CN114400087A

CN114400087A - Method for diagnosing brucellosis based on inspection data and knowledge map

Info

Publication number: CN114400087A
Application number: CN202210029368.5A
Authority: CN
Inventors: 陈超; 宋彪; 张瑞环; 周睿; 韩泽文; 王哲
Original assignee: Inner Mongolia Weishu Data Technology Co ltd
Current assignee: Inner Mongolia Weishu Data Technology Co ltd
Priority date: 2022-01-12
Filing date: 2022-01-12
Publication date: 2022-04-26

Abstract

The invention discloses a method for diagnosing brucellosis based on test data and a knowledge graph. The overall architecture comprises a data acquisition module, a data analysis module and a data analysis module, wherein the data acquisition module is used for acquiring blood conventional data; and the data processing module is used for cleaning the acquired data. And the model construction module is used for constructing the model by utilizing an integrated learning method. And the model prediction module is used for inputting the samples into the ensemble learning model to obtain a result. And the knowledge graph combination module is used for combining the result of the model prediction module into the knowledge graph to acquire related knowledge. And the result output module is used for inquiring the symptoms of the patient according to the result on the basis of the model prediction module and the knowledge graph spectrum combination module and finally outputting the diseased condition of the patient. The method has the advantages that the method used by the invention uses the routine examination data of the clinical laboratory of the hospital, can assist doctors to accurately diagnose the state of an illness, and effectively reduces the misdiagnosis rate. Meanwhile, the medical nursing bed can play a role in warning diseases of patients and help the patients to prevent related diseases in advance.

Description

Method for diagnosing brucellosis based on inspection data and knowledge map

Technical Field

The invention relates to the field of inspection medicine and the field of disease identification, in particular to a method for diagnosing brucellosis based on inspection data and a knowledge map.

Background

Brucellosis is an infectious disease of both human and veterinary diseases caused by the invasion of bacteria of the genus brucella into the body. At present, domestic brucellosis is mainly caused by cattle and sheep. The brucella carried by sheep has the highest infectivity and pathogenicity, and seriously threatens human health. However, brucellosis has a long incubation period, causes irreparable harm to animals, and causes certain damage to bodies of infected people. The existing detection method is mainly to collect and carry brucellosis samples for detection, and the precondition is that the detection is carried out in a laboratory. Meanwhile, a plurality of detections of one sample may occur, which may result in a long time. Some patients with the disease distribution infection go to the hospital and are not easy to be diagnosed as the disease distribution, and the diagnosis is easy to miss. In order to solve the problems, the invention creatively provides a method for diagnosing brucellosis based on inspection data and a knowledge map, so that a doctor can be assisted to reduce complex processes, and the risk of the doctor for culturing and sickening bacteria is avoided. Meanwhile, the system can help the patient to simply diagnose the state of illness and achieve the prompting function of disease distribution diagnosis for the doctor.

Disclosure of Invention

The invention provides a solution for the problem by adopting a machine learning technology, and aims to provide a method for diagnosing brucellosis based on test data and a knowledge map.

The general architecture created by the invention comprises a data acquisition module, a data processing module, a model construction module, a model prediction module, a knowledge graph and spectrum combination module and a result output module, and the detailed steps of the submodules are introduced below.

And the data acquisition module is used for acquiring conventional data of qualified serum samples collected by a hospital.

The data processing module is used for cleaning the acquired data, and specifically comprises the following steps: screening apparent normal data, dimensionless data and oversampling data.

And a model construction module, which constructs the model by using an ensemble learning method.

And the model prediction module is used for inputting the samples into the constructed model and obtaining a prediction result.

A knowledge-graph association module: the module is combined with a model prediction module, and comprehensive evaluation is carried out by combining physiological data with disease symptom conditions to obtain results.

A result output module: the module is used for inquiring according to relevant symptoms of the disease in the knowledge map on the basis of the model prediction module and the knowledge map combination module, and finally outputting the diseased condition of the patient according to the inquiring result.

The method has the advantages that the method is realized on the basis of the inspection data, and can assist doctors to accurately diagnose the state of illness and effectively reduce the misdiagnosis rate. Meanwhile, the medical nursing bed can play a role in warning diseases of patients and help the patients to prevent related diseases in advance.

Drawings

The invention is further described by using the attached drawings, but the embodiments in the attached drawings do not limit the invention at all, and for those skilled in the art, other attached drawings can be obtained according to the following drawings without creative efforts.

FIG. 1 is a schematic diagram of the present invention.

Detailed Description

The overall architecture flow of the embodiment of the present invention is shown in fig. 1, and the following will describe the technical implementation method of the embodiment of the present invention in detail with reference to fig. 1.

Referring to fig. 1, the method for diagnosing brucellosis based on inspection data and knowledge graph of the present embodiment includes a data acquisition module, a data processing module, a model construction module, a model prediction module, a knowledge graph and spectrum combination module, and a result output module.

The data acquisition module of the embodiment mainly acquires regular data of qualified serum sample bleeding collected by a hospital, and the data has great value not only for the embodiment but also for other related researches. The data in this case are mainly obtained from relevant test data of brucellosis.

Preferably, the data processing module of the embodiment comprises the steps of apparent normal data screening, data dimensionless and data oversampling.

And (4) screening apparent normal data, namely screening the apparent normal data according to the domestic reference range of each index of the blood routine data.

And (3) carrying out non-dimensionalization on the data, converting the samples by a certain proportion by using a standardization method, and enabling the samples to fall into a specific interval, so that each numerical value can be ensured to be in the same quantity level, and the data quality is improved.

And (3) data oversampling, namely oversampling the training data by using a SMOTE algorithm, and amplifying a few samples to the same number of samples as the majority of samples so that the number ratio of the two samples reaches 1: 1. The purpose of this is to achieve a relatively uniform distribution of data.

Preferably, the model building module of this embodiment builds a model by using an ensemble learning method as a basic classifier, before building the model, the data is divided into a training set, a test set and a verification set, then the training set is subjected to ten-fold cross-validation training, after building the model, the model is evaluated and tuned, main parameters of tuning are iteration times and the magnitude of data amount of each iteration, and finally, an optimal model is selected by comparing the obtained accuracy with the result of training in the test set.

Preferably, the model prediction module of the embodiment identifies brucellosis on the basis of the optimal model. The sample identified by the model prediction module is conventional inspection data, the data format and standard are determined according to the function realized by the model, and then the existing optimal model is loaded for identification and the result is output.

Preferably, the knowledge-graph combination module of the present embodiment is combined with the model prediction module, and performs comprehensive evaluation to obtain a result by combining physiological data and disease symptom conditions.

Preferably, the result output module of this embodiment queries, based on the model prediction module and the knowledge graph spectrum combination module, according to the symptoms related to the disease in the knowledge graph spectrum, and finally outputs the diseased condition of the patient according to the queried result. Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. The method for diagnosing brucellosis based on inspection data and knowledge graph is characterized by comprising a data acquisition module, a data processing module, a model construction module, a model prediction module, a knowledge graph and spectrum combination module and a result output module, wherein the contents represented by the modules are as follows, and the data acquisition module acquires conventional bleeding data of qualified serum samples collected by hospitals; the data processing module is used for cleaning the acquired data; the model construction module uses the processed data to construct an ensemble learning model; the model prediction module is used for inputting a sample to be tested into the constructed integrated learning model and obtaining a prediction result; the knowledge map combination module is combined with the model prediction module, and comprehensive evaluation is carried out by combining physiological data with disease symptom conditions to obtain a result; the result output module is used for inquiring according to the relevant symptoms of the disease in the knowledge map spectrum on the basis of the model prediction module and the knowledge map spectrum combination module, and finally outputting the diseased condition of the patient according to the inquiring result; the invention can assist doctors to reduce the complicated process and avoid the risk of doctors suffering from diseases caused by culturing bacteria; meanwhile, the system can help the patient to simply diagnose the illness state of the patient.

2. The method for brucellosis diagnosis based on test data and knowledge-map as claimed in claim 1, wherein the data acquisition module is the test data of hospital clinical laboratory, which has great value not only for the present embodiment but also for other related studies, and the data is mainly the conventional test data related to brucellosis.

3. The method for brucellosis diagnosis based on test data and knowledge-graph as claimed in claim 1, wherein the data processing module is used to wash the acquired data, determine the used dimension according to the implemented function, and analyze the data of the selected dimension.

4. The method for diagnosing brucellosis based on test data and knowledge-maps according to claim 1, wherein brucellosis is identified on the basis of an optimal model, the sample identified by the model prediction module is conventional test data, the data format and standard are determined according to the function realized by the model, and then the existing optimal model is loaded for identification and the result is output.