assessment refers to the prediction model fitting of risk factors possibly causing cerebral apoplexy and disease onset risks by applying a statistical method, identifies high risk factors causing cerebral apoplexy based on crowd data, quantitatively assesses and grades the individual cerebral apoplexy suffering risk, screens high risk groups, not only provides scientific support for autonomous health management, but also can monitor and assess the treatment and intervention effects, thereby realizing primary prevention of cerebral apoplexy and promoting individuals to maintain good health conditions.
The current risk prediction model has a long prediction period (mostly 10 years) for the stroke attack in the future, and the risk factors influencing the stroke attack such as diet, motion, psychology and the like are less considered, so that the risk prediction model is not beneficial to guiding the health management of users; the triggering condition of the model is relatively fixed, the requirement on index data provided by a user is strict, and the universality of the prediction model is limited.
Disclosure of Invention
In order to solve the problems, the invention provides a method and a device for predicting the risk of stroke, which are used for efficiently and conveniently screening the early risk of stroke of target people and promoting individualized health management.
The invention adopts a technical scheme that: a method of predicting stroke risk, comprising:
collecting data information of a target user in a classified manner according to a risk assessment index system;
carrying out standardization processing on the data information of the target user;
judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, if so, selecting an adaptive risk prediction model according to the data information of the target user;
calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and comparing the target user risk score with a risk classification threshold value to obtain the risk grade of the target user.
Further, the collecting data information of the target user according to the risk assessment index system by classification includes:
preliminarily screening candidate risk evaluation indexes possibly associated with the occurrence of cerebral apoplexy through likelihood ratio test of logistic regression analysis according to data information of the sampling crowd queue;
performing cyclic screening on the candidate risk assessment indexes by a stepwise regression method to determine the risk assessment indexes for constructing a stroke risk prediction model;
generating a risk assessment index system for predicting the risk of the cerebral apoplexy according to the risk assessment index, wherein the risk assessment index system comprises a basic information module, a life behavior module and a disease characteristic module;
and collecting data information of the target user according to the risk assessment index system in a classified manner, wherein the data information of the target user comprises basic information, life behaviors and disease characteristics of the target user.
Further, the pair of basic information modules comprises age, gender, BMI, education level, marital and regional indexes;
the life behavior module comprises smoking, drinking, exercise activities, eating habits, sleeping time and psychological condition indexes;
the disease characteristic module comprises cerebrovascular disease family history, hypertension family history, diabetes history, blood pressure, blood fat, CRP and drug condition indexes.
Further, the risk prediction model comprises a risk assessment index and a weight of the risk assessment index;
the risk prediction model comprises a model I, a model II and a model III;
the risk assessment indexes of the model I comprise the age, sex, BMI, education degree, marital and regional indexes of the basic information module, the smoking and drinking indexes of the life behavior characteristic module and the blood pressure indexes of the disease characteristic module;
the risk assessment indexes of the model II comprise risk assessment indexes of the model I and exercise activity, eating habits, sleeping time and psychological condition indexes of the life behavior characteristic module;
the risk evaluation indexes of the model III comprise risk evaluation indexes of the model II and cerebrovascular disease family history, hypertension family history, diabetes history, blood fat, CRP and medicine taking condition indexes of the disease characteristic module.
Further, the selecting an adaptive risk prediction model according to the data information of the target user includes: when the data information of the target user meets the triggering conditions of the multiple models, the priority selection sequence of the risk prediction model is as follows: model III, model II, model I;
and the triggering condition is used for judging whether the data information of the target user contains all risk evaluation indexes of the risk prediction model I, II or III.
Further, the step-by-step regression method is used for performing cyclic screening on the candidate risk assessment indexes to determine the risk assessment indexes for constructing the stroke risk prediction model, and the method comprises the following steps:
introducing the candidate risk assessment indexes into a risk prediction model one by one, performing F test after introducing a new candidate risk assessment index, verifying whether the new candidate risk assessment index has statistical significance in the risk prediction model, and otherwise, rejecting the new candidate risk assessment index;
and carrying out t test on the candidate risk assessment indexes introduced into the risk prediction model one by one, and rejecting the candidate risk assessment indexes introduced into the risk prediction model when the candidate risk assessment indexes introduced into the risk prediction model lose significance in the risk prediction model due to the introduction of new candidate risk assessment indexes.
Further, according to the standard regression coefficient of each risk assessment index generated in the construction process of the risk prediction model, the weight of the risk assessment index is obtained after classification and transformation according to the data type of the risk assessment index.
Further, the obtaining of the weight of the risk assessment index after classification and transformation according to the data type of the risk assessment index according to the standard regression coefficient of each risk assessment index generated in the construction process of the risk prediction model includes:
when the risk assessment index is a binary variable, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the integer is obtained;
when the risk assessment index is a multi-categorical variable, the weight is 0 when the ranking of the risk assessment index is a reference level; when the grading of the risk assessment index is other levels, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the rounding is carried out;
when the risk assessment index is a numerical variable, the weight is obtained by multiplying the corresponding standard regression coefficient by the median of the risk assessment index, then multiplying by 10, rounding off and rounding up;
the target user risk score is calculated according to the following formula:
wherein S is1As risk score, independent variableX 1,X 2,...,XmAs risk assessment index, coefficientIs the weight of the risk assessment indicator.
Further, the risk prediction model is constructed for a time period of 4 years, 5 years and 7 years.
According to the method, the invention also provides a stroke risk prediction device, which comprises:
the data acquisition module is used for classifying and acquiring data information of the target user according to the risk assessment index system;
the data processing module is used for carrying out standardization processing on the data information of the target user;
the model selection module is used for judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, and if so, an adaptive risk prediction model is selected according to the data information of the target user;
the calculation module is used for calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and the grading module is used for comparing the risk value of the target user with a risk grading threshold value to obtain the risk grade of the target user.
The invention has the beneficial effects that: through the method or the device of the technical scheme, the method and the device provided by the invention can be used for matching an applicable optimal risk prediction model according to the data information of the target user, performing multi-time-period (4/5/7 years), multi-dimensional (basic information, life behaviors and disease characteristics) and multi-level (low/medium/high risk) stroke risk prediction, generating a risk score and a risk grade, and being used for efficiently and conveniently carrying out early stroke risk screening on the target population and promoting individualized health management.
Drawings
Fig. 1 is a flowchart of a method for predicting stroke risk provided by the present patent.
Fig. 2 is a block diagram of a stroke risk prediction device provided in this patent.
Detailed Description
According to the technical scheme provided by the embodiment of the application, data information of a target user is collected in a classified mode according to a risk assessment index system; carrying out standardization processing on the data information of the target user; judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, if so, selecting an adaptive risk prediction model according to the data information of the target user; calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user; and comparing the target user risk score with a risk classification threshold value to obtain the risk grade of the target user. According to the data information of the target user, an applicable optimal risk prediction model can be matched, stroke risk prediction with multiple time periods (4/5/7 years), multiple dimensions (basic information, life behaviors and disease characteristics) and multiple levels (low/medium/high risk) is carried out, and a risk score and a risk grade are generated, so that the method is used for efficiently and conveniently carrying out early stroke risk screening on the target population and promoting individualized health management.
The main implementation principle, the specific implementation mode and the corresponding beneficial effects of the technical scheme of the embodiment of the present application are explained in detail with reference to the accompanying drawings.
Example one
Referring to fig. 1, an embodiment of the present application provides a method for predicting stroke risk, including:
s101, collecting data information of a target user in a classified manner according to a risk assessment index system;
s102, carrying out standardization processing on the data information of the target user;
s103, judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and age of the target user, if so, selecting an adaptive risk prediction model according to the data information of the target user;
s104, calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and S105, comparing the risk score of the target user with a risk classification threshold value to obtain the risk grade of the target user.
The stroke risk prediction method can be applied to intelligent electronic equipment such as a smart phone and the like, wherein the intelligent electronic equipment is provided with a data input part and a data output part and has data processing capacity. The above method may also be applied to servers and computer systems.
When the method is executed on intelligent electronic equipment, a server or a computer system, S101 is executed to collect data information of a target user in a classified mode according to a risk assessment index system.
Specifically, the risk assessment index system is screened according to the following method:
and preliminarily screening candidate risk assessment indexes possibly associated with the occurrence of the cerebral apoplexy through likelihood ratio test of logistic regression analysis according to the data information of the sampling crowd queue.
Performing cyclic screening on the candidate risk assessment indexes by a stepwise regression method to determine the risk assessment indexes for constructing a stroke risk prediction model;
generating a risk assessment index system for predicting the risk of the cerebral apoplexy according to the risk assessment index, wherein the risk assessment index system comprises a basic information module, a life behavior module and a disease characteristic module;
and collecting data information of the target user according to the risk assessment index system in a classified manner, wherein the data information of the target user comprises basic information, life behaviors and disease characteristics of the target user. The data acquisition of the 3 index systems can be performed according to the application scene and the matching degree of the target user with priority.
Performing cyclic screening on the candidate risk assessment indexes by a stepwise regression method to determine the risk assessment indexes for constructing the stroke risk prediction model, wherein the method comprises the following steps:
introducing the candidate risk assessment indexes into a risk prediction model one by one, performing F test after introducing a new candidate risk assessment index, verifying whether the new candidate risk assessment index has statistical significance in the risk prediction model, and otherwise, rejecting the new candidate risk assessment index;
and carrying out t test on the candidate risk assessment indexes introduced into the risk prediction model one by one, and rejecting the candidate risk assessment indexes introduced into the risk prediction model when the candidate risk assessment indexes introduced into the risk prediction model lose significance in the risk prediction model due to the introduction of new candidate risk assessment indexes.
Specifically, as the candidate risk assessment indexes are introduced into the risk prediction model one by one, a certain degree of correlation may exist between the candidate risk assessment indexes, so that the candidate risk assessment indexes introduced into the risk prediction model lose statistical significance in the model, and thus the risk prediction model may include risk assessment indexes having no significant influence on the occurrence of stroke. In order to establish a relatively optimal risk prediction model, the candidate risk assessment indexes are subjected to circular screening by a screening stepwise regression method, and the specific operation comprises the following steps:
introducing the candidate risk assessment indexes into a risk prediction model one by one, and performing F test after introducing a new candidate risk assessment index to verify whether the candidate risk assessment indexes have statistical significance in the risk prediction model; and carrying out t-test on the candidate risk assessment indexes introduced into the risk prediction model one by one, and removing the candidate risk assessment indexes when the candidate risk assessment indexes lose significance in the risk prediction model due to introduction of new candidate risk assessment indexes. Therefore, before introducing a new candidate risk assessment index each time, the risk prediction model only contains the candidate risk assessment index having significant influence on the candidate risk assessment index, and the statistical efficiency of the risk prediction model is ensured. Finally, the risk assessment indexes which are introduced with other candidate risk assessment indexes and have significant influence on the risk prediction model are determined.
And constructing 3 index systems (a basic information module/a life behavior characteristic module/a disease characteristic module) for predicting the risk of the cerebral apoplexy according to the risk assessment indexes. The basic information module comprises age, gender, BMI, education degree, marital and regional indexes; the life behavior module comprises smoking, drinking, exercise activities, eating habits, sleeping time and psychological condition indexes; the disease characteristic module comprises cerebrovascular disease family history, hypertension family history, diabetes history, blood pressure, blood fat, CRP and drug condition indexes.
The following table can be specifically referred to:
TABLE 1 basic information module risk assessment index system of stroke risk prediction model
TABLE 2 Life behavior feature module risk assessment index system of stroke risk prediction model
TABLE 3 Risk assessment index system for disease feature module of stroke risk prediction model
After the data information of the target user is collected in a classified manner, step S102 is executed to perform a standardization process on the data information of the target user.
Specifically, the data of the target user can be classified and standardized according to the data type (two-classification/multiple-classification/numerical value type) and the level setting of the risk assessment index, the level of the risk assessment index is aligned, the weight is assigned according to the level, and the standardized data information can be matched with the corresponding risk prediction model.
After the data information of the target user is standardized, step S103 is executed, whether the cerebral apoplexy risk prediction of the target user can be started is judged based on the cerebral apoplexy history and age of the target user, if so, an adaptive risk prediction model is selected according to the data information of the target user.
Wherein the risk prediction model comprises a risk assessment indicator and a weight of the risk assessment indicator;
the risk prediction model comprises a model I, a model II and a model III;
the risk assessment indexes of the model I comprise the age, sex, BMI, education degree, marital and regional indexes of the basic information module, the smoking and drinking indexes of the life behavior characteristic module and the blood pressure indexes of the disease characteristic module;
the risk assessment indexes of the model II comprise risk assessment indexes of the model I and exercise activity, eating habits, sleeping time and psychological condition indexes of the life behavior characteristic module;
the risk evaluation indexes of the model III comprise risk evaluation indexes of the model II and cerebrovascular disease family history, hypertension family history, diabetes history, blood fat, CRP and medicine taking condition indexes of the disease characteristic module.
The weight of the risk assessment index can be obtained after classification and transformation according to the data type of the risk assessment index according to the standard regression coefficient of each risk assessment index generated in the construction process of the risk prediction model. The specific method for obtaining the weight may be:
when the risk assessment index is a binary variable, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the integer is obtained;
when the risk assessment index is a multi-categorical variable, the weight is 0 when the ranking of the risk assessment index is a reference level; when the grading of the risk assessment index is other levels, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the rounding is carried out;
and when the risk evaluation index is a numerical variable, the weight is obtained by multiplying the corresponding standard regression coefficient by the median of the risk evaluation index, then multiplying by 10, rounding off and rounding.
Adaptation to a risk prediction model. Whether the target user can start the risk prediction of the stroke can be judged according to the stroke history and the age of the target user, and then the adaptive risk prediction model is automatically selected according to the completeness of data information of the target user. When the triggering conditions of a plurality of models are met, the preferential selection sequence of the models is as follows: model III, model II, model I.
For example, if the data information input by the target user simultaneously meets the risk assessment index requirements of the model I, the model II and the model III, starting the model III; if all risk evaluation indexes of the model I and the model II are simultaneously met, starting the model II; and if only all risk assessment indexes of the model I are met, starting the model I. The method ensures that the risk assessment index data can be utilized to the maximum extent to carry out accurate risk prediction, and when the completeness of the index information input by a target user is not enough, flexible adaptive prediction can be carried out, so that the universality of a prediction model is ensured.
The time periods for the above-described risk prediction model construction are 4 years, 5 years and 7 years, and in particular to the present invention,
(1) 4-year risk prediction model: based on China Health and recovery Long national Study (CHARLS) database, baseline is 450 villages and residences in 150 counties and districts in China between 6 months and 3 months in 2011, and 17708 individuals of 10257 family are visited. People with age more than or equal to 45 years who have not been diagnosed with heart disease and/or stroke.
(2) 5-year risk prediction model: the population with physical examination data in 1992, 1994, 1997 in three follow-up visits was selected, with the physical examination data in 1992 as baseline data. Age 55 years old or older, and people who were not diagnosed with heart disease and/or stroke until 1992.
(3) 7-year risk prediction model: at least three physical examination reports are provided between 6 months 2007 and 10 months 2008, and the patients with age more than or equal to 18 years and without heart disease and/or cerebral apoplexy diagnosis are provided.
And after the standard information data and the adaptive risk prediction model are obtained, executing step S104, calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a target user risk score.
The target user risk score is calculated according to the following formula:
wherein S is1As risk score, independent variableX 1,X 2,...,XmAs risk assessment index, coefficientIs the weight of the risk assessment indicator.
After the risk score of the target user is obtained, step S105 is executed, wherein the risk score of the target user is compared with a risk classification threshold value to obtain the risk grade of the target user. Specifically, all risk scores can be arranged from large to small, and different risk grades (low risk/medium risk/high risk) are divided by a five-score method.
The method adopted by the embodiment of the invention can match an applicable optimal risk prediction model according to the data information of the target user, perform stroke risk prediction with multiple time periods (4/5/7 years), multiple dimensions (basic information, life behaviors and disease characteristics) and multiple stages (low/medium/high risk), generate risk values and risk grades, and is used for efficiently and conveniently screening early-stage risks of stroke for the target population and promoting individualized health management.
Example two
Referring to fig. 2, based on the method of the first embodiment, the present invention further provides a stroke risk prediction apparatus, including:
the data acquisition module is used for classifying and acquiring data information of the target user according to the risk assessment index system;
the data processing module is used for carrying out standardization processing on the data information of the target user;
the model selection module is used for judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, and if so, an adaptive risk prediction model is selected according to the data information of the target user;
the calculation module is used for calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and the grading module is used for comparing the risk value of the target user with a risk grading threshold value to obtain the risk grade of the target user.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
EXAMPLE III
By adopting the method and the device of the embodiment, the data information of the sampling crowd queue provided by the cooperative unit is subjected to target user information acquisition, data standardization processing, prediction adaptation model selection and finally risk prediction and risk classification.
Data information of a target user enters the stroke risk prediction system through the target user information acquisition module and then is transmitted to the data standardization processing module;
the data standardization processing module converts the data information of the target user into a level corresponding to a risk evaluation index in a risk evaluation index system, namely a target evaluation parameter, and transmits the level to the prediction adaptation model selection module. For example, "10 cigarettes/week smoking" in the data information of the target user is converted into "smoking frequency: more than 21 times per month, and 1 time of drinking each day, into the corresponding drinking frequency of the risk assessment index: 10-30 times per month, weights are assigned according to this hierarchy.
And the risk prediction and risk classification module matches a risk prediction model (generated in advance) meeting the trigger condition from the risk prediction and risk classification module according to the target evaluation parameters, preferentially selects a model utilizing the target evaluation parameters to the maximum extent to perform risk prediction, calculates a risk value and classifies the risk grade, and transmits the risk value and the classified risk grade to the conclusion sending module. For example, the target evaluation parameters of the target user A (50 years old) satisfy the trigger conditions in the model I and the model II at the same time, and the risk prediction of the stroke in the next 4 years and 7 years can be carried out (the 5-year risk prediction model cannot be used because the age is required to be more than or equal to 55 years old). And triggering the model II according to the setting of the model priority, generating the risk scores of the target user A suffering from the cerebral apoplexy in the next 4 years and 7 years, and generating the risk grade according to the risk grade threshold value of the model II. For example, target user a has a risk score of 2 and 21 for stroke in the next 4 and 7 years, respectively, with corresponding risk ratings of medium risk (incidence of 0.55-1.14%) and higher risk (incidence of 5.25-7.50%), respectively.
And a conclusion sending module generates a risk assessment report according to the risk prediction result and the data information of the target user.
And finishing the whole process.
The verification result shows that the AUC of the model is over 0.7, and the discrimination capability of the prediction model is good. The calibration degree is shown by a result of goodness-of-fit test of the Hosmer-Lemeshow, and P >0.05 shows that the fitting effect of the prediction model is good; the results of bootstrap resampling and 10-fold cross validation show that P is less than 0.001, which indicates that the constructed prediction model is real and reliable.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is only limited by the appended claims
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A method for predicting stroke risk, comprising:
collecting data information of a target user in a classified manner according to a risk assessment index system;
carrying out standardization processing on the data information of the target user;
judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, if so, selecting an adaptive risk prediction model according to the data information of the target user;
calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and comparing the target user risk score with a risk classification threshold value to obtain the risk grade of the target user.
2. The method for predicting stroke risk according to claim 1, wherein the classifying and collecting data information of the target user according to the risk assessment index system comprises:
preliminarily screening candidate risk evaluation indexes possibly associated with the occurrence of cerebral apoplexy through likelihood ratio test of logistic regression analysis according to data information of the sampling crowd queue;
performing cyclic screening on the candidate risk assessment indexes by a stepwise regression method to determine the risk assessment indexes for constructing a stroke risk prediction model;
generating a risk assessment index system for predicting the risk of the cerebral apoplexy according to the risk assessment index, wherein the risk assessment index system comprises a basic information module, a life behavior module and a disease characteristic module;
collecting data information of a target user according to the risk assessment index system in a classified mode, wherein the data information of the target user comprises basic information, life behaviors and disease characteristics of the target user;
the sampling crowd queue comprises crowds contained in a Chinese health and endowment tracking and surveying database and crowds capable of providing specified physical examination data.
3. A stroke risk prediction method as claimed in claim 2, wherein the basic information module includes age, gender, BMI, education level, marital and regional indicators;
the life behavior module comprises smoking, drinking, exercise activities, eating habits, sleeping time and psychological condition indexes;
the disease characteristic module comprises cerebrovascular disease family history, hypertension family history, diabetes history, blood pressure, blood fat, CRP and drug condition indexes.
4. The method of predicting stroke risk according to claim 3,
the risk prediction model comprises a risk assessment indicator and a weight of the risk assessment indicator;
the risk prediction model comprises a model I, a model II and a model III;
the risk assessment indexes of the model I comprise the age, sex, BMI, education degree, marital and regional indexes of the basic information module, the smoking and drinking indexes of the life behavior characteristic module and the blood pressure indexes of the disease characteristic module;
the risk assessment indexes of the model II comprise risk assessment indexes of the model I and exercise activity, eating habits, sleeping time and psychological condition indexes of the life behavior characteristic module;
the risk evaluation indexes of the model III comprise risk evaluation indexes of the model II and cerebrovascular disease family history, hypertension family history, diabetes history, blood fat, CRP and medicine taking condition indexes of the disease characteristic module.
5. The method of predicting stroke risk as set forth in claim 4, wherein the selecting the adapted risk prediction model according to the data information of the target user comprises: when the data information of the target user meets the triggering conditions of the multiple models, the priority selection sequence of the risk prediction model is as follows: model III, model II, model I;
and the triggering condition is used for judging whether the data information of the target user contains all risk evaluation indexes of the risk prediction model I, II or III.
6. The method for predicting stroke risk according to claim 2, wherein the determining the risk assessment index for constructing the stroke risk prediction model by performing a cyclic screening of the candidate risk assessment indexes through a stepwise regression method comprises:
introducing the candidate risk assessment indexes into a risk prediction model one by one, performing F test after introducing a new candidate risk assessment index, verifying whether the new candidate risk assessment index has statistical significance in the risk prediction model, and otherwise, rejecting the new candidate risk assessment index;
and carrying out t test on the candidate risk assessment indexes introduced into the risk prediction model one by one, and rejecting the candidate risk assessment indexes introduced into the risk prediction model when the candidate risk assessment indexes introduced into the risk prediction model lose significance in the risk prediction model due to the introduction of new candidate risk assessment indexes.
7. The method of claim 4, wherein the weight of the risk assessment index is obtained after classification and transformation according to the data type of the risk assessment index based on the standard regression coefficient of each risk assessment index generated during the construction process of the risk prediction model.
8. The method for predicting stroke risk according to claim 7, wherein the obtaining of the weight of the risk assessment index after the classification and transformation according to the data type of the risk assessment index based on the standard regression coefficient of each risk assessment index generated by the risk prediction model in the construction process comprises:
when the risk assessment index is a binary variable, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the integer is obtained;
when the risk assessment index is a multi-categorical variable, the weight is 0 when the ranking of the risk assessment index is a reference level; when the grading of the risk assessment index is other levels, the weight is rounded after the corresponding standard regression coefficient is multiplied by 10, and the rounding is carried out;
when the risk assessment index is a numerical variable, the weight is obtained by multiplying the corresponding standard regression coefficient by the median of the risk assessment index, then multiplying by 10, rounding off and rounding up;
the target user risk score is calculated according to the following formula:
9. A method for stroke risk prediction according to any of claims 1-8, wherein the risk prediction model is constructed for a time period of 4 years, 5 years and 7 years.
10. A stroke risk prediction device, comprising:
the data acquisition module is used for classifying and acquiring data information of the target user according to the risk assessment index system;
the data processing module is used for carrying out standardization processing on the data information of the target user;
the model selection module is used for judging whether the cerebral apoplexy risk prediction of the target user can be started or not based on the cerebral apoplexy history and the age of the target user, and if so, an adaptive risk prediction model is selected according to the data information of the target user;
the calculation module is used for calculating a risk score based on the data information of the target user after the standardization processing and the adaptive risk prediction model to obtain a risk score of the target user;
and the grading module is used for comparing the risk value of the target user with a risk grading threshold value to obtain the risk grade of the target user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210057344.0A CN114334157A (en) | 2022-01-19 | 2022-01-19 | Cerebral apoplexy risk prediction method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210057344.0A CN114334157A (en) | 2022-01-19 | 2022-01-19 | Cerebral apoplexy risk prediction method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114334157A true CN114334157A (en) | 2022-04-12 |
Family
ID=81028444
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210057344.0A Pending CN114334157A (en) | 2022-01-19 | 2022-01-19 | Cerebral apoplexy risk prediction method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114334157A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117457220A (en) * | 2023-12-26 | 2024-01-26 | 深圳市健怡康医疗器械科技有限公司 | Informationized management method and system for neurological stroke patients |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105787252A (en) * | 2016-01-18 | 2016-07-20 | 贡京京 | Medical decision supporting method and system |
CN107273652A (en) * | 2017-03-10 | 2017-10-20 | 马立伟 | Intelligent risk of stroke monitoring system |
CN110782990A (en) * | 2019-10-12 | 2020-02-11 | 广东徕康医疗科技有限公司 | Method, system and mobile platform for primary public defense service stroke risk assessment |
CN113040711A (en) * | 2021-03-03 | 2021-06-29 | 吾征智能技术(北京)有限公司 | Cerebral stroke attack risk prediction system, equipment and storage medium |
-
2022
- 2022-01-19 CN CN202210057344.0A patent/CN114334157A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105787252A (en) * | 2016-01-18 | 2016-07-20 | 贡京京 | Medical decision supporting method and system |
CN107273652A (en) * | 2017-03-10 | 2017-10-20 | 马立伟 | Intelligent risk of stroke monitoring system |
CN110782990A (en) * | 2019-10-12 | 2020-02-11 | 广东徕康医疗科技有限公司 | Method, system and mobile platform for primary public defense service stroke risk assessment |
CN113040711A (en) * | 2021-03-03 | 2021-06-29 | 吾征智能技术(北京)有限公司 | Cerebral stroke attack risk prediction system, equipment and storage medium |
Non-Patent Citations (3)
Title |
---|
吴乐程;鲍德国;: "非瓣膜性心房颤动脑卒中危险因素的临床分析", 健康研究, no. 03, pages 52 - 54 * |
张玲 等: "重庆市社区居民脑卒中危险因素分析及 个体健康危险评价研究", 中国慢性病预防与控制, vol. 16, no. 3, 30 June 2008 (2008-06-30), pages 228 - 232 * |
李娜;伍亚舟;易东;: "基于疾病风险指数的重庆市社区居民脑卒中发病风险评估方法", 第三军医大学学报, no. 20, 30 October 2010 (2010-10-30), pages 70 - 73 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117457220A (en) * | 2023-12-26 | 2024-01-26 | 深圳市健怡康医疗器械科技有限公司 | Informationized management method and system for neurological stroke patients |
CN117457220B (en) * | 2023-12-26 | 2024-04-12 | 深圳市健怡康医疗器械科技有限公司 | Informationized management method and system for neurological stroke patients |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR20170061222A (en) | The method for prediction health data value through generation of health data pattern and the apparatus thereof | |
JP2016520228A (en) | Contextual awareness prediction in medical systems | |
CN107430640A (en) | System and method for generating the anatomy of stress level and compressive resilience level in colony | |
CN105808906B (en) | The analysis method and its device of patient personal characteristics | |
CN110291542B (en) | Factor estimation device, factor estimation system, and factor estimation method | |
CN112641451B (en) | Multi-scale residual error network sleep staging method and system based on single-channel electroencephalogram signal | |
CN116344050B (en) | Evaluation method based on multidimensional health management model | |
CN111370124A (en) | Health analysis system and method based on facial recognition and big data | |
CN115602325A (en) | Chronic disease risk assessment method and system based on multi-model algorithm | |
KR102342770B1 (en) | A health management counseling system using the distribution of predicted disease values | |
CN114334157A (en) | Cerebral apoplexy risk prediction method and device | |
Bhoomika et al. | Ensemble Learning Approaches for Detecting Parkinson's Disease | |
Dhar | An adaptive intelligent diagnostic system to predict early stage of parkinson's disease using two-stage dimension reduction with genetically optimized lightgbm algorithm | |
CN114943629A (en) | Health management and health care service system and health management method thereof | |
Kalogiannis et al. | Geriatric group analysis by clustering non-linearly embedded multi-sensor data | |
CN116598012A (en) | Chronic disease medical care method and system based on federal learning value evaluation | |
CN108831554B (en) | Medical information processing method and device | |
WO2017149542A1 (en) | Neuropsychological evaluation screening system | |
Mole et al. | An efficient Gait Dynamics classification method for Neurodegenerative Diseases using Brain signals | |
US20230260660A1 (en) | Vaccine hesitancy prediction method based on emergency vaccine | |
JP7364248B2 (en) | Health guidance system, health guidance terminal, and health guidance method | |
CN117476217B (en) | Chronic heart disease state of illness trend prediction system | |
CN117373658B (en) | Data processing-based auxiliary diagnosis and treatment system and method for depression | |
Muthulakshmi et al. | Big Data Analytics for Heart Disease Prediction using Regularized Principal and Quadratic Entropy Boosting | |
CN114628033A (en) | Disease risk prediction method, device, equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |