CN110299207A - For chronic disease detection in based on computer prognosis model data processing method - Google Patents
For chronic disease detection in based on computer prognosis model data processing method Download PDFInfo
- Publication number
- CN110299207A CN110299207A CN201910379450.9A CN201910379450A CN110299207A CN 110299207 A CN110299207 A CN 110299207A CN 201910379450 A CN201910379450 A CN 201910379450A CN 110299207 A CN110299207 A CN 110299207A
- Authority
- CN
- China
- Prior art keywords
- data
- chronic disease
- hypertension
- chronic
- computer
- 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
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention belongs to computer application technologies, disclose a kind of for, based on computer prognosis model data processing system and method, obtaining user's all data in chronic disease detection, the data transmission of acquisition being gone out by internet system；Classifying Sum is carried out to received user's all data, and data are transmitted by internet system；Analysis and prediction is carried out to data according to computer preset model, and result is sent to output module.The present invention is based on computer prognosis models, it completes to the summarizing of user's all data, analyze, arrange, predict, more efficiently self-supervision management is carried out convenient for user itself, convenient for professional person, in summary data provide more accurately health promotion suggestion, the generation of effective preventing hypertension chronic disease simultaneously；The present invention can be improved forecasting efficiency according to hypertension chronic disease predicted value.
Description
Technical Field
The invention belongs to the technical field of computer application, and particularly relates to a computer prediction model-based data processing system and method for chronic disease detection.
Background
The chronic diseases are all called chronic non-infectious diseases, are not specific to a certain disease, but are generalized and general names of diseases which have hidden onset, long course of disease, prolonged illness, lack of exact etiology evidence of infectious organisms, complex etiology and are not completely confirmed. Common chronic diseases mainly include cardiovascular and cerebrovascular diseases, cancer, diabetes and chronic respiratory diseases, wherein the cardiovascular and cerebrovascular diseases comprise hypertension, stroke and coronary heart disease.
Hypertension (hypertension) is a clinical syndrome characterized by an increase in systemic arterial blood pressure (systolic pressure and/or diastolic pressure) (systolic pressure not less than 140 mm hg, diastolic pressure not less than 90 mm hg), which may be accompanied by functional or organic damage to organs such as heart, brain, kidney, etc. Hypertension is the most common chronic disease and also the most major risk factor for cardiovascular and cerebrovascular diseases. The blood pressure of a normal person fluctuates within a certain range along with the changes of internal and external environments. In the whole population, the blood pressure level gradually increases with age, the systolic pressure is more obvious, but the diastolic pressure shows a descending trend after the age of 50, and the pulse pressure is increased. In recent years, people have increasingly deep knowledge on the effects of multiple risk factors of cardiovascular diseases and the protection of target organs of heart, brain and kidney, the diagnosis standard of hypertension is continuously adjusted, and at present, patients with the same blood pressure level are considered to have different cardiovascular disease risks, so that the concept of blood pressure stratification is provided, namely, the patients with different cardiovascular disease risks are different in proper blood pressure level. The evaluation of blood pressure value and risk factor is the main basis for diagnosing and formulating the treatment plan of hypertension, the aim of hypertension management of different patients is different, and doctors judge the most suitable blood pressure range of the patients according to the specific conditions of the patients on the basis of reference standard when facing the patients, and adopt the targeted treatment measures. On the basis of improving the life style, 24-hour long-acting antihypertensive drugs are recommended to control the blood pressure. In addition to assessing office blood pressure, patients should also be aware of home early morning blood pressure monitoring and management to control blood pressure and reduce the incidence of cardiovascular events.
The high-speed development of economy leads to the increase of the living pressure of modern people and poor living habits, leads to the increase of the incidence rate of hypertension chronic diseases, and leads to the increasingly younger disease population, which becomes a serious problem that the health is endangered and the life quality is influenced at present. The conventional prediction of chronic hypertension mainly comprises the steps of classifying the chronic hypertension by an example and mining high-risk factors causing the chronic hypertension by a data mining technology.
In summary, the problems of the prior art are as follows:
(1) the data obtained from the hypertension chronic disease is too complete;
(2) the user data cannot be monitored for a long time, so that the data update is laggard;
(3) chronic hypertensive diseases cannot be predicted in time.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a computer prediction model-based data processing system and method for chronic disease detection.
The invention is realized in such a way that a data processing method based on a computer prediction model for chronic disease detection comprises the following steps:
acquiring various data of a user, and transmitting the acquired data through an Internet system;
classifying and summarizing received user data, and transmitting the data through an Internet system;
and analyzing and predicting the data according to a preset model of the computer, and transmitting the result to an output module.
Further, the method for acquiring various items of data of the user comprises the following steps:
receiving patient identity registration information corresponding to a hospital from a hospital management system, wherein the hospital management system comprises at least one database, and each database stores patient identity information and corresponding illness state information;
receiving patient-related data to be registered from a mobile terminal, wherein the patient-related data comprises monitoring data of identification information of a patient, user age, blood type, height and weight index, familial heredity, eating habits, waist circumference, hip circumference, blood routine, urine routine, renal function, blood sugar, blood fat, blood potassium, echocardiogram, electrocardiogram, chest X-ray, eyeground and dynamic blood pressure;
and after receiving the patient related data request, the server establishes communication with the gateway corresponding to the hospital information.
Further, the method for analyzing and predicting the data by the computer preset model comprises the following steps:
step one, constructing a hypertension chronic disease big data prediction platform containing relational database data, sensor data and controller data based on Hadoop, and turning to step two;
secondly, analyzing and mining in a hypertension chronic disease big data prediction platform by applying an Apriori association rule mining algorithm under a MapReduce framework to obtain hypertension chronic disease influence factors, and turning to the third step;
step three, combining the influence factors of the hypertensive chronic diseases and the historical data of the hypertensive chronic diseases, constructing a neural network model BP, generating an initial weight of the neural network model BP, and turning to step four;
step four, dynamically improving the weight and the threshold of the neural network model BP to obtain a dynamic neural network model DBP, generating the weight and the threshold of the dynamic neural network model DBP, and turning to step five;
fifthly, optimizing a dynamic neural network model DBP by using an Adaptive Immune Genetic Algorithm (AIGA), obtaining a prediction model AIGA-DBP, calculating a prediction value of the chronic hypertension disease according to the prediction model AIGA-DBP, and turning to the sixth step;
step six, judging whether the error between the predicted value of the chronic hypertension disease and the expected value of the chronic hypertension disease meets the set condition, if so, turning to step seven; otherwise, the step five is executed again;
step seven, outputting a predicted value of the chronic hypertension;
step eight: processing the obtained predicted value of the chronic hypertension disease to reduce artifact interference;
step nine: creating a filter, and filtering the processed hypertension chronic disease predicted value to a required frequency band;
step ten: calculating the phase relation between every two channels of the hypertension chronic disease predicted value of each frequency band at each time point by using a phase synchronization analysis method to obtain a dynamic function connection matrix;
step eleven: calculating the time domain entropy of the phase relation value between the two channels one by one to obtain the information entropy of each edge so as to measure the complexity of each edge time domain of the hypertension chronic disease function network;
step twelve: respectively taking the dynamic function connection entropy of each frequency band as the classification characteristic of the hypertension chronic disease function network, training a self-adaptive improvement classifier, and obtaining a plurality of self-adaptive improvement classifiers and corresponding classification accuracy;
step thirteen: and (4) carrying out combined classification on the hypertension chronic diseases in a voting way by utilizing a plurality of trained self-adaptive boosting classifiers.
Further, in step eight, the processing method includes: carrying out 0.5-30Hz band-pass filtering on the collected chronic hypertension disease predicted value data, and then removing an electro-oculogram interference signal and artifact data to obtain a required chronic hypertension disease predicted value;
in the tenth step, the phase locking value PLV is used to calculate the phase relationship between every two channels of the hypertension chronic disease prediction value of each frequency band at each time point, and the specific calculation formula is as follows:
PLV＝|<exp(j{Φi(t)-Φj(t)}))|；
wherein phii(t) and Φj(t) the instantaneous phase of electrodes i and j, respectively;
the phase value of the signal can be calculated by using a hilbert transform, and the specific formula is as follows:
xi(τ) is the continuous time signal of electrode i, τ is a time variable, t represents the time point, and PV is the Cauchy principal value;
the instantaneous phase is calculated as follows:
likewise, canCalculating the instantaneous phase phij(t)；
Setting the number of selected channels for chronic hypertension as M and the number of time points for chronic hypertension as T, constructing different channel pairs by using two channels, and calculating PLV values of all the channel pairs, thereby obtaining an MXM X T three-dimensional matrix K, wherein MXM is an upper triangular matrix of a time point:
each element K of KijtThe PLV value between the ith electrode and the jth electrode at the time point t is the matrix which is a dynamic function connection matrix and not only contains the phase relation between every two different hypertension chronic disease channels, but also contains the spatial information and the time information of the hypertension chronic disease channels;
in the eleventh step, the information entropy of the phase relation value between the two channels, namely the dynamic function connection entropy, is calculated one by one, and the calculation is carried out according to the following steps:
firstly, extracting each PLV value of an M multiplied by M upper triangular matrix of each hypertension chronic disease time point T to obtain a (M multiplied by (M-1)/2) multiplied by T two-dimensional matrix; and then, calculating the information entropy of each edge of the PLV matrix (M x (M-1)/2) multiplied by T to obtain an entropy value matrix (M x (M-1)/2) multiplied by 1.
Further, in the twelfth step, the specific process of obtaining the optimal adaptive boosting classifier includes: for a given hypertensive chronic disease (x)1，y1)，...，(xm，ym) Wherein x isi∈X，yie.Y (-1, 1), X is the training feature, Y is the subject category, and first the weight of each training hypertension chronic disease set is initialized to beThen P iterations are performed, D1(i) Is the weight of each training hypertensive chronic disease set when initializing, i.e. p is 1, and the iterative process is as follows: the variable P is increased from 1 to P, and each weak classifier h is calculated first in each iterationpFor training hypertension chronic diseasesClassification error from line classification
εp＝∑Dp(i)，hp(xi)≠yi，
Wherein h isp(xi) Classification label value, D, obtained for classifying hypertensive chronic disease for the pth weak classifierp(i) Is the weight of each training hypertensive chronic disease set at the p-th iteration, and then the weight of the classification sequence is calculatedFinally, the weight of each training hypertension chronic disease set is updated
Wherein D is+1(i) Is the weight, Z, of each training set after each updatepIn order to normalize the factors, the method comprises the steps of,the weight of the hypertension chronic disease set is adjusted, and when the classification is right, the weight is updatedThe weight of hypertensive chronic disease will decrease; when the classification is misclassified, the weight is updatedThe weight of chronic hypertension will increase;
p weak classifiers h under the frequency band are obtained after P iterations are finishedpAnd finally, combining the P weak classifiers to construct a final classifier which is an optimal self-adaptive improvement classifier:
and then respectively calculating the optimal self-adaption under each frequency band to improve the classification accuracy of the classifier.
In the thirteen step, after combining a plurality of trained self-adaptive improvement classifiers in a voting way, classifying the hypertension chronic diseases:
wherein x isiIs a characteristic of the ith hypertensive chronic disease,is a t-th characteristic of the ith hypertensive chronic disease, wtIs the classification accuracy of the classifier obtained by using the t-th class characteristics, FtIs the classification discrimination of the t-th class feature, F (x)i) The classification accuracy, the true positive rate and the false positive rate of the combination classifier are calculated according to the original label after the classification result of each hypertensive chronic disease combination is obtained by the output of the ith hypertensive chronic disease combination classifier.
Another object of the present invention is to provide a computer-based prediction model data processing program for use in chronic disease detection, which implements the computer-based prediction model data processing method for use in chronic disease detection.
Another object of the present invention is to provide a terminal having a processor for implementing the computer-based predictive model data processing method for chronic disease detection.
It is another object of the present invention to provide a computer-readable storage medium including instructions which, when executed on a computer, cause the computer to perform the computer-predictive model-based data processing method for use in chronic disease detection.
Another object of the present invention is to provide a computer-based prediction model data processing system for use in chronic disease detection, which implements the computer-based prediction model data processing method for use in chronic disease detection, the computer-based prediction model data processing system for use in chronic disease detection, comprising:
the data acquisition module acquires various data of the user and transmits the data to the data storage module and the data analysis module through the Internet system;
the data storage module is used for storing various data transmitted by the data acquisition module, preliminarily summarizing, classifying and sorting the various data, and transmitting the sorted data to the data analysis module through the Internet system;
the data analysis module is used for analyzing the data of the data storage module according to a preset model of a computer, obtaining a prediction result and outputting the prediction result to the output module through an internet system;
and the output module is used for outputting the prediction result transmitted by the data analysis module.
Further, the data acquisition module is connected with the mobile phone client, the portable detection equipment and the physical examination report module to acquire various data of the user;
the physical examination report module is used for acquiring data of blood routine, urine routine, renal function, blood sugar, blood fat, blood potassium, echocardiogram, electrocardiogram, chest X-ray, eyeground and dynamic blood pressure monitoring;
the portable detection equipment comprises a height and weight measuring instrument, a sphygmomanometer and a blood glucose meter.
In summary, the advantages and positive effects of the invention are:
the invention is based on a computer prediction model, completes the collection, analysis, arrangement and prediction of various data of the user, is convenient for the user to carry out more effective self-supervision and management, is convenient for professionals to give more accurate health promotion suggestions by integrating the data, and effectively prevents the occurrence of hypertension chronic diseases.
The method comprises the steps of firstly constructing a hypertension chronic disease big data prediction platform, then excavating hypertension chronic disease influence factors by using an association rule algorithm, constructing a neural network model BP, dynamically improving the weight and the threshold of the neural network model BP to obtain a dynamic neural network model DBP, then optimizing the dynamic neural network model DBP by using a self-adaptive immune genetic AIGA algorithm to obtain a prediction model AIGA-DBP, finally calculating a hypertension chronic disease prediction value by using the prediction model AIGA-DBP, and improving the prediction efficiency according to the hypertension chronic disease prediction value.
The dynamic neural network model DBP in the invention can adapt to various changes caused by time lapse.
According to the invention, a big data analysis technology is applied, so that the excavation of the influence factors of the chronic hypertension is more efficient and accurate, the influence factors of the chronic hypertension are more comprehensively considered, and the prediction accuracy is effectively improved.
The method processes the acquired predicted value of the chronic hypertension disease to reduce artifact interference; creating a filter, and filtering the processed hypertension chronic disease predicted value to a required frequency band; calculating the phase relation between every two channels of the hypertension chronic disease predicted value of each frequency band at each time point by using a phase synchronization analysis method to obtain a dynamic function connection matrix; calculating the time domain entropy of the phase relation value between the two channels one by one to obtain the information entropy of each edge so as to measure the complexity of each edge time domain of the hypertension chronic disease function network;
respectively taking the dynamic function connection entropy of each frequency band as the classification characteristic of the hypertension chronic disease function network, training a self-adaptive improvement classifier, and obtaining a plurality of self-adaptive improvement classifiers and corresponding classification accuracy;
and (4) carrying out combined classification on the hypertension chronic diseases in a voting way by utilizing a plurality of trained self-adaptive boosting classifiers. Accurate hypertension chronic disease prediction data information can be obtained.
Drawings
Fig. 1 is a schematic diagram of a data processing system based on computer prediction model for chronic disease detection according to an embodiment of the present invention.
In the figure: 1. a data acquisition module; 2. a data storage module; 3. a data analysis module; 4. an output module; 5. a mobile phone client; 6. a portable detection device; 7. a physical examination report module.
Fig. 2 is a flow diagram of a computer-based predictive model data processing system and method for use in chronic disease detection.
FIG. 3 is a flowchart of a method for analyzing and predicting data by using a computer default model according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The acquired data is too one-sided;
the user data cannot be monitored for a long time, so that the data update is laggard;
chronic hypertensive diseases cannot be predicted in time.
In order to solve the above technical problems, the following detailed description is provided for the application principle of the present invention with reference to specific embodiments.
As shown in fig. 1, a computer-based predictive model data processing system for chronic disease detection according to an embodiment of the present invention includes:
the data acquisition module 1 acquires various data of a user and transmits the data to the data storage module and the data analysis module through the Internet system;
the data storage module 2 is used for storing various data transmitted by the data acquisition module, preliminarily summarizing, classifying and sorting the various data, and transmitting the sorted data to the data analysis module through the Internet system;
the data analysis module 3 is used for analyzing the data of the data storage module according to a preset model of a computer, obtaining a prediction result and outputting the prediction result to the output module through an internet system;
and the output module 4 outputs the prediction result transmitted by the data analysis module.
In the embodiment of the present invention, the data acquisition module 1 provided by the present invention includes:
the mobile phone client 5 acquires data such as user age, blood type, height and weight index, family inheritance, eating habits and the like;
a portable detection device 6, including a height and weight measuring instrument, a tape, a sphygmomanometer and a blood glucose meter; acquiring data such as height, weight, waist circumference, hip circumference, family heredity, eating habits and the like;
the physical examination report module 7 is used for acquiring data such as blood routine, urine routine (including protein, sugar and urinary sediment microscopic examination), renal function, blood sugar, blood fat, blood potassium, echocardiogram, electrocardiogram, chest X-ray, eyeground, dynamic blood pressure monitoring and the like.
As shown in fig. 2, a computer-based prediction model data processing method for chronic disease detection according to an embodiment of the present invention includes:
s101: the data acquisition module acquires various data of the user through the mobile phone client, the portable detection equipment and the physical examination report, and transmits the acquired data to the data storage module through the Internet system;
s102: the data storage module classifies and summarizes the received user data and transmits the data to the data analysis module through the Internet system;
s103: the data analysis module analyzes and predicts the data according to a preset model of the computer and transmits the result to the output module;
s104: the output module outputs the prediction result.
As shown in fig. 3, in the embodiment of the present invention, a method for analyzing and predicting data by using a computer predetermined model includes:
s301, constructing a hypertension chronic disease big data prediction platform containing relational database data, sensor data and controller data based on Hadoop, and turning to the second step;
s302, analyzing and mining in a hypertension chronic disease big data prediction platform by using an Apriori association rule mining algorithm under a MapReduce framework to obtain hypertension chronic disease influence factors, and turning to the third step;
s303, establishing a neural network model BP by combining the influence factors of the chronic hypertension diseases and the historical data of the chronic hypertension diseases, generating an initial weight of the neural network model BP, and turning to the fourth step;
s304, dynamically improving the weight and the threshold of the neural network model BP to obtain a dynamic neural network model DBP, generating the weight and the threshold of the dynamic neural network model DBP, and turning to the fifth step;
s305, optimizing a dynamic neural network model DBP by using an Adaptive Immune Genetic Algorithm (AIGA), obtaining a prediction model AIGA-DBP, calculating a prediction value of the chronic hypertension disease according to the prediction model AIGA-DBP, and turning to the sixth step;
s306, judging whether the error between the predicted value and the expected value of the chronic hypertension disease meets the set condition, if so, turning to the seventh step; otherwise, the step five is executed again;
s307, outputting a hypertension chronic disease prediction value;
s308, processing the acquired chronic hypertension disease prediction value to reduce artifact interference;
s309, creating a filter, and filtering the processed hypertension chronic disease predicted value to a required frequency band;
s310, calculating the phase relation between every two channels of the hypertension chronic disease predicted value of each frequency band at each time point by using a phase synchronization analysis method to obtain a dynamic function connection matrix;
s311, calculating the time domain entropy of the phase relation value between the two channels one by one to obtain the information entropy of each edge so as to measure the complexity of each edge time domain of the hypertension chronic disease function network;
s312, training a self-adaptive improvement classifier by using the dynamic function connection entropy of each frequency band as the classification characteristic of the hypertension chronic disease function network respectively to obtain a plurality of self-adaptive improvement classifiers and corresponding classification accuracy;
and S313, performing combined classification on the hypertension chronic diseases in a voting mode by using a plurality of trained adaptive boosting classifiers.
In the embodiment of the present invention, in step eight, the processing method includes: carrying out 0.5-30Hz band-pass filtering on the collected chronic hypertension disease predicted value data, and then removing an electro-oculogram interference signal and artifact data to obtain a required chronic hypertension disease predicted value;
in the tenth step, the phase locking value PLV is used to calculate the phase relationship between every two channels of the hypertension chronic disease prediction value of each frequency band at each time point, and the specific calculation formula is as follows:
PLV＝|<exp(j{Φi(t)-Φj(t)})>|；
wherein phii(t) and Φj(t) the instantaneous phase of electrodes i and j, respectively;
the phase value of the signal can be calculated by using a hilbert transform, and the specific formula is as follows:
xi(τ) is the continuous time signal of electrode i, τ is a time variable, t represents the time point, and PV is the Cauchy principal value;
the instantaneous phase is calculated as follows:
likewise, the instantaneous phase Φ can be calculatedj(t)；
Setting the number of selected channels for chronic hypertension as M and the number of time points for chronic hypertension as T, constructing different channel pairs by using two channels, and calculating PLV values of all the channel pairs, thereby obtaining an MXM X T three-dimensional matrix K, wherein MXM is an upper triangular matrix of a time point:
each element K of KijtThe PLV value between the ith electrode and the jth electrode at the time point t is the matrix which is a dynamic function connection matrix and not only contains the phase relation between every two different hypertension chronic disease channels, but also contains the spatial information and the time information of the hypertension chronic disease channels;
in the eleventh step, the information entropy of the phase relation value between the two channels, namely the dynamic function connection entropy, is calculated one by one, and the calculation is carried out according to the following steps:
firstly, extracting each PLV value of an M multiplied by M upper triangular matrix of each hypertension chronic disease time point T to obtain a (M multiplied by (M-1)/2) multiplied by T two-dimensional matrix; and then, calculating the information entropy of each edge of the PLV matrix (M x (M-1)/2) multiplied by T to obtain an entropy value matrix (M x (M-1)/2) multiplied by 1.
In the embodiment of the present invention, in the twelfth step, the specific process of obtaining the optimal adaptive boosting classifier includes: for a given hypertensive chronic disease (x)1，y1)，...，(xm，ym) Wherein x isi∈X，yie.Y (-1, 1), X is the training feature, Y is the subject category, and first the weight of each training hypertension chronic disease set is initialized to beThen P iterations are performed, D1(i) Is the weight of each training hypertensive chronic disease set when initializing, i.e. p is 1, and the iterative process is as follows: the variable P is increased from 1 to P, and each weak classifier h is calculated first in each iterationpClassifying error epsilon obtained by classifying training hypertension chronic disease setp＝∑Dp(i)，hp(xi)≠yi，
Wherein h isp(xi) Classification label value, D, obtained for classifying hypertensive chronic disease for the pth weak classifierp(i) Is the weight of each training hypertensive chronic disease set at the p-th iteration, and then the weight of the classification sequence is calculatedFinally, the weight of each training hypertension chronic disease set is updatedWherein D is+1(i) Is the weight, Z, of each training set after each updatepIn order to normalize the factors, the method comprises the steps of,the weight of the hypertension chronic disease set is adjusted, and when the classification is right, the weight is updatedThe weight of hypertensive chronic disease will decrease; when the classification is misclassified, the weight is updatedThe weight of chronic hypertension will increase;
p weak classifiers h under the frequency band are obtained after P iterations are finishedpAnd finally, combining the P weak classifiers to construct a final classifier which is an optimal self-adaptive improvement classifier:
and then respectively calculating the optimal self-adaption under each frequency band to improve the classification accuracy of the classifier.
In the thirteen step, after combining a plurality of trained self-adaptive improvement classifiers in a voting way, classifying the hypertension chronic diseases:
wherein x isiIs a characteristic of the ith hypertensive chronic disease,is a t-th characteristic of the ith hypertensive chronic disease, wtIs the classification accuracy of the classifier obtained by using the t-th class characteristics, FtIs the classification discrimination of the t-th class feature, F (x)i) The classification accuracy, the true positive rate and the false positive rate of the combination classifier are calculated according to the original label after the classification result of each hypertensive chronic disease combination is obtained by the output of the ith hypertensive chronic disease combination classifier.
In the embodiment of the present invention, the first step specifically includes the following steps:
uploading the relational database data, the sensor data and the controller data to a distributed file system (HDFS) through Sqoop, and storing the data in a NoSQL database; and mining and analyzing the relational database data, the sensor data and the controller data by using a MapReduce calculation framework, writing the analyzed data into a NoSQL database, and displaying the data through Web.
In step S2, the mining algorithm using Apriori association rules under the MapReduce framework specifically includes the following steps:
s201, obtaining a set L of frequent 1 item sets by using a MapReduce calculation model1Generating a set C of candidate k-term setsk(k≥2)；
S202, in the Map function processing stage, each Map task calculates that each transaction record in the transaction data set processed by the Map task is contained in CkIf a certain item set (containing k items) of the candidate k item set appears in a transaction record for each Map task, the Map function generates and outputs<A certain set of items, 1>The key value pair is given to a Combiner function, processed by the Combiner function and then given to a Reduce function;
s203, in the Reduce function processing stage, the Reduce function accumulates CkThe number of occurrences of the item set in (1) is obtained as the support frequency of all the item sets, and all the item sets with the support frequency more than or equal to the set minimum support frequency form a frequent item set LkIf k is less than the maximum iteration number and is not empty, executing k + +, and going to step S202; otherwise, ending the operation.
The method for generating the initial weight of the neural network model BP in step S3 is to randomly select an initial weight between intervals [ -1, 1 ];
in the embodiment of the present invention, the step four specifically includes the following steps:
s401, adjusting weight w between a neural network model BP hidden layer and an output layerkj；
Adjusting wkjIs intended to output a new output o of node j* pjIs more than the current output opjCloser to the target value tpjDefining:
where α represents closeness, remains unchanged at each training period, and becomes smaller as the number of hidden layer nodes H is adjusted, regardless of the threshold, there are:
wherein wkjAnd w* kjWeight before and after updating, ypkFor hidden layer output,. DELTA.wkjIs wkjThe amount of change of (d);
obtaining Δ w according to the following formulakjThe solution equation of (c):
wherein,
solving equation according to least squares sum error principleObtaining Δ wkjApproximate solution of (2):
for each hidden layer node k connected to an output node j, calculating the weight change Deltaw between k and jkjUpdating the weight value and calculating the square sum error E, and then belonging to [1, H ] at k]Selecting an optimal k from the interval to minimize E;
s402, adjusting weight v between the BP input layer and the hidden layer of the neural network modelik；
Adjustment vikThe purpose is that once the neural network algorithm falls into a local minimum point, the modified weight can jump out of the minimum point, and the condition that the neural network algorithm falls into the local minimum point is judged to be the change of an error EThe chemical conversion rate Delta E is 0, and E>0；
Regardless of the threshold, the change in the weights of the hidden layer node k is solved by the following equation:
wherein deltapj＝f-1(ypk+Δypk)-f-1(ypk) M is a natural number, then the hidden layer outputs ypkThe solving formula is as follows:
wherein Δ ypkIs ypkThe change amount of (c) is:
solving formula according to least squares sum error principleThe matrix equation can be constructed to calculate:
calculating the dynamic average change of weight between hidden layer and output layer
Calculating the dynamic average change of the weight between the input layer and the hidden layer
And M is a natural number between 10 and 20, a dynamic average weight of the neural network model BP is obtained, and a dynamic neural network model DBP is obtained according to the dynamic average weight of the neural network model BP.
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 and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A computer-based prediction model data processing method for chronic disease detection is characterized by comprising the following steps:
acquiring various data of a user, and transmitting the acquired data through an Internet system;
classifying and summarizing received user data, and transmitting the data through an Internet system;
and analyzing and predicting the data according to a preset model of the computer, and transmitting the result to an output module.
2. The computer-based predictive model data processing method for use in chronic disease detection as set forth in claim 1, wherein the method of obtaining user item data includes:
receiving patient identity registration information corresponding to a hospital from a hospital management system, wherein the hospital management system comprises at least one database, and each database stores patient identity information and corresponding illness state information;
receiving patient-related data to be registered from a mobile terminal, wherein the patient-related data comprises monitoring data of identification information of a patient, user age, blood type, height and weight index, familial heredity, eating habits, waist circumference, hip circumference, blood routine, urine routine, renal function, blood sugar, blood fat, blood potassium, echocardiogram, electrocardiogram, chest X-ray, eyeground and dynamic blood pressure;
and after receiving the patient related data request, the server establishes communication with the gateway corresponding to the hospital information.
3. The computer-based predictive model data processing method for chronic disease detection as set forth in claim 1, wherein the method for analyzing and predicting data by the computer predictive model comprises:
step one, constructing a hypertension chronic disease big data prediction platform containing relational database data, sensor data and controller data based on Hadoop, and turning to step two;
secondly, analyzing and mining in a hypertension chronic disease big data prediction platform by applying an Apriori association rule mining algorithm under a MapReduce framework to obtain hypertension chronic disease influence factors, and turning to the third step;
step three, combining the influence factors of the hypertensive chronic diseases and the historical data of the hypertensive chronic diseases, constructing a neural network model BP, generating an initial weight of the neural network model BP, and turning to step four;
step four, dynamically improving the weight and the threshold of the neural network model BP to obtain a dynamic neural network model DBP, generating the weight and the threshold of the dynamic neural network model DBP, and turning to step five;
fifthly, optimizing a dynamic neural network model DBP by using an Adaptive Immune Genetic Algorithm (AIGA), obtaining a prediction model AIGA-DBP, calculating a prediction value of the chronic hypertension disease according to the prediction model AIGA-DBP, and turning to the sixth step;
step six, judging whether the error between the predicted value of the chronic hypertension disease and the expected value of the chronic hypertension disease meets the set condition, if so, turning to step seven; otherwise, the step five is executed again;
step seven, outputting a predicted value of the chronic hypertension;
step eight: processing the obtained predicted value of the chronic hypertension disease to reduce artifact interference;
step nine: creating a filter, and filtering the processed hypertension chronic disease predicted value to a required frequency band;
step ten: calculating the phase relation between every two channels of the hypertension chronic disease predicted value of each frequency band at each time point by using a phase synchronization analysis method to obtain a dynamic function connection matrix;
step eleven: calculating the time domain entropy of the phase relation value between the two channels one by one to obtain the information entropy of each edge so as to measure the complexity of each edge time domain of the hypertension chronic disease function network;
step twelve: respectively taking the dynamic function connection entropy of each frequency band as the classification characteristic of the hypertension chronic disease function network, training a self-adaptive improvement classifier, and obtaining a plurality of self-adaptive improvement classifiers and corresponding classification accuracy;
step thirteen: and (4) carrying out combined classification on the hypertension chronic diseases in a voting way by utilizing a plurality of trained self-adaptive boosting classifiers.
4. The computer-based predictive model data processing method for use in chronic disease detection as claimed in claim 3, wherein in step eight, the processing method includes: carrying out 0.5-30Hz band-pass filtering on the collected chronic hypertension disease predicted value data, and then removing an electro-oculogram interference signal and artifact data to obtain a required chronic hypertension disease predicted value;
in the tenth step, the phase locking value PLV is used to calculate the phase relationship between every two channels of the hypertension chronic disease prediction value of each frequency band at each time point, and the specific calculation formula is as follows:
PLV＝|<exp(j{Φi(t)-Φj(t)})〉|；
wherein phii(t) and Φj(t) the instantaneous phase of electrodes i and j, respectively;
the phase value of the signal can be calculated by using a hilbert transform, and the specific formula is as follows:
xi(τ) is the continuous time signal of electrode i, τ is a time variable, t represents the time point, and PV is the Cauchy principal value;
the instantaneous phase is calculated as follows:
likewise, the instantaneous phase Φ can be calculatedj(t)；
Setting the number of selected channels for chronic hypertension as M and the number of time points for chronic hypertension as T, constructing different channel pairs by using two channels, and calculating PLV values of all the channel pairs, thereby obtaining an MXM X T three-dimensional matrix K, wherein MXM is an upper triangular matrix of a time point:
each element K of KijtThe PLV value between the ith electrode and the jth electrode at the time point t is the matrix which is a dynamic function connection matrix and not only contains the phase relation between every two different hypertension chronic disease channels, but also contains the spatial information and the time information of the hypertension chronic disease channels;
in the eleventh step, the information entropy of the phase relation value between the two channels, namely the dynamic function connection entropy, is calculated one by one, and the calculation is carried out according to the following steps:
firstly, extracting each PLV value of an M multiplied by M upper triangular matrix of each hypertension chronic disease time point T to obtain a (M multiplied by (M-1)/2) multiplied by T two-dimensional matrix; and then, calculating the information entropy of each edge of the PLV matrix (M x (M-1)/2) multiplied by T to obtain an entropy value matrix (M x (M-1)/2) multiplied by 1.
5. The computer-based predictive model data processing method for chronic disease detection as set forth in claim 3, wherein the specific process of obtaining the optimal adaptive boosting classifier in the twelfth step includes: for a given hypertensive chronic disease (x)1，y1)，...，(xm，ym) Wherein x isi∈X，yie.Y (-1, 1), X is the training feature, Y is the subject category, and first the weight of each training hypertension chronic disease set is initialized to beThen P iterations are performed, D1(i) Is the weight of each training hypertensive chronic disease set when initializing, i.e. p is 1, and the iterative process is as follows: the variable P is increased from 1 to P, and each weak classifier h is calculated first in each iterationpClassifying error epsilon obtained by classifying training hypertension chronic disease setp＝∑Dp(i)，hp(xi)≠yi，
Wherein h isp(xi) Classification label value, D, obtained for classifying hypertensive chronic disease for the pth weak classifierp(i) Is the weight of each training hypertensive chronic disease set at the p-th iteration, and then the weight of the classification sequence is calculatedFinally, the weight of each training hypertension chronic disease set is updatedWherein D is+1(i) Is the weight, Z, of each training set after each updatepIn order to normalize the factors, the method comprises the steps of,the weight of the hypertension chronic disease set is adjusted, and when the classification is right, the weight is updatedThe weight of hypertensive chronic disease will decrease; when the classification is misclassified, the weight is updatedThe weight of chronic hypertension will increase;
p weak classifiers h under the frequency band are obtained after P iterations are finishedpAnd finally, combining the P weak classifiers to construct a final classifier which is an optimal self-adaptive improvement classifier:
then respectively calculating the optimal self-adaption under each frequency band to improve the classification accuracy of the classifier;
in the thirteen step, after combining a plurality of trained self-adaptive improvement classifiers in a voting way, classifying the hypertension chronic diseases:
wherein x isiIs a characteristic of the ith hypertensive chronic disease,is a t-th characteristic of the ith hypertensive chronic disease, wtIs the classification accuracy of the classifier obtained by using the t-th class characteristics, FtIs the classification discrimination of the t-th class feature, F (x)i) The classification accuracy, the true positive rate and the false positive rate of the combination classifier are calculated according to the original label after the classification result of each hypertensive chronic disease combination is obtained by the output of the ith hypertensive chronic disease combination classifier.
6. A computer-based prediction model data processing program for chronic disease detection, which implements the computer-based prediction model data processing method for chronic disease detection according to any one of claims 1 to 5.
7. A terminal, characterized in that the terminal is provided with a processor for realizing the computer prediction model-based data processing method for chronic disease detection according to any one of claims 1 to 5.
8. A computer-readable storage medium comprising instructions which, when executed on a computer, cause the computer to perform the computer-predictive model-based data processing method for use in chronic disease detection according to any one of claims 1 to 5.
9. A computer-based predictive model data processing system for use in chronic disease detection implementing the computer-based predictive model data processing method for use in chronic disease detection of claim 1, the computer-based predictive model data processing system for use in chronic disease detection comprising:
the data acquisition module acquires various data of the user and transmits the data to the data storage module and the data analysis module through the Internet system;
the data storage module is used for storing various data transmitted by the data acquisition module, preliminarily summarizing, classifying and sorting the various data, and transmitting the sorted data to the data analysis module through the Internet system;
the data analysis module is used for analyzing the data of the data storage module according to a preset model of a computer, obtaining a prediction result and outputting the prediction result to the output module through an internet system;
and the output module is used for outputting the prediction result transmitted by the data analysis module.
10. The computer-based predictive model data processing system for chronic disease detection as recited in claim 9, wherein the data collection module is connected to the mobile phone client, the portable detection device, and the physical examination report module for obtaining various data of the user;
the physical examination report module is used for acquiring data of blood routine, urine routine, renal function, blood sugar, blood fat, blood potassium, echocardiogram, electrocardiogram, chest X-ray, eyeground and dynamic blood pressure monitoring;
the portable detection equipment comprises a height and weight measuring instrument, a sphygmomanometer and a blood glucose meter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910379450.9A CN110299207A (en) | 2019-05-08 | 2019-05-08 | For chronic disease detection in based on computer prognosis model data processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910379450.9A CN110299207A (en) | 2019-05-08 | 2019-05-08 | For chronic disease detection in based on computer prognosis model data processing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110299207A true CN110299207A (en) | 2019-10-01 |
Family
ID=68026690
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910379450.9A Pending CN110299207A (en) | 2019-05-08 | 2019-05-08 | For chronic disease detection in based on computer prognosis model data processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110299207A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111128380A (en) * | 2019-12-24 | 2020-05-08 | 郑州科技学院 | Method and system for constructing chronic disease health management model for simulating doctor diagnosis and accurate intervention strategy |
| CN112037910A (en) * | 2020-08-31 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Health information management method, device, equipment and storage medium |
| CN112270994A (en) * | 2020-10-14 | 2021-01-26 | 中国医学科学院阜外医院 | Construction method, device, terminal and storage medium of a risk prediction model |
| CN112599252A (en) * | 2021-03-04 | 2021-04-02 | 北京冠新医卫软件科技有限公司 | Chronic medical record feature classification processing method and device and terminal |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636479A (en) * | 2015-02-15 | 2015-05-20 | 西安电子科技大学 | Industrial big data driven total completion time prediction method |
| CN108717875A (en) * | 2018-05-18 | 2018-10-30 | 贵州大学 | A kind of chronic disease intelligent management system based on big data |
| CN108735279A (en) * | 2018-06-21 | 2018-11-02 | 广西虚拟现实科技有限公司 | A kind of virtual reality headstroke rehabilitation training of upper limbs system and control method |
| CN108764043A (en) * | 2018-04-26 | 2018-11-06 | 太原理工大学 | The brain electricity sorting technique of entropy based on dynamic function connection |
| US20200160147A1 (en) * | 2017-08-02 | 2020-05-21 | National Institute For Materials Science | Human brain like intelligent decision-making machine |
-
2019
- 2019-05-08 CN CN201910379450.9A patent/CN110299207A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636479A (en) * | 2015-02-15 | 2015-05-20 | 西安电子科技大学 | Industrial big data driven total completion time prediction method |
| US20200160147A1 (en) * | 2017-08-02 | 2020-05-21 | National Institute For Materials Science | Human brain like intelligent decision-making machine |
| CN108764043A (en) * | 2018-04-26 | 2018-11-06 | 太原理工大学 | The brain electricity sorting technique of entropy based on dynamic function connection |
| CN108717875A (en) * | 2018-05-18 | 2018-10-30 | 贵州大学 | A kind of chronic disease intelligent management system based on big data |
| CN108735279A (en) * | 2018-06-21 | 2018-11-02 | 广西虚拟现实科技有限公司 | A kind of virtual reality headstroke rehabilitation training of upper limbs system and control method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111128380A (en) * | 2019-12-24 | 2020-05-08 | 郑州科技学院 | Method and system for constructing chronic disease health management model for simulating doctor diagnosis and accurate intervention strategy |
| CN112037910A (en) * | 2020-08-31 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Health information management method, device, equipment and storage medium |
| CN112037910B (en) * | 2020-08-31 | 2023-04-18 | 康键信息技术（深圳）有限公司 | Health information management method, device, equipment and storage medium |
| CN112270994A (en) * | 2020-10-14 | 2021-01-26 | 中国医学科学院阜外医院 | Construction method, device, terminal and storage medium of a risk prediction model |
| CN112599252A (en) * | 2021-03-04 | 2021-04-02 | 北京冠新医卫软件科技有限公司 | Chronic medical record feature classification processing method and device and terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11017902B2 (en) | System and method for processing human related data including physiological signals to make context aware decisions with distributed machine learning at edge and cloud | |
| CN104287706B (en) | A kind of health status real time monitor and diagnosis and treatment commending system | |
| CN109009017B (en) | An intelligent health monitoring system and data processing method thereof | |
| US20060025931A1 (en) | Method and apparatus for real time predictive modeling for chronically ill patients | |
| US20220122735A1 (en) | System and method for processing human related data including physiological signals to make context aware decisions with distributed machine learning at edge and cloud | |
| Tervonen et al. | Personalized mental stress detection with self-organizing map: From laboratory to the field | |
| CN110299207A (en) | For chronic disease detection in based on computer prognosis model data processing method | |
| Pal et al. | Deep learning techniques for prediction and diagnosis of diabetes mellitus | |
| CN111095232A (en) | Exploring genomes for use in machine learning techniques | |
| Chen et al. | Edge2Analysis: a novel AIoT platform for atrial fibrillation recognition and detection | |
| KR102342770B1 (en) | A health management counseling system using the distribution of predicted disease values | |
| US20240206821A1 (en) | Apparatus and a method for predicting a physiological indicator | |
| CN112768074A (en) | Artificial intelligence-based serious disease risk prediction method and system | |
| CN105808906A (en) | Method for analyzing individual characteristics of patient and apparatus therefor | |
| EP4224373A1 (en) | System for forecasting a mental state of a subject and method | |
| Paviglianiti et al. | VITAL-ECG: A de-bias algorithm embedded in a gender-immune device | |
| EP1810197A2 (en) | Method and apparatus for real time predictive modeling for chronically ill patients | |
| US8428965B2 (en) | System for clinical research and clinical management of cardiovascular risk using ambulatory blood pressure monitoring and actigraphy | |
| KR20210112041A (en) | Smart Healthcare Monitoring System and Method for Heart Disease Prediction Based On Ensemble Deep Learning and Feature Fusion | |
| JP2013148996A (en) | Seriousness determination device, and seriousness determination method | |
| Dileep et al. | IGHOA Based Modified Convolutional Neural Network for Prediction of Cardiovascular Disease | |
| Chen et al. | Heterogeneous sensing and predictive modeling of postoperative outcomes | |
| Rajmohan et al. | G-Sep: A deep learning algorithm for detection of long-term sepsis using bidirectional gated recurrent unit | |
| Kotkar et al. | An IoT enabled healthcare framework for arrhythmia detection based on Qos aware trust aided osprey routing protocol and ensemble learning | |
| Malik et al. | Heart Disease Prediction Using Artificial Intelligence |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191001 |