CN116579503B - 5G intelligent hospital basic data processing method and database platform - Google Patents
5G intelligent hospital basic data processing method and database platform Download PDFInfo
- Publication number
- CN116579503B CN116579503B CN202310827489.9A CN202310827489A CN116579503B CN 116579503 B CN116579503 B CN 116579503B CN 202310827489 A CN202310827489 A CN 202310827489A CN 116579503 B CN116579503 B CN 116579503B
- Authority
- CN
- China
- Prior art keywords
- data
- model
- index
- hospital
- representing
- 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.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 52
- 238000012545 processing Methods 0.000 claims abstract description 89
- 238000004364 calculation method Methods 0.000 claims abstract description 73
- 238000009826 distribution Methods 0.000 claims abstract description 59
- 238000007726 management method Methods 0.000 claims abstract description 57
- 238000005457 optimization Methods 0.000 claims abstract description 57
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000012795 verification Methods 0.000 claims abstract description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 78
- 238000000605 extraction Methods 0.000 claims description 49
- 238000004458 analytical method Methods 0.000 claims description 34
- 238000003066 decision tree Methods 0.000 claims description 30
- 230000006870 function Effects 0.000 claims description 28
- 238000007781 pre-processing Methods 0.000 claims description 28
- 230000004044 response Effects 0.000 claims description 28
- 230000000694 effects Effects 0.000 claims description 25
- 238000012549 training Methods 0.000 claims description 19
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 238000007477 logistic regression Methods 0.000 claims description 11
- 238000000611 regression analysis Methods 0.000 claims description 11
- 238000004140 cleaning Methods 0.000 claims description 10
- 238000000546 chi-square test Methods 0.000 claims description 9
- 238000013075 data extraction Methods 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 8
- 230000005540 biological transmission Effects 0.000 description 7
- 230000006872 improvement Effects 0.000 description 7
- 230000008569 process Effects 0.000 description 5
- 238000013459 approach Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000013178 mathematical model Methods 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000036772 blood pressure Effects 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000013468 resource allocation Methods 0.000 description 2
- 230000000452 restraining effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- 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
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W36/00—Hand-off or reselection arrangements
- H04W36/14—Reselecting a network or an air interface
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/50—Allocation or scheduling criteria for wireless resources
- H04W72/54—Allocation or scheduling criteria for wireless resources based on quality criteria
- H04W72/542—Allocation or scheduling criteria for wireless resources based on quality criteria using measured or perceived quality
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention relates to the technical field of intelligent medical treatment, in particular to a 5G intelligent hospital basic data processing method and a database platform. The method comprises the following steps: acquiring hospital basic data, analyzing the hospital basic data by utilizing JSON based on the hospital basic data, and generating API direct connection interface data; performing cross index verification by using a scene optimization processing formula based on the API direct connection interface data to generate task type distribution index data; performing distribution calculation by using a 5G network slice based on task type distribution index data to generate a management index data set; according to the invention, the hospital basic data is subjected to data processing, and the 5G network slicing technology is used for combining the artificial intelligent model to perform data processing on the hospital basic data, so that the network delay in the data processing is reduced, the processing efficiency of the hospital basic data is improved, and the bandwidth required by the data processing is reduced.
Description
Technical Field
The invention relates to the technical field of intelligent medical treatment, in particular to a 5G intelligent hospital basic data processing method and a database platform.
Background
With the rapid development of 5G technology and the rising of intelligent hospitals, the efficiency and accuracy of processing medical data become particularly important, and traditional hospital data processing methods often rely on manual operation or simple network transmission, but when facing large-scale medical data and complex data processing requirements, the problems of high network transmission delay, low data processing efficiency and high bandwidth requirements are encountered, so how to develop a 5G intelligent hospital basic data processing method with low network delay, high efficiency and low bandwidth becomes a problem to be solved.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a 5G intelligent hospital base data processing method and database platform to solve at least one of the above-mentioned problems.
In order to achieve the above purpose, the present invention provides a 5G intelligent hospital basic data processing method, which includes the following steps:
step S1: acquiring hospital basic data, analyzing the hospital basic data by utilizing JSON based on the hospital basic data, and generating API direct connection interface data;
step S2: performing cross index verification by using a scene optimization processing formula based on the API direct connection interface data to generate task type distribution index data;
Step S3: performing distribution calculation by using a 5G network slice based on task type distribution index data to generate a management index data set;
step S4: performing feature extraction by using a feature decision extraction algorithm based on the management index data set to generate model feature preprocessing data;
step S5: model training processing is carried out by utilizing an optimization algorithm based on model feature preprocessing data, a tuning processing method model is generated, data feedback is carried out based on the tuning processing method model, and basic data processing of the 5G intelligent hospital is achieved.
The invention provides a 5G intelligent hospital basic data processing method, which improves data processing efficiency, accuracy and decision support capability, realizes standardized management and intelligent processing of hospital data, promotes development of intelligent hospitals, improves medical service quality, generates direct interface data through JSON analysis and API, improves data readability and operability, directly interfaces with interface data on the basis of traditional data processing, improves data processing efficiency, generates accurate and reliable task type distribution index data through cross index verification and scene optimization processing formulas, provides reliable basis for hospital decision, utilizes 5G network slices for distribution calculation, ensures real-time processing and high-performance calculation of management index data, improves data processing efficiency and response speed, realizes feature extraction and model training of data through feature decision extraction algorithm and optimization algorithm, generates a tuning processing method model, and continuously improves and optimizes the processing method through data feedback and self optimization of the model, thereby providing continuous improvement and optimization capability for 5G intelligent hospital data processing.
Preferably, step S1 comprises the steps of:
step S11: analyzing the hospital basic data by utilizing JSON based on the hospital basic data to generate JSON analysis data;
step S12: based on the JSON analysis data, utilizing a JSON analysis database to analyze the interface request data, and generating interface request data;
step S13: extracting HTTP response header TOKEN data based on the interface request data to generate response header TOKEN data;
step S14: and carrying out data preprocessing based on the response header TOKEN data, the interface request data and a preset regular expression to generate API direct connection interface data.
According to the invention, the data is directly obtained from the interface, and is analyzed, preprocessed and extracted, so that the processed API direct connection interface data which can be directly used is generated, the hospital basic data is converted into a structured data format through JSON analysis data, the readability and the treatability of the data are provided, the JSON analysis database is used for further processing and analyzing the data obtained through analysis, so that the data conforming to the interface request format is generated, the subsequent data interaction and interface request can be smoothly carried out, the effective transmission and correct analysis of the data are ensured, TOKEN data in the HTTP response head are extracted, and the TOKEN data are used as marks for tracing and tracking the data. The TOKEN can be uniquely identified and related data can be associated through the TOKEN mark, a tracing and tracking basis is provided for subsequent data operation, the reliability and reliability of the data are enhanced, and the data preprocessing is carried out on the response header TOKEN data and the interface request data by utilizing the preset regular expression, so that the response header TOKEN data and the interface request data meet the requirements and specifications of an API (application program interface) direct interface.
Preferably, step S2 comprises the steps of:
step S21: performing data cleaning and noise reduction based on the API direct connection interface data to generate preprocessing interface data;
step S22: medical data extraction is carried out based on the preprocessing interface data, and physiological index data, equipment data and archive identity data are generated;
step S23: performing multiple regression analysis based on the physiological index data, the equipment data and the archive identity data to generate a Bayesian network model of the hospital;
step S24: data extraction is carried out based on a Bayesian network model of a hospital, and primary optimization management index data is generated;
step S25: acquiring hospital task type distribution rule data, performing task data index calculation by utilizing a data index calculation formula based on primary optimization management index data and the hospital task type distribution rule data, and generating task type distribution index data.
According to the invention, firstly, noise and abnormal values in data are removed through data cleaning and noise reduction processing, the accuracy and reliability of the data are improved, specific medical data including physiological index data (such as blood pressure, heart rate and the like), equipment data (such as equipment state, working parameters and the like) and archive identity data (such as patient identity information) are extracted from preprocessing interface data, the subsequent data analysis and modeling are convenient, the characteristics and relevance of the medical data are deeply mined, a relational model among the physiological index data, the equipment data, the archive identity data and target variables is established through a multiple regression analysis method, the generated hospital Bayesian network model can be used for deducing and predicting, decision support is provided for data analysis and management of a hospital, the data is processed and optimized through a scene optimization processing formula based on the hospital Bayesian network model, the primary optimization management index data is generated through optimization processing, the index condition of operation and management of the hospital can be better reflected, and decision reference is provided for a hospital management layer.
Preferably, the data index calculation formula in step S25 is specifically:
;
wherein ,indicate->Seed task type distribution index data, < >>Representing the number of all task data in the primary optimization management index data,/for each task data>Representing the input element as +.>Time->The task type corresponds to->Output probability density function of individual patient, +.>Is->The individual patient is at->Mean under task type>Respectively represent->The individual patient is at->Variance under the task type>Indicate->Weight coefficient corresponding to the task type, < ->For->Adjustment coefficients for the task type.
The invention utilizes a data index calculation formula for calculating task type distribution index data in primary optimization management index dataBased on the input physiological index data, equipment data and archive identity data, a Bayesian network model of the hospital is built by multiple regression analysis, then the model is used for data extraction, task type distribution index data is generated by calculation of a formula, and the formula utilizes the eenthaspect by calculating and weighting and summing probability density functions on patient data of each task type>The individual patient is at- >Mean ∈under task type>First->The individual patient is at->Variance under task type +.>First->Weight coefficient corresponding to the task type +.>In a functional relationship ofThe method comprises the steps of carrying out a first treatment on the surface of the Quantifying and evaluating the distribution of different task types, thus optimizing the data processing and management of hospitals by adjusting the coefficients +.>And the task type distribution index calculation of the primary optimization management index data is realized by utilizing multiple regression analysis and probability density function calculation and combining the weight coefficient and the adjustment coefficient.
Preferably, step S3 comprises the steps of:
step S31: acquiring a format conversion regular expression, and performing data format conversion by using a preset code instruction set based on task type distribution index data and the format conversion regular expression to generate standard format data;
step S32: constructing a 5G network switching allocation algorithm based on the standard format data to generate the 5G network switching allocation algorithm;
step S33: and carrying out standard format data slicing processing based on a 5G network switching allocation algorithm to generate a management index data set.
The invention converts task type distribution index data into data conforming to a standard format by acquiring the format conversion regular expression and utilizing a preset code instruction set. The standard format data generated in this way has unified data structure and specification, the consistency and the processibility of the data are improved, the standard format data are utilized for analysis and modeling, a 5G network switching distribution algorithm is constructed, the algorithms can realize intelligent 5G network switching and resource distribution according to task type distribution and other network parameters in the standard format data, the network performance and the user experience are improved, the standard format data are sliced and processed by the 5G network switching distribution algorithm, an index data set aiming at optimization management is generated, and accurate data support is provided for network optimization and management decision.
Preferably, the 5G network handover allocation algorithm in step S32 is specifically:
;
wherein ,for switching performance index data, +.>For the number of task types in the standard format data, < >>Is->The number of medical devices corresponding to the task type, < >>Representing the%>The devices of the individual task types combine resource consumption data with respect to quantity data +.>Is->Resource capacity relative capacity data of data slice, < ->Representing the computing resource capacity of a data center, +.>Computing resource utilization for the system as a whole, +.>Indicate->No. 5 of individual Hospital>Medical device is at->The proportion of the computing resources occupied in the individual tasks, +.>Indicate->Medical device is at->Weights in individual hospitals,/->Indicate->Computing resource capacity of individual data slices, +.>Indicate->Network quality of individual hospitals>To calculate resource capacity adjustment coefficients.
The invention utilizes a 5G network switching allocation algorithm, the goal of the formula is to calculate switching performance index data to evaluate the effect and quality of 5G network switching, which is based on task type, medical equipment quantity, equipment resource consumption and slice resource capacityThe amount, the computing resource capacity, the system resource utilization rate, the equipment occupation resource proportion, the weight, the network quality and other factors are comprehensively calculated, the task type number, the medical equipment number and the equipment combination resource consumption reflect the switching requirements and the resource consumption conditions of different types of tasks and equipment, and the slicing resource capacity and the computing resource capacity reflect the availability and the limitation of network slicing and computing resources. The utilization rate of system resources considers the utilization condition of the whole resources, and the formula utilizes the task type quantity data in the standard format data First->Medical equipment number corresponding to each task type>First->Task resource storage space occupancy of individual task types +.>First->Resource capacity relative capacity data of data slice +.>Computing resource capacity of data center>Overall computing resource utilization of the system->First->No. 5 of individual Hospital>Medical device is at->The proportion of computing resources occupied in the individual tasks +.>Computing resource capacity->Network quality->By a functional relationshipThe method comprises the steps of carrying out a first treatment on the surface of the Realizing the switching performance index data->Is calculated by the computer.
Preferably, step S4 comprises the steps of:
step S41: performing feature analysis by using chi-square test based on the management index data set to generate a feature subset;
step S42: performing feature extraction algorithm construction by using a preset node purity calculation formula, a preset decision tree division effect calculation formula and a preset chi-square value calculation formula based on the feature subset to generate a feature extraction algorithm;
step S43: carrying out data feature extraction by utilizing a feature extraction algorithm based on the feature subset to generate decision result data;
step S44: and selecting decision result data as model preprocessing data.
The invention realizes the characteristic analysis by carrying out chi-square test on the management index data set. Chi-square test can be used to determine the correlation and significance between different features, thereby screening out features that have an important impact on the target variable. Based on the result of the chi-square test, a feature subset is generated that contains features that have a significant impact on the target variable. The feature subset can help to reduce the data dimension, improve the interpretation and efficiency of the model, the feature subset is utilized to perform feature extraction through a feature decision extraction algorithm to generate decision result data, the feature decision extraction algorithm can select the features with the most representation and information richness from the feature subset to be used for constructing the model or further processing the data, the generated decision result data contains important features subjected to feature extraction, and the generated decision result data has higher information value and prediction capability, can be used for subsequent modeling, analysis and prediction tasks, and improves the accuracy and interpretation of the model.
Preferably, the feature extraction algorithm, the node purity calculation formula, the decision tree division effect calculation formula, and the chi-square value calculation formula in step S42 are respectively as follows:
the node purity calculation formula in step S42 is as follows:
;
wherein ,for node purity data, ++>Representing the number of categories of the current sample, +.>Indicate->Probability of occurrence of individual categories in the sample.
The decision tree division effect calculation formula in step S42 is as follows:
;
wherein ,representing sample relative parameters in a current nodeCount (n)/(l)>Dividing the function for the node->For the node sample data quantity, +.>Representing node correspondence characteristic data, < >>And the classification standard value of the corresponding node.
The chi-square value calculation formula in step S42 is as follows:
;
wherein ,representing the sample relative parameters in the current node, +.>For the classification criterion->Representing the corresponding characteristic data of the node,representation feature->Classification result and classification criterion->The number of co-occurrences in the sample, +.>Representing expected value data.
The feature extraction algorithm in step S42 is as follows:
+H+/>;
wherein ,for decision result data, ++>Representing the sample relative parameters in the current node, +.>Weights representing corresponding characteristic parameters +. >Representing characteristic parameters +.>Decision tree function of->Representing the number of leaf nodes of the decision tree, +.>For node purity data, ++>For chi-square value data, < >>The effect data is partitioned for the decision tree.
The invention utilizes a feature extraction algorithm which takes feature data attached to nodes in a decision tree as basic data, wherein the purity of the nodes is represented by an entropy concept, and node purity data is carried out based on sample category data and probability data of sample occurrence by utilizing a node purity calculation formulaIs calculated by using the relative parameters of the sample in the current node, the corresponding characteristic data of the node, the expected value data, the characteristic classification result and the common occurrence number of the classification standard result in the sample, and chi-square value data is generated>Calculating by using a decision tree division effect calculation formula to generate decision tree division effect data +.>Based on chi-square value data->Decision tree partitioning effect dataNode purity data->Connecting sample relative parameters in the current node by utilizing the characteristic data attached to the leaf node>Leaf node number of decision tree->By functional relation->+H+/>The method comprises the steps of carrying out a first treatment on the surface of the Calculating to generate decision result data +. >。
Preferably, step S5 comprises the steps of:
step S51: based on decision result data, carrying out model parameter adjustment by using a model optimization algorithm to generate a comparison index;
step S52: acquiring a service scene, and performing intelligent equipment connection by using a 5G network based on the service scene and the comparison index to generate an index processing method model;
preferably, the model optimization algorithm is specifically:
;
wherein ,for comparison index, ->Representing the number of samples of the training dataset, +.>Representing the number of features->Indicate->Feature vector of individual samples, +_>Indicate->Target variable of individual samples, +.>Representing model parameter vector, ++>Predictive value representing a logistic regression model, +.>For regularization parameters, ++>Representing the loss function pair->Partial derivatives of the individual parameters.
The invention utilizes a model optimization algorithm, the formula optimizes the parameters of a logistic regression model by minimizing a loss function, so that the logistic regression model can better fit a training data set and improve the prediction accuracy, the introduction of regularization terms can balance the complexity of the model and prevent the occurrence of over-fitting problems, and the formula firstly utilizes the predicted value of the logistic regression modelLogarithmic probability and->Target variable +. >The product calculation is carried out, the matching degree of the result of the loss function and the actual label can be expressed as a continuous function by multiplying the logarithmic probability, and smaller loss value can be obtained when the model prediction is correct, when the prediction result of the model is consistent with the actual label, the result of the multiplication of the logarithmic probability approaches 1, the value of the loss function approaches 0, the prediction of the model is accurate, the loss value of each sample is accumulated by using a summation symbol to obtain the total loss of the whole training data set, the total matching degree between the prediction result and the actual label on the whole data set can be considered by adding the loss value of each sample, and the sample number of the training data set is utilized based on the principle>First->Target variable +.>Predictive value of logistic regression model +.>Forming a functional relationshipThe method comprises the steps of carrying out a first treatment on the surface of the Parallel level regularization term->The method comprises the steps of carrying out a first treatment on the surface of the The complexity of the control model is realized, the overfitting is prevented, the regularization term has the function of restraining model parameters, so that the parameter value of the model is not overlarge, the overfitting of the model to training data is prevented, the fitting capacity and the generalization capacity of the model can be balanced by adding the regularization term, and the phenomenon that the model performs poorly on new data due to overdependence on certain characteristics is avoided, thereby realizing the comparison index +. >Is calculated by the computer.
Step S53: acquiring output data based on the index processing method model, and generating model output data;
step S54: based on the model output data and the API direct connection interface data, utilizing the 5G network intelligent equipment to conduct scene data prediction output, and generating real-time equipment output data;
step S55: model data feedback is carried out based on real-time equipment output data, and 5G intelligent hospital basic data processing is achieved.
According to the invention, parameter adjustment is carried out on model preprocessing data through a model optimization algorithm, better adaptability and accuracy of the model are realized, performance and effect of the model are improved, a 5G network is utilized to connect intelligent equipment in combination with service scenes and comparison indexes, an index processing method model aiming at a specific scene is generated, a customized data processing scheme is provided, real-time data acquisition and scene data prediction output are carried out by utilizing the 5G network intelligent equipment based on model output data and API direct connection port data, rapid generation and feedback of real-time equipment output data are realized, rapid and accurate processing and analysis of the 5G intelligent hospital basic data are realized by utilizing the 5G network and intelligent equipment for data processing, instantaneity and efficiency of data processing are improved, data feedback transmission of the model is realized based on real-time equipment output data, and the performance and quality of the 5G intelligent hospital basic data processing are improved through continuous improvement and optimization.
In one embodiment of the present specification, there is provided a 5G intelligent hospital base data processing database platform comprising:
at least one processor;
a memory communicatively coupled to the at least one processor;
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the 5G intelligent hospital base data processing method of any of the above.
The invention provides a 5G intelligent hospital basic data processing database platform, which can realize any 5G intelligent hospital basic data processing method, realize data acquisition, operation and generation, perform distribution calculation through hospital basic data, operate image-text information in the data according to a designed instruction sequence to generate pre-processed image-text information, perform model processing training through the pre-processed image-text information to generate a tuning processing method model, perform data processing according to the tuning processing method model, realize 5G intelligent hospital basic data processing, and complete the method operation steps by following a set instruction set in the system to push the 5G intelligent hospital basic data processing method.
The application provides a 5G intelligent hospital basic data processing method, which solves the problems of low efficiency and high bandwidth requirement in the traditional hospital data processing by comprehensively applying a multidisciplinary and multiclass model, and realizes the 5G intelligent hospital basic data processing method with high efficiency, low bandwidth, low delay and dynamic network switching function.
Drawings
FIG. 1 is a flowchart showing steps of a basic data processing method for a 5G intelligent hospital according to the present application;
FIG. 2 is a detailed implementation step flow diagram of step S1;
fig. 3 is a detailed implementation step flow diagram of step S2.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The embodiment of the application provides a 5G intelligent hospital basic data processing method. The execution subject of the 5G intelligent hospital basic data processing method comprises, but is not limited to, the system: mechanical devices, data processing platforms, cloud server nodes, network transmission devices, etc. may be considered general purpose computing nodes of the present application. The data processing platform includes, but is not limited to: at least one of an audio management system, an image management system and an information management system.
Referring to fig. 1 to 3, the present invention provides a 5G intelligent hospital basic data processing method, which includes the following steps:
step S1: acquiring hospital basic data, analyzing the hospital basic data by utilizing JSON based on the hospital basic data, and generating API direct connection interface data;
step S2: performing cross index verification by using a scene optimization processing formula based on the API direct connection interface data to generate task type distribution index data;
step S3: performing distribution calculation by using a 5G network slice based on task type distribution index data to generate a management index data set;
step S4: performing feature extraction by using a feature decision extraction algorithm based on the management index data set to generate model feature preprocessing data;
step S5: model training processing is carried out by utilizing an optimization algorithm based on model feature preprocessing data, a tuning processing method model is generated, data feedback is carried out based on the tuning processing method model, and basic data processing of the 5G intelligent hospital is achieved.
The invention provides a 5G intelligent hospital basic data processing method, which improves data processing efficiency, accuracy and decision support capability, realizes standardized management and intelligent processing of hospital data, promotes development of intelligent hospitals, improves medical service quality, generates direct interface data through JSON analysis and API, improves data readability and operability, directly interfaces with interface data on the basis of traditional data processing, improves data processing efficiency, generates accurate and reliable task type distribution index data through cross index verification and scene optimization processing formulas, provides reliable basis for hospital decision, utilizes 5G network slices for distribution calculation, ensures real-time processing and high-performance calculation of management index data, improves data processing efficiency and response speed, realizes feature extraction and model training of data through feature decision extraction algorithm and optimization algorithm, generates a tuning processing method model, and continuously improves and optimizes the processing method through data feedback and self optimization of the model, thereby providing continuous improvement and optimization capability for 5G intelligent hospital data processing.
In the embodiment of the invention, please refer to fig. 1, the method for processing basic data of a 5G intelligent hospital includes the following steps:
step S1: acquiring hospital basic data, analyzing the hospital basic data by utilizing JSON based on the hospital basic data, and generating API direct connection interface data;
in the embodiment of the invention, for example, hospital basic data is firstly organized according to a JSON format, including key value pairs, arrays and nested structures, values of specific fields or key value pairs in the JSON data are queried and used as parameters or data of an interface request, required information can be extracted from the JSON data and interface request data are generated by using a JSON parser and a database query language such as SQL, after HTTP request is carried out, a server returns an HTTP response containing a response header, containing TOKEN data related to identity verification or access control, the required TOKEN data is extracted by parsing the response header of the HTTP response, the acquired TOKEN data of the response header is combined with the interface request data, and data processing and cleaning including matching, replacing, extracting or other data operation operations are carried out by applying a preset regular expression mode to obtain final API direct interface data.
Step S2: performing cross index verification by using a scene optimization processing formula based on the API direct connection interface data to generate task type distribution index data;
in the embodiment of the invention, for example, a proper data cleaning technology is adopted, values of specific fields or key value pairs are analyzed and extracted to obtain physiological index data, equipment data and archive identity data, for the data such as physiological measurement values of patients, states and parameters of equipment, personal information of the patients and medical archive data, the physiological index data, the equipment data and the archive identity data are used as independent variables, the input data are processed and calculated to generate primary optimized management index data, the index data can be used for evaluating and monitoring operation conditions, resource allocation and performance improvement of hospitals, hospital task type distribution rule data are obtained, the rules describe distribution conditions of different task types in the hospitals, then the primary optimized management index data and the task type distribution rule data are matched and marked, and the data are distributed to different tasks according to the rules.
Step S3: performing distribution calculation by using a 5G network slice based on task type distribution index data to generate a management index data set;
In the embodiment of the invention, for example, a regular expression for data format conversion is obtained, a preset code instruction set is applied in combination with task type distribution index data, a conversion code is written by using a programming language or a script, the data is converted into a standard data format, the standard format data is used as input, a switching allocation algorithm suitable for a 5G network is designed and constructed by analyzing data characteristics and performance indexes, a proper mathematical model and an optimization method are adopted to improve the efficiency and performance of network switching, the data in the standard format is sliced by using the 5G network switching allocation algorithm, the data can be divided into proper time periods, space areas or service types to generate a data set with certain granularity and relevance, and a management index data set can be generated by analyzing and calculating the sliced data set for evaluating and improving the performance, resource utilization and service quality of the 5G network.
Step S4: performing feature extraction by using a feature decision extraction algorithm based on the management index data set to generate model feature preprocessing data;
in the embodiment of the invention, for example, a management index data set is used as input, the statistical significance of characteristics is selected by calculating chi-square statistics and corresponding significance level, a characteristic subset is generated, the characteristic subset is used as input, a characteristic decision extraction algorithm is applied to further screen and extract the characteristics with the most information quantity and importance, the characteristic decision extraction algorithm can evaluate the importance of the characteristics and the predictive capability of the target variable by using various technologies such as information gain, a radix index or a correlation coefficient, and the like, and the characteristic decision extraction algorithm is used for carrying out characteristic extraction by connecting a node purity calculation formula, a decision tree division effect calculation formula and a chi-square value calculation formula which are used in the concrete calculation process of decision tree data by selecting the characteristics with high correlation and differentiation degree, so as to generate decision result data.
Step S5: model training processing is carried out by utilizing an optimization algorithm based on model feature preprocessing data, a tuning processing method model is generated, data feedback is carried out based on the tuning processing method model, and basic data processing of the 5G intelligent hospital is achieved.
According to the invention, parameter adjustment is carried out on model preprocessing data through a model optimization algorithm, so that better adaptability and accuracy of the model are realized, performance and effect of the model are improved, a 5G network is used for connecting intelligent equipment in combination with service scenes and comparison indexes, an index processing method model aiming at a specific scene is generated, a customized data processing scheme is provided, real-time data acquisition and scene data prediction output are carried out by using the 5G network intelligent equipment based on model output data and API direct connection port data, rapid and accurate processing and analysis of 5G intelligent hospital basic data are realized, instantaneity and efficiency of data processing are improved, data feedback of the model is realized based on real-time equipment output data, and performance and quality of 5G intelligent hospital basic data processing are improved through continuous improvement and optimization.
In the embodiment of the present invention, referring to fig. 2, the detailed implementation steps of step S1 include:
Step S11: analyzing the hospital basic data by utilizing JSON based on the hospital basic data to generate JSON analysis data;
step S12: based on the JSON analysis data, utilizing a JSON analysis database to analyze the interface request data, and generating interface request data;
step S13: extracting HTTP response header TOKEN data based on the interface request data to generate response header TOKEN data;
step S14: and carrying out data preprocessing based on the response header TOKEN data, the interface request data and a preset regular expression to generate API direct connection interface data.
According to the invention, the data is directly obtained from the interface, and is analyzed, preprocessed and extracted, so that the processed API direct connection interface data which can be directly used is generated, the hospital basic data is converted into a structured data format through JSON analysis data, the readability and the treatability of the data are provided, the JSON analysis database is used for further processing and analyzing the data obtained through analysis, so that the data conforming to the interface request format is generated, the subsequent data interaction and interface request can be smoothly carried out, the effective transmission and correct analysis of the data are ensured, TOKEN data in the HTTP response head are extracted, and the TOKEN data are used as marks for tracing and tracking the data. The TOKEN can be uniquely identified and related data can be associated through the TOKEN mark, a tracing and tracking basis is provided for subsequent data operation, the reliability and reliability of the data are enhanced, and the data preprocessing is carried out on the response header TOKEN data and the interface request data by utilizing the preset regular expression, so that the response header TOKEN data and the interface request data meet the requirements and specifications of an API (application program interface) direct interface.
In the embodiment of the invention, for example, hospital basic data is firstly organized according to a JSON format and comprises key value pairs, arrays and nested structures, then, a JSON parser or a related programming language library, such as a JSON module in Python, is used for parsing the hospital basic data, converting the hospital basic data into JSON objects or data structures in a memory, querying values of specific fields or key value pairs in the JSON data and using the values as parameters or data of interface requests, and by using the JSON parser and a database query language, such as SQL, required information can be extracted from the JSON data and interface request data can be generated, and based on the interface request data, the extraction of HTTP response header TOKEN data is performed. After performing the HTTP request, the server returns an HTTP response including a response header, including TOKEN data related to authentication or access control, extracts the required TOKEN data by parsing the response header of the HTTP response using an appropriate method (e.g., a regular expression, a string operation, or a special HTTP response header parsing library), merges the obtained TOKEN data with the interface request data, and performs data processing and cleaning by applying a preset regular expression pattern, including matching, replacing, extracting, or other data operation operations, to obtain final API direct interface data.
In the embodiment of the present invention, referring to fig. 3, the detailed implementation steps of step S2 include:
step S21: performing data cleaning and noise reduction based on the API direct connection interface data to generate preprocessing interface data;
step S22: medical data extraction is carried out based on the preprocessing interface data, and physiological index data, equipment data and archive identity data are generated;
step S23: performing multiple regression analysis based on the physiological index data, the equipment data and the archive identity data to generate a Bayesian network model of the hospital;
step S24: data extraction is carried out based on a Bayesian network model of a hospital, and primary optimization management index data is generated;
step S25: acquiring hospital task type distribution rule data, performing task data index calculation by utilizing a data index calculation formula based on primary optimization management index data and the hospital task type distribution rule data, and generating task type distribution index data.
According to the invention, firstly, noise and abnormal values in data are removed through data cleaning and noise reduction processing, the accuracy and reliability of the data are improved, specific medical data including physiological index data (such as blood pressure, heart rate and the like), equipment data (such as equipment state, working parameters and the like) and archive identity data (such as patient identity information) are extracted from the preprocessing interface data, the subsequent data analysis and modeling are convenient, the characteristics and the relevance of the medical data are deeply mined, a multiple regression analysis method is adopted, a relational model among the physiological index data, the equipment data, the archive identity data and target variables is established, the generated hospital Bayesian network model can be used for deducing and predicting, decision support is provided for data analysis and management of a hospital, primary optimization management index data is generated by carrying out data extraction based on the hospital Bayesian network model, the index condition of hospital operation and management can be better reflected, and decision reference is provided for a hospital management layer.
In the embodiment of the invention, for example, a proper data cleaning technology is adopted, such as abnormal value removal, missing data filling, error correction data correction and the like, so as to ensure the quality and consistency of interface data, a data cleaning and noise reduction algorithm is applied to generate preprocessed interface data for subsequent processing, the values of specific fields or key value pairs are analyzed and extracted according to the structure and content of the preprocessed interface data to obtain physiological index data, equipment data and archive identity data, physiological index data, equipment data and archive identity data of patients, such as physiological measurement values of patients, states and parameters of equipment, personal information and medical archive data of the patients are taken as independent variables, a mathematical model is established through multiple regression analysis, a Bayesian network model is constructed through a statistical method, a machine learning algorithm or an artificial intelligence technology to understand the association and potential between different variables, primary optimization management index data is generated by analyzing and extracting values of specific fields or key value pairs according to the structure and content of the preprocessed interface data, the index data can be used for evaluating and monitoring the condition, resource allocation and performance improvement of hospitals, the physiological index data, the equipment data and archive identity data are taken as independent variables, the physiological index data and the archive identity data are distributed in a multiple regression analysis, the Bayesian network model is constructed through a statistical method, a mechanical learning algorithm or an artificial intelligence technology is applied to understand the association and potential between different variables, the task management rule is obtained, the task rule is distributed according to the task rule is distributed to the task rule of different task rule and the task rule is distributed and the task rule of the task rule is distributed and the task rule.
In the embodiment of the present invention, the data index calculation formula in step S25 specifically includes:
;
wherein ,indicate->Seed task type distribution index data, < >>Representing the number of all task data in the primary optimization management index data,/for each task data>Representing the input element as +.>Time->The task type corresponds to->Output probability density function of individual patient, +.>Is->The individual patient is at->Mean under task type>Respectively represent->The individual patient is at->Variance under the task type>Indicate->Weight coefficient corresponding to the task type, < ->For->Adjustment coefficients for the task type.
The invention utilizes a data index calculation formula for calculating task type distribution index data in primary optimization management index dataBased on the input physiological index data, equipment data and archive identity data, a Bayesian network model of the hospital is built by multiple regression analysis, then the model is used for data extraction, task type distribution index data is generated by calculation of a formula, and the formula utilizes the eenthaspect by calculating and weighting and summing probability density functions on patient data of each task type >The individual patient is at->Mean ∈under task type>First->The individual patient is at->Variance under task type +.>First->Weight coefficient corresponding to the task type +.>In a functional relationship ofThe method comprises the steps of carrying out a first treatment on the surface of the Quantifying and evaluating the distribution of different task types, thus optimizing the data processing and management of hospitals by adjusting the coefficients +.>And the task type distribution index calculation of the primary optimization management index data is realized by utilizing multiple regression analysis and probability density function calculation and combining the weight coefficient and the adjustment coefficient.
In the embodiment of the invention, the specific steps of step S3 are as follows:
step S31: acquiring a format conversion regular expression, and performing data format conversion by using a preset code instruction set based on task type distribution index data and the format conversion regular expression to generate standard format data;
step S32: constructing a 5G network switching allocation algorithm based on the standard format data to generate the 5G network switching allocation algorithm;
step S33: and carrying out standard format data slicing processing based on a 5G network switching allocation algorithm to generate a management index data set.
The invention converts task type distribution index data into data conforming to a standard format by acquiring the format conversion regular expression and utilizing a preset code instruction set. The standard format data generated in this way has unified data structure and specification, the consistency and the processibility of the data are improved, the standard format data are utilized for analysis and modeling, a 5G network switching distribution algorithm is constructed, the algorithms can realize intelligent 5G network switching and resource distribution according to task type distribution and other network parameters in the standard format data, the network performance and the user experience are improved, the standard format data are sliced and processed by the 5G network switching distribution algorithm, an index data set aiming at optimization management is generated, and accurate data support is provided for network optimization and management decision.
In the embodiment of the invention, for example, a regular expression for data format conversion is obtained, the expression defines a data format rule which needs to be matched and converted, a preset code instruction set is applied in combination with task type distribution index data, a conversion code is written by using a programming language or script, format conversion operation is carried out on the data, the data is converted into a standard data format, the data in the standard format is used as input, a switching allocation algorithm suitable for a 5G network is designed and constructed by analyzing data characteristics and performance indexes, a proper mathematical model and an optimization method are adopted to improve the efficiency and performance of network switching, the 5G network switching allocation algorithm is used for slicing the data in the standard format, the data can be divided into a proper time period, a space area or a service type to generate a data set with certain granularity and relevance, and a management index data set can be generated by analyzing and calculating the sliced data set for evaluating and improving the performance, resource utilization and service quality of the 5G network.
In the embodiment of the present invention, the 5G network switching allocation algorithm in step S32 specifically includes:
;
wherein ,for switching performance index data, +.>For the number of task types in the standard format data, < >>Is->The number of medical devices corresponding to the task type, < >>Representing the%>The devices of the individual task types combine resource consumption data with respect to quantity data +.>Is->Resource capacity relative capacity data of data slice, < ->Representing the computing resource capacity of a data center, +.>Computing resource utilization for the system as a whole, +.>Indicate->No. 5 of individual Hospital>Medical device is at->The proportion of the computing resources occupied in the individual tasks, +.>Indicate->Medical device is at->Weights in individual hospitals,/->Indicate->Computing resource capacity of individual data slices, +.>Indicate->Network quality of individual hospitals>To calculate resource capacity adjustment coefficients.
The invention utilizes a 5G network switching allocation algorithm, the goal of the formula is to calculate switching performance index data to evaluate the effect and quality of 5G network switching, the method is to comprehensively calculate based on factors such as task type, medical equipment quantity, equipment resource consumption, slice resource capacity, calculation resource capacity, system resource utilization rate, equipment occupation resource proportion, weight, network quality and the like, the task type quantity, medical equipment quantity and equipment combination resource consumption reflect the switching requirements and resource consumption conditions of different types of tasks and equipment, and the slice resource capacity and calculation resource capacity reflect the availability and limitation of network slicing and calculation resources. The utilization rate of system resources considers the utilization condition of the whole resources, and the formula utilizes the task type quantity data in the standard format data First->Medical equipment number corresponding to each task type>First->Task resource storage space occupancy of individual task types +.>First->Resource capacity relative capacity data of data slice +.>Computing resource capacity of data center>Overall computing resource utilization of the system->First->No. 5 of individual Hospital>Medical device is at->The proportion of computing resources occupied in the individual tasks +.>Computing resource capacity->Network quality->By a functional relationshipThe method comprises the steps of carrying out a first treatment on the surface of the Realizing the switching performance index data->Calculation of (2)。
In the embodiment of the invention, the specific steps of step S4 are as follows:
step S41: performing feature analysis by using chi-square test based on the management index data set to generate a feature subset;
step S42: performing feature extraction algorithm construction by using a preset node purity calculation formula, a preset decision tree division effect calculation formula and a preset chi-square value calculation formula based on the feature subset to generate a feature extraction algorithm;
step S43: carrying out data feature extraction by utilizing a feature extraction algorithm based on the feature subset to generate decision result data;
step S44: and selecting decision result data as model preprocessing data.
The invention realizes the characteristic analysis by carrying out chi-square test on the management index data set. Chi-square test can be used to determine the correlation and significance between different features, thereby screening out features that have an important impact on the target variable. Based on the result of the chi-square test, a feature subset is generated that contains features that have a significant impact on the target variable. The feature subset can help to reduce the data dimension, improve the interpretation and efficiency of the model, the feature subset is utilized to perform feature extraction through a feature decision extraction algorithm to generate decision result data, the feature decision extraction algorithm can select the features with the most representation and information richness from the feature subset to be used for constructing the model or further processing the data, the generated decision result data contains important features subjected to feature extraction, and the generated decision result data has higher information value and prediction capability, can be used for subsequent modeling, analysis and prediction tasks, and improves the accuracy and interpretation of the model.
In the embodiment of the invention, for example, a management index data set is used as input, the statistical significance of characteristics is selected by calculating chi-square statistics and corresponding significance level, a characteristic subset is generated, the characteristic subset is used as input, a characteristic decision extraction algorithm is applied to further screen and extract the characteristics with the most information quantity and importance, the characteristic decision extraction algorithm can evaluate the importance of the characteristics and the predictive capability of the target variable by using various technologies such as information gain, a radix index or a correlation coefficient, and the like, and the characteristic decision extraction algorithm is used for carrying out characteristic extraction by connecting a node purity calculation formula, a decision tree division effect calculation formula and a chi-square value calculation formula which are used in the concrete calculation process of decision tree data by selecting the characteristics with high correlation and differentiation degree, so as to generate decision result data.
Preferably, the feature extraction algorithm, the node purity calculation formula, the decision tree division effect calculation formula, and the chi-square value calculation formula in step S42 are respectively as follows:
the node purity calculation formula in step S42 is as follows:
;
wherein ,for node purity data, ++>Representing the number of categories of the current sample, +. >Indicate->Probability of occurrence of individual categories in the sample.
The decision tree division effect calculation formula in step S42 is as follows:
;
wherein ,representing the sample relative parameters in the current node, +.>Dividing the function for the node->For the node sample data quantity, +.>Representing node correspondence characteristic data, < >>And the classification standard value of the corresponding node.
The chi-square value calculation formula in step S42 is as follows:
;
wherein ,representing the sample relative parameters in the current node, +.>For the classification criterion->Representing the corresponding characteristic data of the node,representation feature->Classification result and classification criterion->The number of co-occurrences in the sample, +.>Representing expected value data.
The feature extraction algorithm in step S42 is as follows:
+H+/>;
wherein ,for decision result data, ++>Representing the sample relative parameters in the current node, +.>Weights representing corresponding characteristic parameters +.>Representing characteristic parameters +.>Decision tree function of->Representing the number of leaf nodes of the decision tree, +.>For node purity data, ++>For chi-square value data, < >>The effect data is partitioned for the decision tree.
The invention utilizes a feature extraction algorithm which takes feature data attached to nodes in a decision tree as basic data, wherein the purity of the nodes is represented by an entropy concept, and node purity data is carried out based on sample category data and probability data of sample occurrence by utilizing a node purity calculation formula Using sample relative in the current nodeCalculating the number of co-occurrence of the parameter, the node corresponding characteristic data, the expected value data and the characteristic classification result and the classification standard result in the sample into chi-square value to generate chi-square value data +.>Calculating by using a decision tree division effect calculation formula to generate decision tree division effect data +.>Based on chi-square value data->Decision tree partitioning effect dataNode purity data->Connecting sample relative parameters in the current node by utilizing the characteristic data attached to the leaf node>Leaf node number of decision tree->By functional relation->+H+/>The method comprises the steps of carrying out a first treatment on the surface of the Calculating to generate decision result data +.>。
In the embodiment of the invention, the specific steps of step S5 are as follows:
step S51: based on the model preprocessing data, carrying out model parameter adjustment by using a model optimization algorithm to generate a comparison index;
step S52: acquiring a service scene, and performing intelligent equipment connection by using a 5G network based on the service scene and the comparison index to generate an index processing method model;
in the embodiment of the invention, the model optimization algorithm is specifically:
;
wherein ,for comparison index, ->Representing the number of samples of the training dataset, +.>Representing the number of features- >Indicate->Feature vector of individual samples, +_>Indicate->Target variable of individual samples, +.>Representing model parameter vector, ++>Predictive value representing a logistic regression model, +.>For regularization parameters, ++>Representing the loss function pair->Partial derivatives of the individual parameters.
The invention utilizes a model optimization algorithm, the formula optimizes the parameters of a logistic regression model by minimizing a loss function, so that the logistic regression model can better fit a training data set and improve the prediction accuracy, the introduction of regularization terms can balance the complexity of the model and prevent the occurrence of over-fitting problems, and the formula firstly utilizes the predicted value of the logistic regression modelLogarithmic probability and->Target variable +.>The product calculation is carried out, the matching degree of the result of the loss function and the actual label can be expressed as a continuous function by multiplying the logarithmic probability, and smaller loss value can be obtained when the model prediction is correct, when the prediction result of the model is consistent with the actual label, the result of the multiplication of the logarithmic probability approaches 1, the value of the loss function approaches 0, the prediction of the model is accurate, the loss value of each sample is accumulated by using a summation symbol to obtain the total loss of the whole training data set, the total matching degree between the prediction result and the actual label on the whole data set can be considered by adding the loss value of each sample, and the sample number of the training data set is utilized based on the principle >First->Target variable +.>Predictive value of logistic regression model +.>Forming a functional relationshipThe method comprises the steps of carrying out a first treatment on the surface of the Parallel level regularization term->The method comprises the steps of carrying out a first treatment on the surface of the The complexity of the control model is realized, the overfitting is prevented, the regularization term has the function of restraining model parameters, so that the parameter value of the model is not overlarge, the overfitting of the model to training data is prevented, the fitting capacity and the generalization capacity of the model can be balanced by adding the regularization term, and the phenomenon that the model performs poorly on new data due to overdependence on certain characteristics is avoided, thereby realizing the comparison index +.>Is calculated by the computer. />
Step S53: acquiring output data based on the index processing method model, and generating model output data;
step S54: based on the model output data and the API direct connection interface data, utilizing the 5G network intelligent equipment to conduct scene data prediction output, and generating real-time equipment output data;
step S55: model data feedback is carried out based on real-time equipment output data, and 5G intelligent hospital basic data processing is achieved.
According to the invention, parameter adjustment is carried out on model preprocessing data through a model optimization algorithm, better adaptability and accuracy of the model are realized, performance and effect of the model are improved, a 5G network is utilized to connect intelligent equipment in combination with service scenes and comparison indexes, an index processing method model aiming at a specific scene is generated, a customized data processing scheme is provided, real-time data acquisition and scene data prediction output are carried out by utilizing the 5G network intelligent equipment based on model output data and API direct connection port data, rapid generation and feedback of real-time equipment output data are realized, rapid and accurate processing and analysis of the 5G intelligent hospital basic data are realized by utilizing the 5G network and intelligent equipment for data processing, instantaneity and efficiency of data processing are improved, data feedback transmission of the model is realized based on real-time equipment output data, and the performance and quality of the 5G intelligent hospital basic data processing are improved through continuous improvement and optimization.
In the embodiment of the invention, data are preprocessed through an analysis model, a model optimization algorithm is applied, model parameters are adjusted and optimized, a set of comparison indexes are generated through the optimized model parameters, relevant data are collected according to specific business scene requirements, the previously generated comparison indexes are combined, a 5G network is used for connecting intelligent equipment, output data acquisition is performed based on an index processing method model, model output data are generated, the model output data and API direct connection port data are combined, the 5G network is used for connecting intelligent equipment, scene data prediction and analysis are performed, real-time equipment output data are generated through real-time interaction with the intelligent equipment, the result based on the model output data and the scene data prediction is included, the data are fed back to the model for further analysis and processing according to the real-time equipment output data, and processing and optimization of basic data of a 5G intelligent hospital are realized, so that the accuracy and efficiency of data processing are improved.
In one embodiment of the present specification, there is provided a 5G intelligent hospital base data processing database platform comprising:
at least one processor;
a memory communicatively coupled to the at least one processor;
The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the 5G intelligent hospital base data processing method of any of the above.
The invention provides a 5G intelligent hospital basic data processing database platform, which can realize any 5G intelligent hospital basic data processing method, realize data acquisition, operation and generation, perform distribution calculation through hospital basic data, operate image-text information in the data according to a designed instruction sequence to generate pre-processed image-text information, perform model processing training through the pre-processed image-text information to generate a tuning processing method model, perform data processing according to the tuning processing method model, realize 5G intelligent hospital basic data processing, and complete the method operation steps by following a set instruction set in the system to push the 5G intelligent hospital basic data processing method.
The invention provides a 5G intelligent hospital basic data processing method, which solves the problems of low efficiency and high bandwidth requirement in the traditional hospital data processing by comprehensively applying a multidisciplinary and multiclass model, and realizes the 5G intelligent hospital basic data processing method with high efficiency, low bandwidth, low delay and dynamic network switching function.
The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. The 5G intelligent hospital basic data processing method is characterized by comprising the following steps of:
step S1: acquiring hospital basic data, analyzing the hospital basic data by utilizing JSON based on the hospital basic data, and generating API direct connection interface data;
step S2: performing cross index verification by using a scene optimization processing formula based on the API direct connection interface data to generate task type distribution index data;
step S3: performing distribution calculation by using a 5G network slice based on task type distribution index data to generate a management index data set;
step S4: performing feature extraction by using a feature decision extraction algorithm based on the management index data set to generate model feature preprocessing data;
Step S5: model training processing is carried out by utilizing an optimization algorithm based on model feature preprocessing data, a tuning processing method model is generated, data feedback is carried out based on the tuning processing method model, and basic data processing of the 5G intelligent hospital is achieved.
2. The method according to claim 1, wherein the specific steps of step S1 are:
step S11: analyzing the hospital basic data by utilizing JSON based on the hospital basic data to generate JSON analysis data;
step S12: based on the JSON analysis data, utilizing a JSON analysis database to analyze the interface request data, and generating interface request data;
step S13: extracting HTTP response header TOKEN data based on the interface request data to generate response header TOKEN data;
step S14: and carrying out data preprocessing based on the response header TOKEN data, the interface request data and a preset regular expression to generate API direct connection interface data.
3. The method according to claim 1, wherein the specific steps of step S2 are:
step S21: performing data cleaning and noise reduction based on the API direct connection interface data to generate preprocessing interface data;
step S22: medical data extraction is carried out based on the preprocessing interface data, and physiological index data, equipment data and archive identity data are generated;
Step S23: performing multiple regression analysis based on the physiological index data, the equipment data and the archive identity data to generate a Bayesian network model of the hospital;
step S24: data extraction is carried out based on a Bayesian network model of a hospital, and primary optimization management index data is generated;
step S25: acquiring hospital task type distribution rule data, performing task data index calculation by utilizing a data index calculation formula based on primary optimization management index data and the hospital task type distribution rule data, and generating task type distribution index data.
4. The method according to claim 3, wherein the data index calculation formula in step S25 is specifically:
;
wherein ,indicate->Seed task type distribution index data, < >>Representing primary optimization management index dataAll task data amounts in>Representing the input element as +.>Time->The task type corresponds to->Output probability density function of individual patient, +.>Is->The individual patient is at->Mean under task type>Respectively represent->The individual patient is at->Variance under the task type>Indicate->Weight coefficient corresponding to the task type, < ->For->Adjustment coefficients for the task type.
5. The method according to claim 4, wherein the specific steps of step S3 are:
step S31: acquiring a format conversion regular expression, and performing data format conversion by using a preset code instruction set based on task type distribution index data and the format conversion regular expression to generate standard format data;
step S32: constructing a 5G network switching allocation algorithm based on the standard format data to generate the 5G network switching allocation algorithm;
step S33: and carrying out standard format data slicing processing based on a 5G network switching allocation algorithm to generate a management index data set.
6. The method according to claim 5, wherein the 5G network handover allocation algorithm in step S32 is specifically:
;
wherein ,for switching performance index data, +.>For the number of task types in the standard format data, < >>Is->The number of medical devices corresponding to the task type, < >>Representing the%>Task resource storage space occupancy of individual task types, < ->Is->Resource capacity relative capacity data of data slice, < ->Representing the computing resource capacity of a data center, +.>Computing resource utilization for the system as a whole, +.>Indicate->No. 5 of individual Hospital >Medical device is at->The proportion of the computing resources occupied in the individual tasks, +.>Indicate->Medical device is at->Weights in individual hospitals,/->Indicate->Computing resource capacity of individual data slices, +.>Indicate->Network quality of individual hospitals>To calculate resource capacity adjustment coefficients.
7. The method according to claim 1, wherein the specific steps in step S4 are:
step S41: performing feature analysis by using chi-square test based on the management index data set to generate a feature subset;
step S42: performing feature extraction algorithm construction by using a preset node purity calculation formula, a preset decision tree division effect calculation formula and a preset chi-square value calculation formula based on the feature subset to generate a feature extraction algorithm;
step S43: carrying out data feature extraction by utilizing a feature extraction algorithm based on the feature subset to generate decision result data;
step S44: and selecting decision result data as model preprocessing data.
8. The method according to claim 7, wherein the feature extraction algorithm, the node purity calculation formula, the decision tree partition effect calculation formula, and the chi-square value calculation formula in step S42 are respectively as follows:
The node purity calculation formula in step S42 is as follows:
;
wherein ,for node purity data, ++>Representing the number of categories of the current sample, +.>Indicate->Probability of occurrence of individual categories in the sample;
the decision tree division effect calculation formula in step S42 is as follows:
;
wherein ,representing the sample relative parameters in the current node, +.>Dividing the function for the node->For the node sample data quantity, +.>Representing node correspondence characteristic data, < >>A classification standard value of the corresponding node;
the chi-square value calculation formula in step S42 is as follows:
;
wherein ,representing the sample relative parameters in the current node, +.>For the classification criterion->Representing node correspondence characteristic data, < >>Representation feature->Classification result and classification criterion->The number of co-occurrences in the sample, +.>Representing expected value data;
the feature extraction algorithm in step S42 is as follows:
+H+/>;
wherein ,for decision result data, ++>Representing a current nodeSample relative parameters of->Weights representing corresponding characteristic parameters +.>Representing characteristic parameters +.>Decision tree function of->Representing the number of leaf nodes of the decision tree, +.>For the node purity data to be used,for chi-square value data, < >>The effect data is partitioned for the decision tree.
9. The method according to claim 1, wherein the specific step of step S5 is:
step S51: based on decision result data, carrying out model parameter adjustment by using a model optimization algorithm to generate a comparison index;
step S52: acquiring a service scene, and performing intelligent equipment connection by using a 5G network based on the service scene and the comparison index to generate an index processing method model;
the model optimization algorithm specifically comprises the following steps:
;
wherein ,for comparison index, ->Representing the number of samples of the training dataset, +.>Representing the number of features->Indicate->Feature vector of individual samples, +_>Indicate->Target variable of individual samples, +.>Representing model parameter vector, ++>Predictive value representing a logistic regression model, +.>For regularization parameters, ++>Representing the loss function pair->Partial derivatives of the individual parameters;
step S53: acquiring output data based on the index processing method model, and generating model output data;
step S54: based on the model output data and the API direct connection interface data, utilizing the 5G network intelligent equipment to conduct scene data prediction output, and generating real-time equipment output data;
step S55: model data feedback is carried out based on real-time equipment output data, and 5G intelligent hospital basic data processing is achieved.
10. A 5G intelligent hospital base data processing database platform, comprising:
at least one processor;
a memory coupled to the at least one processor;
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the 5G intelligent hospital base data processing method according to any one of claims 1 to 9.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310827489.9A CN116579503B (en) | 2023-07-07 | 2023-07-07 | 5G intelligent hospital basic data processing method and database platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310827489.9A CN116579503B (en) | 2023-07-07 | 2023-07-07 | 5G intelligent hospital basic data processing method and database platform |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116579503A CN116579503A (en) | 2023-08-11 |
CN116579503B true CN116579503B (en) | 2023-09-08 |
Family
ID=87536130
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310827489.9A Active CN116579503B (en) | 2023-07-07 | 2023-07-07 | 5G intelligent hospital basic data processing method and database platform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116579503B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117076454B (en) * | 2023-08-21 | 2024-03-12 | 广州地铁集团有限公司 | Engineering quality acceptance form data structured storage method and system |
CN117910850A (en) * | 2023-12-18 | 2024-04-19 | 北京宇信科技集团股份有限公司 | Index data analysis engine, index data calculation device and calculation method |
CN118173214B (en) * | 2024-05-10 | 2024-07-19 | 湖南尚医康医疗科技有限公司 | Intelligent communication interaction method and system for medical information |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108766542A (en) * | 2018-05-28 | 2018-11-06 | 镇江市第人民医院 | A kind of data analysis processing method and system |
CN111984408A (en) * | 2020-08-14 | 2020-11-24 | 薛亮 | Data cooperative processing method based on big data and edge computing and edge cloud platform |
CN115662560A (en) * | 2022-10-31 | 2023-01-31 | 单亦文 | Internet-based intelligent hospital management method and system |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11182177B2 (en) * | 2020-03-30 | 2021-11-23 | Hayssam Hamze | Dynamic modeler |
-
2023
- 2023-07-07 CN CN202310827489.9A patent/CN116579503B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108766542A (en) * | 2018-05-28 | 2018-11-06 | 镇江市第人民医院 | A kind of data analysis processing method and system |
CN111984408A (en) * | 2020-08-14 | 2020-11-24 | 薛亮 | Data cooperative processing method based on big data and edge computing and edge cloud platform |
CN115662560A (en) * | 2022-10-31 | 2023-01-31 | 单亦文 | Internet-based intelligent hospital management method and system |
Also Published As
Publication number | Publication date |
---|---|
CN116579503A (en) | 2023-08-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116579503B (en) | 5G intelligent hospital basic data processing method and database platform | |
US10127477B2 (en) | Distributed event prediction and machine learning object recognition system | |
US9367799B2 (en) | Neural network based cluster visualization that computes pairwise distances between centroid locations, and determines a projected centroid location in a multidimensional space | |
US10832174B1 (en) | Distributed hyperparameter tuning system for active machine learning | |
US10929762B1 (en) | Distributable event prediction and machine learning recognition system | |
US10883345B2 (en) | Processing of computer log messages for visualization and retrieval | |
RU2693324C2 (en) | Method and a server for converting a categorical factor value into its numerical representation | |
US10956825B1 (en) | Distributable event prediction and machine learning recognition system | |
CN117151870B (en) | Portrait behavior analysis method and system based on guest group | |
CN117670066B (en) | Questor management method, system, equipment and storage medium based on intelligent decision | |
CN108665148B (en) | Electronic resource quality evaluation method and device and storage medium | |
CN113610240A (en) | Method and system for performing predictions using nested machine learning models | |
CN113707323B (en) | Disease prediction method, device, equipment and medium based on machine learning | |
CN109389424B (en) | Flow distribution method and device, electronic equipment and storage medium | |
CN116611546B (en) | Knowledge-graph-based landslide prediction method and system for target research area | |
CN116861331A (en) | Expert model decision-fused data identification method and system | |
Talagala et al. | Meta‐learning how to forecast time series | |
CN117436768A (en) | Unified supervision index method based on data management | |
Li et al. | An improved genetic-XGBoost classifier for customer consumption behavior prediction | |
CN116504401A (en) | Intelligent physical examination project recommendation method and device | |
CN116257758A (en) | Model training method, crowd expanding method, medium, device and computing equipment | |
CN113590691A (en) | Target object processing method and device | |
CN118193855B (en) | Online response method and device and computer equipment | |
CN116414360B (en) | Artificial intelligence-based application system integrated management method and system | |
Ingram et al. | Glint: An MDS Framework for Costly Distance Functions. |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |