WO2023155441A1

WO2023155441A1 - Medical resource recommendation method and apparatus, device, and storage medium

Info

Publication number: WO2023155441A1
Application number: PCT/CN2022/122991
Authority: WO
Inventors: 候学智
Original assignee: 康键信息技术（深圳）有限公司
Priority date: 2022-02-15
Filing date: 2022-09-30
Publication date: 2023-08-24
Also published as: CN114420279A

Abstract

A medical resource recommendation method, comprising: using historical patient treatment data to extract an association feature between a historical patient information feature set and a historical medical resource feature set; using the historical patient information feature set, the historical medical resource feature set, and the association feature to construct a historical patient medical recommendation sample set; training a pre-constructed medical resource recommendation model on the basis of the historical patient medical recommendation sample set to obtain a trained medical resource recommendation model; and using the trained medical resource recommendation model to perform matching analysis on patient information to be analyzed and available medical resources to obtain recommended medical resource data. In addition, the present application further relates to the blockchain technology, and the historical patient medical recommendation sample set may be stored in a node of a blockchain. Also provided are a medical resource recommendation apparatus, an electronic device, and a storage medium. The medical resource recommendation accuracy can be improved.

Description

Medical resource recommendation method, device, equipment and storage medium

This application claims the priority of the Chinese patent application with the application number 202210137916.6 filed on February 15, 2022, and the title of the invention is "medical resource recommendation method, device, equipment and storage medium", the entire content of which is incorporated by reference in this application.

technical field

The present application relates to the technical field of artificial intelligence, and in particular to a medical resource recommendation method, device, electronic equipment, and computer-readable storage medium.

Background technique

Seeking a doctor is a common thing in people's life. It is very important to choose the right hospital and doctor. However, the inventor realized that most patients cannot correctly evaluate their own symptoms due to the limitation of medical knowledge. At the same time, the Internet information is huge and complex. Patients have no way of distinguishing authenticity from falsehoods, and due to lack of medical expertise, it is also difficult to find hospitals and doctors who are good at their own conditions, thereby reducing medical efficiency and wasting medical resources. By configuring the medical recommendation system, it can help patients and doctors greatly shorten the time of seeing a doctor and save manpower and material resources.

Existing medical recommendation systems often use fixed search methods, or simply use historical interaction information between doctors and patients as input to recommend relevant medical information. The medical resources recommended by this method are not accurate enough.

Contents of the invention

In order to achieve the above purpose, a method for recommending medical resources provided by this application includes:

Obtaining a historical patient data set and a historical medical resource set, extracting a historical patient information feature set and historical patient visit data from the historical patient data set, and extracting a historical medical resource feature set from the historical medical resource set;

Using the historical patient visit data, extract the associated features between the historical patient information feature set and the historical medical resource feature set, use the historical patient information feature set, the historical medical resource feature set and the association features, constructing a medical recommendation sample set for historical patients;

Based on the historical patient medical recommendation sample set, using the contrastive divergence algorithm to train the pre-built medical resource recommendation model to obtain the trained medical resource recommendation model;

Obtain patient information to be analyzed and available medical resources, use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources, and obtain recommended medical resource data.

The present application also provides a device for recommending medical resources, the device comprising:

The feature extraction module is used to obtain historical patient data sets and historical medical resource sets, extract historical patient information feature sets and historical patient visit data from the historical patient data sets, and extract historical medical resource feature sets from the historical medical resource sets ;

The sample construction module is used to use the historical patient visit data to extract the associated features between the historical patient information feature set and the historical medical resource feature set, and use the historical patient information feature set and the historical medical resource feature set The feature set and the associated features are used to construct a medical recommendation sample set for historical patients;

The model training module is used to train the pre-built medical resource recommendation model based on the historical patient medical recommendation sample set using the contrastive divergence algorithm to obtain the trained medical resource recommendation model;

The recommendation module is used to obtain the patient information to be analyzed and the available medical resources, and use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and the available medical resources to obtain recommended medical resource data.

The present application also provides an electronic device, the electronic device comprising:

at least one processor; and,

a memory communicatively coupled to the at least one processor; wherein,

The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor, so that the at least one processor can execute the medical resource recommendation method as follows:

The present application also provides a computer-readable storage medium, at least one computer program is stored in the computer-readable storage medium, and the at least one computer program is executed by a processor in an electronic device to implement medical resource recommendation as described below method:

Description of drawings

FIG. 1 is a schematic flowchart of a method for recommending medical resources provided by an embodiment of the present application;

FIG. 2 is a schematic diagram of a detailed implementation process of one of the steps in the medical resource recommendation method shown in FIG. 1;

FIG. 3 is a schematic diagram of a detailed implementation process of another step in the method for recommending medical resources shown in FIG. 1;

FIG. 4 is a functional block diagram of a medical resource recommendation device provided by an embodiment of the present application;

FIG. 5 is a schematic structural diagram of an electronic device implementing the medical resource recommendation method provided by an embodiment of the present application.

The realization, functional features and advantages of the present application will be further described in conjunction with the embodiments and with reference to the accompanying drawings.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present application, and are not intended to limit the present application.

An embodiment of the present application provides a method for recommending medical resources. The subject of execution of the method for recommending medical resources includes, but is not limited to, at least one of electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for recommending medical resources can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server can be an independent server, or it can provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery network (ContentDelivery Network) , CDN), and cloud servers for basic cloud computing services such as big data and artificial intelligence platforms.

Referring to FIG. 1 , it is a schematic flowchart of a method for recommending medical resources provided by an embodiment of the present application. In this embodiment, the method for recommending medical resources includes:

S1. Acquire historical patient data sets and historical medical resource sets, extract historical patient information feature sets and historical patient visit data from the historical patient data sets, and extract historical medical resource feature sets from the historical medical resource sets.

In one embodiment of the present application, the historical patient data set includes: historical patient medical record data, and historical patient visit data.

In the embodiment of the present application, the historical patient medical record data includes the patient's basic information, disease type, symptom description, physical condition, previous treatment conditions, pathological type, etc. Among them, the basic information includes: name, age, weight, gender, etc.; the types of diseases include: cardiovascular disease, lung disease, rheumatic immune system disease, urinary system disease, kidney disease, diabetes and endocrine system disease, ENT diseases, digestive system diseases, gynecological diseases, nervous system diseases, eye diseases, orthopedic diseases, etc.; the description of the symptoms is the abnormal sensation of the patient himself, such as dizziness, vertigo, wheezing, coughing, etc.; According to the Eastern Cooperative Oncology Group (ECOG) score standard, the ECOG score is an index to understand the patient's general health status and tolerance to treatment from the patient's physical strength, wherein, 0 points indicate that the activity ability is completely normal, and There was no difference in activity ability before the onset of the disease; 1 point means that you can walk around freely and engage in light physical activities, including general housework or office work, but you cannot engage in heavy physical activities; 2 points means you can move around freely and take care of yourself, but you have lost Ability to work, able to get up and move about at least half of the time during the day; 3 points means that they can only take care of themselves partially, and more than half of the time during the day is bedridden or in a wheelchair; 4 points means that they are bedridden and unable to take care of themselves; the previous treatment status is The patient's previous treatment conditions, including whether they have received surgical treatment, etc.

Further, the historical patient visit data includes the hospital data and doctor data that the historical patient actually visited finally, such as hospital name, hospital ranking, hospital grade, doctor name, doctor's department, doctor's current position, doctor's specialty, etc.

Further, the historical medical resource set includes hospital data disclosed by domestic and foreign hospital official websites and authoritative websites, and doctor data disclosed by domestic and foreign hospital official websites and authoritative websites.

Among them, the hospital data may specifically include hospital rankings, whether the hospital participates in formulating disease-related diagnosis and treatment guidelines, whether the hospital has doctors working in disease-related professional academic institutions, the number of disease-related papers published by the hospital, and disease-related clinical trials conducted by the hospital. The number of doctors and the number of expert teams in the hospital; the doctor data can include the name of the doctor, the center/project to which the doctor belongs, the department to which the doctor belongs, the current position of the doctor, the academic title of the doctor, the clinical expertise of the doctor, whether the doctor participates in the formulation of disease-related diagnosis and treatment guidelines, whether the doctor Affiliations in disease-related associations and foundations, number of disease-related papers published by physicians, etc.

In one of the embodiments of the present application, the historical patient data set may include various data forms such as text and images. Therefore, after obtaining the historical patient data set, it is necessary to use existing text conversion technology (such as OCR recognition technologies, etc.), and convert historical patient data sets in various data forms into text form in order to extract historical patient information feature sets and historical patient visit data.

In detail, the feature set of historical patient information extracted from the historical patient data set in S1 includes:

Converting the historical patient data set into a text format uniformly to obtain the historical patient text data set;

Carrying out word segmentation and part-of-speech tagging on the historical patient text data set to obtain the results of word segmentation and part-of-speech tagging;

Nouns and noun phrases are extracted from the word segmentation according to the results of the part-of-speech tagging, and according to the nouns and noun phrases, statistically obtain historical patient information characteristic frequencies, and generate frequent pattern trees according to the historical patient information characteristic frequencies;

identifying features in the frequent pattern tree to obtain a feature set of candidate historical patient information;

Calculate the point mutual information value of each feature in the candidate historical patient information feature set, and filter out the historical patient information features whose point mutual information value is less than a preset standard threshold from the candidate historical patient information feature set to obtain the historical patient information feature set.

In one of the embodiments of the present application, the following algorithm can be used to calculate the point mutual information value PMI of each feature in the candidate historical patient information feature set:

Among them, the historical patient represents the name of the historical patient, and the feature represents each feature in the candidate historical patient information feature set; the hit ("historical patient") represents the number obtained by searching the historical patient; hit ("feature") represents the search for the feature The number obtained; hit("historical patient" and "feature") indicates the number obtained by searching the feature and the historical patient; the higher the point mutual information value, the higher the degree of association between the historical patient and the feature, when the When the point mutual information value is lower than the preset standard threshold, filter out the corresponding historical patient information features; when the point mutual information value is higher than the preset standard threshold, keep the corresponding historical patient information features, and combine to obtain the Describe the feature set of historical patient information.

Further, in this embodiment of the present application, the same method as that for extracting the feature set of historical patient information can be used to extract the historical patient visit data and the feature set of historical medical resources, which will not be repeated here.

S2. Using the historical patient visit data, extract the associated features between the historical patient information feature set and the historical medical resource feature set, and use the historical patient information feature set, the historical medical resource feature set and all Based on the above association features, a medical recommendation sample set for historical patients is constructed.

Specifically, in S2, using the historical patient visit data to extract the associated features between the historical patient information feature set and the historical medical resource feature set includes:

Segmenting each historical patient information feature in the historical patient information feature set and each historical medical resource feature in the historical medical resource feature set, and extracting key words and relational words from the word segmentation results;

The key words and relational words irrelevant to the historical patient visit data are deleted to obtain the associated features between the historical patient information feature set and the historical medical resource feature set.

Specifically, the embodiment of the present application deletes key words and relational words that are not related to the historical patient visit data, and deletes features with too much missing data in historical patient information features and historical medical resource features, so as to screen out key words and phrases that are key to medical resource recommendation. The characteristics of the function, such as the pathological type of the patient, whether the doctor participates in the formulation of disease-related diagnosis and treatment guidelines, etc., the selected features that play a key role in the recommendation of medical resources are used as the difference between the historical patient information feature set and the historical medical resource feature set. interrelationship features.

Further, in the embodiment of the present application, before constructing the historical patient medical recommendation sample set, it is necessary to unify the representation forms of the features in the historical patient information feature set, the historical medical resource feature set, and the associated features, for example, in the Among the historical patient information feature set, the historical medical resource feature set, and the associated features, some of the time representations are "****year**month**day**hour**minute**second", and some The time representation form is "**** year/** month/** day ** hour ** minute ** second", then the embodiment of this application can unify the time representation form as "**** year* *Month**day**hour**minute**second"; for another example, some of the names of hospitals in the feature are abbreviated names of hospitals, and some are full names of hospitals, so all the names of hospitals can be unified as hospitals Full name.

S3. Based on the historical patient medical recommendation sample set, use the contrastive divergence algorithm to train the pre-built medical resource recommendation model to obtain a trained medical resource recommendation model.

In the implementation of this application, the medical resource recommendation model is constructed based on a multi-layer restricted Boltzmann machine (RBM for short). Wherein, the RBM is a probability generation model based on deep learning, comprising a visible layer and a hidden layer, the visible layer is composed of a plurality of visible layer neurons, and the hidden layer is composed of a plurality of hidden layers Composed of neurons, there is a connection between the visible layer and the hidden layer, but there is no connection between the units in the layer, and the multi-layer restricted Boltzmann machine network connection is determined by the weight value under self-orientation. In one of the embodiments of the present application, the multilayer restricted Boltzmann machine may be composed of three RBMs.

In detail, referring to Figure 2, the S3 includes:

S31. Divide the historical patient medical recommendation sample set into a training set and a test set;

S32. According to the training set, use the contrastive divergence algorithm to adjust the parameters of the medical resource recommendation model, and iteratively train the medical resource recommendation model to obtain a trained medical resource recommendation model;

S33. Use the test set to test and adjust the trained medical resource recommendation model to obtain a trained medical resource recommendation model.

In one embodiment of the present application, 80% of the samples in the historical patient medical recommendation sample set may be divided into a training set, and the remaining 20% of the samples may be divided into a test set.

The contrastive divergence algorithm (contrastive divergence, CD algorithm for short) described in the embodiment of the present application reconstructs the value of the neuron of the visible layer through the value of the neuron of the hidden layer, according to the initial value of the neuron of the visible layer The parameters of the medical resource recommendation model are adjusted according to the errors of the reconstructed values of neurons in the visible layer.

Further, referring to Fig. 3, the S32 includes:

S321. Convert the historical patient information feature set, the historical medical resource feature set, and the associated features in the training set into feature vectors, and perform mean value, variance, and whitening processing on the feature vectors, and calculate all The processed eigenvectors are evenly normalized to between 0 and 1 in order of small to large, to obtain a normalized sample set;

S322. Initialize the weight value between the neurons in the visible layer and the neurons in the hidden layer in the medical resource recommendation model, and the bias parameters of the neurons in the visible layer and the bias parameters of the neurons in the hidden layer;

S323. Using the normalized sample set, the weight value, the bias parameter of the neuron in the visible layer, and the bias parameter of the neuron in the hidden layer, calculate the neuron in the visible layer and the neuron in the hidden layer. The hidden layer neuron is looped and iterated, and the activation probability of the hidden layer neuron is calculated;

In one embodiment of the present application, the activation probability of the hidden layer neurons can be calculated by the following formula:

Among them, f() represents the sigmoid function, h _j represents the value of the jth neuron in the hidden layer, v represents the value of the visible layer neuron, θ=(w _ij , b _i , b _c ), c _j represents the The bias parameter of the jth neuron in the hidden layer, w _ij represents the weight value between the jth hidden layer neuron and the i-th visible layer neuron, b _i represents the weight value of the i-th neuron in the visible layer Bias parameter, v _i represents the value of the ith neuron in the visible layer.

S324. Using the hidden layer neurons after the loop iteration, iterate the visible layer neurons in a reverse loop, and calculate the activation probability of the visible layer neurons;

In one embodiment of the present application, the activation probability of neurons in the visible layer can be calculated by the following formula:

Where f() represents the sigmoid function, v _i represents the value of the ith neuron in the visible layer, h represents the value of the neuron in the hidden layer, θ=(w _ij , bi _, c _j ), c _j represents the hidden The bias parameter of the j-th neuron in the layer, w _ij represents the weight value between the hidden layer neuron j and the visible layer neuron i, b _i represents the bias parameter of the i-th neuron in the visible layer, b _j represents the value of the jth neuron of the visible layer.

S325. According to the activation probability of the neuron in the hidden layer and the activation probability of the neuron in the visible layer, use a contrastive divergence algorithm to calculate the weight value, the bias parameter of the neuron in the visible layer, and the hidden adjust the bias parameters of the neurons in the visible layer, use the adjusted weight value and the bias parameters to recalculate the activation probability of the neurons in the visible layer, and according to the activation probability of the neurons in the visible layer The output of the medical resource recommendation model corresponding to the maximum value is used as the predicted recommended medical resource;

S326. Using a loss function to calculate the loss value between the predicted and recommended medical resources and the historical patient visit data, and determine whether the loss value is less than a preset loss threshold;

When the loss value is greater than or equal to the preset loss threshold, execute S327, adjust the parameters of the medical resource recommendation model, and return to the step corresponding to S322;

When the loss value is less than the preset loss threshold, S328 is executed to obtain and output the trained medical resource recommendation model.

Further, the use of contrastive divergence algorithm described in S325 to adjust the weight value, the bias parameter of the visible layer neuron and the bias parameter of the hidden layer neuron, including:

Step A, using the normalized sample set to initialize the visible layer neuron, the state vector of the hidden layer neuron, the weight value between the visible layer neuron and the hidden layer neuron ;

Step B, using the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer to perform K-step Gibbs sampling to obtain the state vector and hidden The state vector of neurons in the layer, the state vector of neurons in the visible layer corresponding to time t, and the state vector of neurons in the hidden layer;

The Gibbs sampling (Gibbs sampling) described in the embodiment of the present application is an algorithm used in Markov Monte Carlo (MCMC) in statistics, which is used to extract from a certain multivariate probability distribution when it is difficult to sample directly. Approximately draws a sequence of samples.

Step C, using the state vector of the visible layer neuron and the state vector of the hidden layer neuron corresponding to the time t-1, the state vector of the visible layer neuron and the state vector of the hidden layer neuron corresponding to the time t, performing circular calculation on the normalized sample set to obtain a comparison error value;

Step D, using the comparison error value to adjust the weight value, the bias parameter of the neuron in the visible layer, and the bias parameter of the neuron in the hidden layer.

The embodiment of the present application uses the medical resource recommendation model trained by the contrastive divergence algorithm, so that the update direction of the model parameters depends on the result of the previous iteration, so that the algorithm jumps out of the local optimum, speeds up the model convergence speed, and greatly reduces the model iteration. The number of times greatly improves the recommendation efficiency and improves the accuracy of medical resource recommendation.

S4. Obtain patient information to be analyzed and available medical resources, and use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources to obtain recommended medical resource data.

In the embodiment of the present application, the available medical resources include hospital data published by domestic and foreign hospital official websites and authoritative websites, and doctor data published by domestic and foreign hospital official websites and authoritative websites.

In detail, said S4 includes:

Obtain patient information to be analyzed and available medical resources uploaded by the client;

extracting the characteristics of the patient information to be analyzed from the patient information to be analyzed and the characteristics of available medical resources from the available medical resources;

Using the trained medical resource recommendation model to perform matching analysis on the characteristics of the patient information to be analyzed and the characteristics of the available medical resources to obtain recommended medical resource data;

Feedback the recommended medical resource data to the client.

In one implementation of the present application, the patient to be analyzed can enter the relevant case data through the label filter items related to the condition set on the user terminal to obtain the patient's information to be analyzed.

In one implementation of the present application, there may be multiple options for hospital data and doctor data in the recommended medical resource data, and the current user can re-screen according to his actual needs.

In the embodiment of the present application, the historical patient information feature set and the historical medical resource feature set are respectively extracted from the historical patient data set and the historical medical resource set; the historical patient information feature set and the historical patient information feature set are extracted using the historical patient data The associated features between the historical medical resource feature sets, using the historical patient information feature set, the historical medical resource feature set, and the associated features, construct a historical patient medical recommendation sample set, which is conducive to improving the relationship between patients and recommended medical treatment. The matching degree of resources improves the accuracy of recommended medical resources; the medical resource recommendation model is obtained by using contrastive divergence algorithm training, which speeds up model convergence, reduces the number of model iterations, and improves the accuracy of medical resource recommendation. Therefore, the medical resource recommendation method proposed in this application can solve the problem of low accuracy in medical resource recommendation.

As shown in FIG. 4 , it is a functional block diagram of a medical resource recommendation device provided by an embodiment of the present application. The medical resource recommendation apparatus 100 described in this application may be installed in electronic equipment. According to the functions realized, the medical resource recommendation device 100 may include a feature extraction module 101 , a sample construction module 102 , a model training module 103 and a recommendation module 104 . The module described in this application can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of the electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

In this embodiment, the functions of each module/unit are as follows:

The feature extraction module 101 is configured to acquire historical patient data sets and historical medical resource sets, extract historical patient information feature sets and historical patient visit data from the historical patient data sets, and extract historical medical treatment data from the historical medical resource sets. resource feature set;

The sample construction module 102 is configured to use the historical patient visit data to extract the associated features between the historical patient information feature set and the historical medical resource feature set, and use the historical patient information feature set, the The historical medical resource feature set and the associated features are used to construct a historical patient medical recommendation sample set;

The model training module 103 is used to train the pre-built medical resource recommendation model based on the historical patient medical recommendation sample set by using the contrastive divergence algorithm, so as to obtain the trained medical resource recommendation model;

The recommendation module 104 is used to obtain patient information to be analyzed and available medical resources, and use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources to obtain recommended medical resource data.

In detail, each module described in the medical resource recommendation device 100 in the embodiment of the present application adopts the same technical means as the medical resource recommendation method described in the above-mentioned Fig. 1 to Fig. 3 , and can generate the same The technical effect will not be repeated here.

As shown in FIG. 5 , it is a schematic structural diagram of an electronic device implementing a method for recommending medical resources provided by an embodiment of the present application.

The electronic device 1 may include a processor 10, a memory 11, a communication bus 12, and a communication interface 13, and may also include computer programs stored in the memory 11 and operable on the processor 10, such as medical resource recommendation program.

Wherein, the processor 10 may be composed of integrated circuits in some embodiments, for example, may be composed of a single packaged integrated circuit, or may be composed of multiple integrated circuits with the same function or different functions, including one or A combination of multiple central processing units (Central Processing unit, CPU), microprocessors, digital processing chips, graphics processors and various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect the various components of the entire electronic device, by running or executing programs or modules stored in the memory 11 (for example, executing medical resource recommendation program, etc.), and call the data stored in the memory 11 to execute various functions of the electronic device and process data.

The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card type memory (for example: SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. , the computer-readable storage medium may be non-volatile or volatile. The storage 11 may be an internal storage unit of the electronic device in some embodiments, such as a mobile hard disk of the electronic device. The memory 11 can also be an external storage device of an electronic device in other embodiments, such as a plug-in mobile hard disk equipped on an electronic device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD ) card, flash card (Flash Card), etc. Further, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 can not only be used to store application software and various data installed in the electronic device, such as the code of the medical resource recommendation program, etc., but also can be used to temporarily store the data that has been output or will be output.

The communication bus 12 may be a peripheral component interconnect (PCI for short) bus or an extended industry standard architecture (EISA for short) bus or the like. The bus can be divided into address bus, data bus, control bus and so on. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10 and the like.

The communication interface 13 is used for communication between the electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which are generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display) or an input unit (such as a keyboard (Keyboard)). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, and the like. Wherein, the display may also be properly referred to as a display screen or a display unit, and is used for displaying information processed in the electronic device and for displaying a visualized user interface.

FIG. 5 only shows an electronic device with components. Those skilled in the art can understand that the structure shown in FIG. 5 does not constitute a limitation to the electronic device 1, and may include fewer or more components, or combinations of certain components, or different arrangements of components.

For example, although not shown, the electronic device may also include a power supply (such as a battery) for supplying power to various components. Preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so that Realize functions such as charge management, discharge management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power failure detection circuits, power converters or inverters, power status indicators and other arbitrary components. The electronic device may also include various sensors, a Bluetooth module, a Wi-Fi module, etc., which will not be repeated here.

It should be understood that the embodiments are only for illustration, and are not limited by the structure in the scope of the patent application.

The medical resource recommendation program stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can realize:

Specifically, for the specific implementation method of the above instructions by the processor 10, reference may be made to the description of relevant steps in the corresponding embodiments in the drawings, and details are not repeated here.

Further, if the integrated modules/units of the electronic device 1 are realized in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory).

The present application also provides a computer-readable storage medium, the readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, it can realize:

Obtain patient information to be analyzed and available medical resources, use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources, and obtain recommended medical resource data. In the several embodiments provided in this application, it should be understood that the disclosed devices, devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional module in each embodiment of the present application may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software function modules.

It will be apparent to those skilled in the art that the present application is not limited to the details of the exemplary embodiments described above, but that the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

Therefore, the embodiments should be regarded as exemplary and not restrictive in all points of view, and the scope of the application is defined by the appended claims rather than the foregoing description, and it is intended that the scope of the present application be defined by the appended claims rather than by the foregoing description. All changes within the meaning and range of equivalents of the elements are embraced in this application. Any reference sign in a claim should not be construed as limiting the claim concerned.

The blockchain referred to in this application is a new application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain (Blockchain), essentially a decentralized database, is a series of data blocks associated with each other using cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify its Validity of information (anti-counterfeiting) and generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

The embodiments of the present application may acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. .

In addition, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or devices stated in the system claims may also be realized by one unit or device through software or hardware. The terms first, second, etc. are used to denote names and do not imply any particular order.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application without limitation. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be Make modifications or equivalent replacements without departing from the spirit and scope of the technical solutions of the present application.

Claims

A method for recommending medical resources, wherein the method includes:

Obtaining a historical patient data set and a historical medical resource set, extracting a historical patient information feature set and historical patient visit data from the historical patient data set, and extracting a historical medical resource feature set from the historical medical resource set;

Using the historical patient visit data, extract the associated features between the historical patient information feature set and the historical medical resource feature set, use the historical patient information feature set, the historical medical resource feature set and the association features, constructing a medical recommendation sample set for historical patients;

Based on the historical patient medical recommendation sample set, using the contrastive divergence algorithm to train the pre-built medical resource recommendation model to obtain the trained medical resource recommendation model;

Obtain patient information to be analyzed and available medical resources, use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources, and obtain recommended medical resource data.
The medical resource recommendation method according to claim 1, wherein said extracting a feature set of historical patient information from said historical patient data set comprises:

Converting the historical patient data set into a text format uniformly to obtain the historical patient text data set;

Carrying out word segmentation and part-of-speech tagging on the historical patient text data set to obtain the results of word segmentation and part-of-speech tagging;

Nouns and noun phrases are extracted from the word segmentation according to the results of the part-of-speech tagging, and according to the nouns and noun phrases, statistically obtain historical patient information characteristic frequencies, and generate frequent pattern trees according to the historical patient information characteristic frequencies;

identifying features in the frequent pattern tree to obtain a feature set of candidate historical patient information;

Calculate the point mutual information value of each feature in the candidate historical patient information feature set, and filter out the historical patient information features whose point mutual information value is less than a preset standard threshold from the candidate historical patient information feature set to obtain the historical patient information feature set.
The method for recommending medical resources according to claim 1, wherein said using said historical patient visit data to extract the associated features between said historical patient information feature set and said historical medical resource feature set comprises:

Segmenting each historical patient information feature in the historical patient information feature set and each historical medical resource feature in the historical medical resource feature set, and extracting key words and relational words from the word segmentation results;

The key words and relational words irrelevant to the historical patient visit data are deleted to obtain the associated features between the historical patient information feature set and the historical medical resource feature set.
The medical resource recommendation method according to claim 1, wherein, based on the historical patient medical recommendation sample set, the pre-built medical resource recommendation model is trained using a contrastive divergence algorithm to obtain the trained medical resource recommendation models, including:

Dividing the historical patient medical recommendation sample set into a training set and a test set;

According to the training set, using a contrastive divergence algorithm to adjust the parameters of the medical resource recommendation model, and iteratively training the medical resource recommendation model to obtain a trained medical resource recommendation model;

The test set is used to test and adjust the trained medical resource recommendation model to obtain a trained medical resource recommendation model.
The medical resource recommendation method according to claim 4, wherein, according to the training set, the parameters of the medical resource recommendation model are adjusted using a contrastive divergence algorithm, and the medical resource recommendation model is iteratively trained to obtain the Trained medical resource recommendation models, including:

Convert the historical patient information feature set, the historical medical resource feature set, and the associated features in the training set into feature vectors, and perform mean value, variance, and whitening processing on the feature vectors, and process the processed The eigenvectors of are uniformly normalized to between 0 and 1 in order from small to large to obtain a normalized sample set;

Initializing the weight value between the neurons of the visible layer and the neurons of the hidden layer in the medical resource recommendation model, and the bias parameters of the neurons of the visible layer and the bias parameters of the neurons of the hidden layer;

Using the normalized sample set, the weight value, the bias parameter of the neuron in the visible layer and the bias parameter of the neuron in the hidden layer, the neuron in the visible layer and the neuron in the hidden layer Layer neurons perform loop iterations, and calculate the activation probability of the hidden layer neurons;

Using the hidden layer neurons after the loop iteration, iterate the visible layer neurons in a reverse cycle, and calculate the activation probability of the visible layer neurons;

According to the activation probability of the neuron in the hidden layer and the activation probability of the neuron in the visible layer, use a contrastive divergence algorithm to adjust the weight value and the bias parameter, and use the adjusted weight value and The bias parameter is to recalculate the activation probability of the visible layer neurons, and use the output result of the medical resource recommendation model corresponding to the maximum value of the activation probability of the visible layer neurons as the predicted recommended medical resources;

Using a loss function to calculate the loss value between the predicted and recommended medical resources and the historical patient visit data, and adjust the parameters of the medical resource recommendation model according to the loss value, and return to the initialization of the medical resource recommendation model The step of weighting the neurons in the visible layer and the neurons in the hidden layer until the loss value is less than a preset loss threshold, to obtain a trained medical resource recommendation model.
The method for recommending medical resources as claimed in claim 5, wherein, according to the activation probability of neurons in the hidden layer and the activation probability of neurons in the visible layer, the weight value and The bias parameters to adjust include:

Using the normalized sample set to initialize the visible layer neuron, the state vector of the hidden layer neuron, the weight value between the visible layer neuron and the hidden layer neuron;

Utilize the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer to perform K-step Gibbs sampling to obtain the state vector of the neurons in the visible layer and the neurons in the hidden layer corresponding to the t-1 moment The state vector of , the state vector of the visible layer neuron and the state vector of the hidden layer neuron corresponding to time t;

Utilize the state vector of the visible layer neuron corresponding to the t-1 moment and the state vector of the hidden layer neuron, the state vector of the visible layer neuron corresponding to the t moment and the state vector of the hidden layer neuron, to the described The normalized sample set is cyclically calculated to obtain the comparison error value;

The weight value, the bias parameter of the visible layer neuron and the bias parameter of the hidden layer neuron are adjusted by using the comparison error value.
The method for recommending medical resources according to any one of claims 1 to 6, wherein said acquisition of patient information to be analyzed and available medical resources is performed using the trained medical resource recommendation model for said patient information to be analyzed and Available medical resources can be matched and analyzed to obtain recommended medical resource data, including:

Obtain patient information to be analyzed and available medical resources uploaded by the client;

extracting the characteristics of the patient information to be analyzed from the patient information to be analyzed and the characteristics of available medical resources from the available medical resources;

Using the trained medical resource recommendation model to perform matching analysis on the characteristics of the patient information to be analyzed and the characteristics of the available medical resources to obtain recommended medical resource data;

Feedback the recommended medical resource data to the client.
A device for recommending medical resources, wherein the device includes:

The feature extraction module is used to obtain historical patient data sets and historical medical resource sets, extract historical patient information feature sets and historical patient visit data from the historical patient data sets, and extract historical medical resource feature sets from the historical medical resource sets ;

The sample construction module is used to use the historical patient visit data to extract the associated features between the historical patient information feature set and the historical medical resource feature set, and use the historical patient information feature set and the historical medical resource feature set The feature set and the associated features are used to construct a medical recommendation sample set for historical patients;

The model training module is used to train the pre-built medical resource recommendation model based on the historical patient medical recommendation sample set using the contrastive divergence algorithm to obtain the trained medical resource recommendation model;

The recommendation module is used to obtain the patient information to be analyzed and the available medical resources, and use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and the available medical resources to obtain recommended medical resource data.
An electronic device, wherein the electronic device includes:

at least one processor; and,

a memory communicatively coupled to the at least one processor; wherein,

The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor, so that the at least one processor can execute the medical resource recommendation method as follows:

Obtaining a historical patient data set and a historical medical resource set, extracting a historical patient information feature set and historical patient visit data from the historical patient data set, and extracting a historical medical resource feature set from the historical medical resource set;

Using the historical patient visit data, extract the associated features between the historical patient information feature set and the historical medical resource feature set, use the historical patient information feature set, the historical medical resource feature set and the association features, constructing a medical recommendation sample set for historical patients;

Based on the historical patient medical recommendation sample set, using the contrastive divergence algorithm to train the pre-built medical resource recommendation model to obtain the trained medical resource recommendation model;

Obtain patient information to be analyzed and available medical resources, use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources, and obtain recommended medical resource data.
The electronic device according to claim 9, wherein said extracting a feature set of historical patient information from said historical patient data set comprises:

Converting the historical patient data set into a text format uniformly to obtain the historical patient text data set;

Carrying out word segmentation and part-of-speech tagging on the historical patient text data set to obtain the results of word segmentation and part-of-speech tagging;

Nouns and noun phrases are extracted from the word segmentation according to the results of the part-of-speech tagging, and according to the nouns and noun phrases, statistically obtain historical patient information characteristic frequencies, and generate frequent pattern trees according to the historical patient information characteristic frequencies;

identifying features in the frequent pattern tree to obtain a feature set of candidate historical patient information;

Calculate the point mutual information value of each feature in the candidate historical patient information feature set, and filter out the historical patient information features whose point mutual information value is less than a preset standard threshold from the candidate historical patient information feature set to obtain the historical patient information feature set.
The electronic device according to claim 9, wherein said extracting the associated features between the historical patient information feature set and the historical medical resource feature set by using the historical patient visit data comprises:

Segmenting each historical patient information feature in the historical patient information feature set and each historical medical resource feature in the historical medical resource feature set, and extracting key words and relational words from the word segmentation results;

The key words and relational words irrelevant to the historical patient visit data are deleted to obtain the associated features between the historical patient information feature set and the historical medical resource feature set.
The electronic device according to claim 9, wherein, based on the historical patient medical recommendation sample set, the pre-built medical resource recommendation model is trained using a contrastive divergence algorithm to obtain a trained medical resource recommendation model, include:

Dividing the historical patient medical recommendation sample set into a training set and a test set;

According to the training set, using a contrastive divergence algorithm to adjust the parameters of the medical resource recommendation model, and iteratively training the medical resource recommendation model to obtain a trained medical resource recommendation model;

The test set is used to test and adjust the trained medical resource recommendation model to obtain a trained medical resource recommendation model.
The electronic device according to claim 12, wherein, according to the training set, the parameters of the medical resource recommendation model are adjusted using a contrastive divergence algorithm, and the medical resource recommendation model is iteratively trained to obtain a trained Medical resource recommendation model, including:

Convert the historical patient information feature set, the historical medical resource feature set, and the associated features in the training set into feature vectors, and perform mean value, variance, and whitening processing on the feature vectors, and process the processed The eigenvectors of are uniformly normalized to between 0 and 1 in order from small to large to obtain a normalized sample set;

Initializing the weight value between the neurons of the visible layer and the neurons of the hidden layer in the medical resource recommendation model, and the bias parameters of the neurons of the visible layer and the bias parameters of the neurons of the hidden layer;

Using the normalized sample set, the weight value, the bias parameter of the neuron in the visible layer and the bias parameter of the neuron in the hidden layer, the neuron in the visible layer and the neuron in the hidden layer Layer neurons perform loop iterations, and calculate the activation probability of the hidden layer neurons;

Using the hidden layer neurons after the loop iteration, iterate the visible layer neurons in a reverse cycle, and calculate the activation probability of the visible layer neurons;

According to the activation probability of the neuron in the hidden layer and the activation probability of the neuron in the visible layer, use a contrastive divergence algorithm to adjust the weight value and the bias parameter, and use the adjusted weight value and The bias parameter is to recalculate the activation probability of the visible layer neurons, and use the output result of the medical resource recommendation model corresponding to the maximum value of the activation probability of the visible layer neurons as the predicted recommended medical resources;

Using a loss function to calculate the loss value between the predicted and recommended medical resources and the historical patient visit data, and adjust the parameters of the medical resource recommendation model according to the loss value, and return to the initialization of the medical resource recommendation model The step of weighting the neurons in the visible layer and the neurons in the hidden layer until the loss value is less than a preset loss threshold, to obtain a trained medical resource recommendation model.
The electronic device according to claim 13, wherein, according to the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer, the contrastive divergence algorithm is used to compare the weight value and the bias Adjust the configuration parameters, including:

Using the normalized sample set to initialize the visible layer neuron, the state vector of the hidden layer neuron, the weight value between the visible layer neuron and the hidden layer neuron;

Utilize the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer to perform K-step Gibbs sampling to obtain the state vector of the neurons in the visible layer and the neurons in the hidden layer corresponding to the t-1 moment The state vector of , the state vector of the visible layer neuron and the state vector of the hidden layer neuron corresponding to time t;

Utilize the state vector of the visible layer neuron corresponding to the t-1 moment and the state vector of the hidden layer neuron, the state vector of the visible layer neuron corresponding to the t moment and the state vector of the hidden layer neuron, to the described The normalized sample set is cyclically calculated to obtain the comparison error value;

The weight value, the bias parameter of the visible layer neuron and the bias parameter of the hidden layer neuron are adjusted by using the comparison error value.
A computer-readable storage medium, storing a computer program, wherein, when the computer program is executed by a processor, the following method for recommending medical resources is implemented:

Obtaining a historical patient data set and a historical medical resource set, extracting a historical patient information feature set and historical patient visit data from the historical patient data set, and extracting a historical medical resource feature set from the historical medical resource set;

Using the historical patient visit data, extract the associated features between the historical patient information feature set and the historical medical resource feature set, use the historical patient information feature set, the historical medical resource feature set and the association features, constructing a medical recommendation sample set for historical patients;

Based on the historical patient medical recommendation sample set, using the contrastive divergence algorithm to train the pre-built medical resource recommendation model to obtain the trained medical resource recommendation model;

Obtain patient information to be analyzed and available medical resources, use the trained medical resource recommendation model to perform matching analysis on the patient information to be analyzed and available medical resources, and obtain recommended medical resource data.
The computer-readable storage medium of claim 15, wherein said extracting a feature set of historical patient information from said historical patient data set comprises:

Converting the historical patient data set into a text format uniformly to obtain the historical patient text data set;

Carrying out word segmentation and part-of-speech tagging on the historical patient text data set to obtain the results of word segmentation and part-of-speech tagging;

Nouns and noun phrases are extracted from the word segmentation according to the results of the part-of-speech tagging, and according to the nouns and noun phrases, statistically obtain historical patient information characteristic frequencies, and generate frequent pattern trees according to the historical patient information characteristic frequencies;

identifying features in the frequent pattern tree to obtain a feature set of candidate historical patient information;

Calculate the point mutual information value of each feature in the candidate historical patient information feature set, and filter out the historical patient information features whose point mutual information value is less than a preset standard threshold from the candidate historical patient information feature set to obtain the historical patient information feature set.
The computer-readable storage medium according to claim 15, wherein said extracting the associated features between the historical patient information feature set and the historical medical resource feature set by using the historical patient visit data comprises:

Segmenting each historical patient information feature in the historical patient information feature set and each historical medical resource feature in the historical medical resource feature set, and extracting key words and relational words from the word segmentation results;

The key words and relational words irrelevant to the historical patient visit data are deleted to obtain the associated features between the historical patient information feature set and the historical medical resource feature set.
The computer-readable storage medium according to claim 15, wherein, based on the historical patient medical recommendation sample set, a contrastive divergence algorithm is used to train a pre-built medical resource recommendation model to obtain a trained medical resource Recommended models, including:

Dividing the historical patient medical recommendation sample set into a training set and a test set;

According to the training set, using a contrastive divergence algorithm to adjust the parameters of the medical resource recommendation model, and iteratively training the medical resource recommendation model to obtain a trained medical resource recommendation model;

The test set is used to test and adjust the trained medical resource recommendation model to obtain a trained medical resource recommendation model.
The computer-readable storage medium according to claim 18, wherein, according to the training set, the parameters of the medical resource recommendation model are adjusted using a contrastive divergence algorithm, and the medical resource recommendation model is iteratively trained to obtain Trained medical resource recommendation models, including:

Convert the historical patient information feature set, the historical medical resource feature set, and the associated features in the training set into feature vectors, and perform mean value, variance, and whitening processing on the feature vectors, and process the processed The eigenvectors of are uniformly normalized to between 0 and 1 in order from small to large to obtain a normalized sample set;

Initializing the weight value between the neurons of the visible layer and the neurons of the hidden layer in the medical resource recommendation model, and the bias parameters of the neurons of the visible layer and the bias parameters of the neurons of the hidden layer;

Using the normalized sample set, the weight value, the bias parameter of the neuron in the visible layer and the bias parameter of the neuron in the hidden layer, the neuron in the visible layer and the neuron in the hidden layer Layer neurons perform loop iterations, and calculate the activation probability of the hidden layer neurons;

Using the hidden layer neurons after the loop iteration, iterate the visible layer neurons in a reverse cycle, and calculate the activation probability of the visible layer neurons;

According to the activation probability of the neuron in the hidden layer and the activation probability of the neuron in the visible layer, use a contrastive divergence algorithm to adjust the weight value and the bias parameter, and use the adjusted weight value and The bias parameter is to recalculate the activation probability of the visible layer neurons, and use the output result of the medical resource recommendation model corresponding to the maximum value of the activation probability of the visible layer neurons as the predicted recommended medical resources;

Using a loss function to calculate the loss value between the predicted and recommended medical resources and the historical patient visit data, and adjust the parameters of the medical resource recommendation model according to the loss value, and return to the initialization of the medical resource recommendation model The step of weighting the neurons in the visible layer and the neurons in the hidden layer until the loss value is less than a preset loss threshold, to obtain a trained medical resource recommendation model.
The computer-readable storage medium according to claim 19, wherein, according to the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer, the contrastive divergence algorithm is used to compare the weight values and The bias parameters to adjust include:

Using the normalized sample set to initialize the visible layer neuron, the state vector of the hidden layer neuron, the weight value between the visible layer neuron and the hidden layer neuron;

Utilize the activation probability of the neurons in the hidden layer and the activation probability of the neurons in the visible layer to perform K-step Gibbs sampling to obtain the state vector of the neurons in the visible layer and the neurons in the hidden layer corresponding to the t-1 moment The state vector of , the state vector of the visible layer neuron and the state vector of the hidden layer neuron corresponding to time t;

Utilize the state vector of the visible layer neuron corresponding to the t-1 moment and the state vector of the hidden layer neuron, the state vector of the visible layer neuron corresponding to the t moment and the state vector of the hidden layer neuron, to the described The normalized sample set is cyclically calculated to obtain the comparison error value;

The weight value, the bias parameter of the visible layer neuron and the bias parameter of the hidden layer neuron are adjusted by using the comparison error value.