Method and device for processing medical assessment data based on machine learning

Abstract

The present application provides a method and device for processing medical assessment data based on machine learning, which relates to the field of artificial intelligence and digital medicine, and its main purpose is to improve the problem of low business processing capability of intelligent medical systems caused by manual review of medical assessment data. It includes: obtaining the medical information and medical credit parameters of the admission request, and calling the execution thread of the admission operation to perform the admission process; predicting and processing the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data of the admission request; if the medical assessment data is an assessment failure result, calling the credit identifier that matches the admission request and locking the credit identifier; if the time length of locking the credit identifier is greater than the preset time threshold, intercepting and suspending the execution thread, and reporting the admission request to update the medical information and the medical credit parameters.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence and digital medical technology, and in particular to a method and device for processing medical assessment data based on machine learning. Background Art

[0002] In order to greatly improve and speed up the speed and efficiency of medical treatment, various hospitals are gradually adopting intelligent medical systems to realize the intelligent data processing between medical and patients. Especially for patients who apply for hospitalization claims, verification is required to determine whether they meet the standards for hospitalization claims, so as to avoid defrauding claims through hospitalization claims.

[0003] At present, the existing underwriting of hospitalization claims is only through manual data review. However, patients who need hospitalization may have acute or serious illnesses, and manual review is slow, which affects the effectiveness of the actual hospitalization claims. Hospitalization before review will affect the authenticity of the hospitalization claims, resulting in fraudulent withdrawal of hospitalization claim expenses, affecting the business processing capabilities of the intelligent medical system. Therefore, a method for processing medical assessment data is urgently needed to solve the above problems. Summary of the invention

[0004] In view of this, the present application provides a method and device for processing medical assessment data based on machine learning, the main purpose of which is to improve the existing technical problem of low business processing capabilities of intelligent medical systems caused by manual review of medical assessment data.

[0005] According to one aspect of the present application, a method for processing medical assessment data based on machine learning is provided, comprising:

[0006] Obtain the medical information and medical credit parameters of the admission request, and call the execution thread of the admission operation to perform the admission process. The medical credit parameters are evaluation parameters for the historical medical payment status;

[0007] The medical information and medical credit parameters are predicted and processed based on the trained credit assessment processing model to obtain the medical assessment data of the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted and trained based on the additional information of the medical credit parameters;

[0008] If the medical assessment data is an assessment failure result, the credit identifier matching the admission request is retrieved and the credit identifier is locked, wherein the credit identifier is used to represent the functional information of marking whether to execute in the process of executing different medical consultation operations;

[0009] If the length of time for locking the credit identifier is greater than a preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported so that the medical information and the medical credit parameter are updated.

[0010] Preferably, before obtaining the medical information and medical credit parameters of the admission request, the method further includes:

[0011] Acquire a medical assessment data training sample set, wherein the medical assessment data training sample set includes structured and unstructured medical sample data and medical credit sample data;

[0012] A credit assessment processing model consisting of two tower networks is constructed, and the credit assessment processing model is trained based on the medical assessment data training sample set shown to obtain a trained credit assessment processing model, wherein the weight between the two tower networks is configured based on the weight value calculated based on the additional information.

[0013] Preferably, the performing model training on the credit assessment processing model based on the medical assessment data training sample set includes:

[0014] Determining whether the medical credit parameter is greater than a preset credit threshold;

[0015] If it is greater, obtaining a first additional function, and calculating a weight value matching the medical credit parameter based on the first additional function, so as to configure the weight between the two tower networks based on the weight value;

[0016] If it is less than, a second additional function is obtained, and a weight value matching the medical credit parameter is calculated based on the second additional function, so as to configure the weight between the two tower networks based on the weight value.

[0017] Preferably, before obtaining the medical information and medical credit parameters of the admission request, the method further includes:

[0018] Acquire historical medical information matching the admission request, and parse historical claim information and historical payment data based on the historical medical information;

[0019] Based on the difference between the historical claim information and the historical payment information, a medical credit parameter matching the medical information is searched from a preset payment evaluation relationship.

[0020] Preferably, before retrieving the credit identifier matching the admission request and locking the credit identifier, the method further includes:

[0021] At least one diagnosis and treatment operation matching the medical information is determined, and an initialized credit identifier is configured for the diagnosis and treatment operation to determine whether to lock the credit identifier according to the medical evaluation data.

[0022] Preferably, the method further comprises:

[0023] If the medical assessment data is a successful assessment result, the credit identification corresponding to the diagnosis and treatment operation is cleared, and claim processing information for the admission request is generated to indicate the approval operation of the admission request.

[0024] Preferably, the method further comprises:

[0025] If an update operation on the medical information is detected, re-acquiring the medical credit parameter;

[0026] The step of re-executing the step of predicting the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data of the admission request.

[0027] According to another aspect of the present application, a device for processing medical assessment data based on machine learning is provided, comprising:

[0028] The retrieval module is used to obtain the medical information and medical credit parameters of the admission request, and to retrieve the execution thread of the admission operation to perform the admission process. The medical credit parameters are evaluation parameters for the historical medical payment status;

[0029] A prediction module, used to perform prediction processing on the medical information and medical credit parameters based on a trained credit assessment processing model to obtain medical assessment data for the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted based on additional information of the medical credit parameters to complete model training;

[0030] A locking module, used for retrieving a credit identifier that matches the hospital admission request and locking the credit identifier if the medical assessment data is an assessment failure result, wherein the credit identifier is used to represent functional information of marking whether to execute in the process of executing different medical consultation operations;

[0031] The reporting module is used to intercept and suspend the execution thread if the length of time for locking the credit identifier is greater than a preset time threshold, and report the admission request to update the medical information and the medical credit parameter.

[0032] Preferably, before the calling module, the device further includes:

[0033] An acquisition module, used to acquire a medical evaluation data training sample set, wherein the medical evaluation data training sample set includes structured and unstructured medical sample data and medical credit sample data;

[0034] A construction module is used to construct a credit assessment processing model consisting of two tower networks, and to perform model training on the credit assessment processing model based on the medical assessment data training sample set to obtain a trained credit assessment processing model, wherein the weight between the two tower networks is configured based on the weight value calculated based on the additional information.

[0035] Preferably, the building blocks include:

[0036] A judging unit, used for judging whether the medical credit parameter is greater than a preset credit threshold;

[0037] a configuration unit, configured to obtain a first additional function if it is greater than , and calculate a weight value matching the medical credit parameter based on the first additional function, so as to configure the weight between the two tower networks based on the weight value;

[0038] The configuration unit is further configured to obtain a second additional function if it is less than , and calculate a weight value matching the medical credit parameter based on the second additional function, so as to configure the weight between the two tower networks based on the weight value.

[0039] Preferably, before the calling module, the device further includes:

[0040] A parsing module, used for obtaining historical medical information matching the admission request, and parsing historical claim information and historical payment data based on the historical medical information;

[0041] A search module is used to search for medical credit parameters matching the medical information from a preset payment evaluation relationship based on the difference between the historical claim information and the historical payment information.

[0042] Preferably, before the locking module, the device further comprises:

[0043] A configuration module is used to determine at least one diagnosis and treatment operation that matches the medical information, and configure an initialized credit identifier for the diagnosis and treatment operation to determine whether to lock the credit identifier based on the medical evaluation data.

[0044] Preferably, the device further comprises:

[0045] A clearing module is used to clear the credit identification corresponding to the diagnosis and treatment operation if the medical assessment data is a successful assessment result, and generate claim processing information for the admission request to indicate the approval operation of the admission request.

[0046] Preferably, the device further comprises:

[0047] A detection module, configured to reacquire the medical credit parameter if an update operation on the medical information is detected;

[0048] The updating module is used to re-execute the step of predicting and processing the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data of the admission request.

[0049] According to another aspect of the present application, a storage medium is provided, wherein at least one executable instruction is stored in the storage medium, and the executable instruction enables a processor to perform operations corresponding to the above-mentioned method for processing medical assessment data based on machine learning.

[0050] According to another aspect of the present application, there is provided a computer device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0051] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned method for processing medical assessment data based on machine learning.

[0052] By means of the above technical solution, the technical solution provided by the embodiment of the present application has at least the following advantages:

[0053] The present application provides a method and device for processing medical evaluation data based on machine learning. Compared with the prior art, the embodiment of the present application obtains the medical information and medical credit parameters of the admission request, and calls the execution thread of the admission operation to perform the admission processing, the medical credit parameter is an evaluation parameter for the historical medical payment status; the medical information and medical credit parameters are predicted and processed based on the trained credit evaluation processing model to obtain the medical evaluation data of the admission request, the credit evaluation processing model is constructed based on two tower networks and the tower networks are weighted and configured based on the additional information of the medical credit parameters to complete the model training; if the medical evaluation data is an evaluation failure result, the credit identifier matching the admission request is called and the credit identifier is locked, the credit identifier is used to characterize the functional information of marking whether to execute in the process of executing different medical treatment operations; if the time length of locking the credit identifier is greater than the preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported to update the medical information and the medical credit parameters. It avoids the occurrence of fraudulent hospitalization claims and improves the processing efficiency of real hospitalization claims and the business processing capabilities of the intelligent medical system.

[0054] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0056] Figure 1 A flow chart of a method for processing medical assessment data based on machine learning provided in an embodiment of the present application is shown;

[0057] Figure 2 A schematic diagram of the network structure of a credit assessment processing model provided in an embodiment of the present application is shown;

[0058] Figure 3 A block diagram of a medical assessment data processing device based on machine learning provided in an embodiment of the present application is shown;

[0059] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0060] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0061] The embodiments of the present application can 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.

[0062] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0063] Based on this, in one embodiment, Figure 1 As shown, a method for processing medical assessment data based on machine learning is provided, and the method is applied to a computer device such as a server as an example for explanation, wherein the server can be an independent server, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms, such as intelligent medical systems, digital medical platforms, etc. The above method includes the following steps:

[0064] 101. Obtain the medical information and medical credit parameters of the admission request, and call the execution thread of the admission operation to perform the admission process.

[0065] Among them, the medical credit parameter is an evaluation parameter for the historical medical payment status.

[0066] In the embodiment of the present application, users can establish an account by registering for the smart medical system. If the user has a need for hospitalization, he can first apply for admission based on the smart medical system, and at the same time verify the admission information based on the computer background to speed up the efficiency of admission and avoid the situation where admission cannot be processed due to emergency situations. That is, the admission operation can be executed through the execution thread of the admission operation, so that the application and admission process are processed at the same time. Among them, the medical credit parameter is an evaluation parameter for the user's historical medical payment situation. The medical information reported for admission includes but is not limited to user identity information, user medical information, doctor diagnosis information, etc., combined with the medical credit parameter as a processing object for predicting whether admission can be performed.

[0067] Among them, the admission process includes the process of generating an admission account based on medical information, which may include generating ward numbers, corresponding resident physician information, admission departments, etc., so that users who really need to be admitted to the hospital can enter the admission process, and at the same time, anti-fraud assessment and processing can be performed on this user.

[0068] 102. Based on the trained credit assessment processing model, predict and process the medical information and medical credit parameters to obtain the medical assessment data of the admission request.

[0069] The credit assessment processing model is constructed based on two tower networks and the tower networks are weighted based on the additional information of medical credit parameters to complete the model training. The medical credit parameters are evaluation parameters for the user's historical medical payment situation, which have been mentioned above and will not be repeated here.

[0070] In the embodiment of the present application, the credit assessment processing model is constructed as two tower-shaped networks, one of which serves as a processing network structure for structured data and unstructured data, and the other serves as a processing network structure for the patient's medical information and medical credit parameters, thereby predicting hospital admission declarations.

[0071] 103. If the medical assessment data shows an assessment failure result, the credit identifier that matches the admission request is retrieved and the credit identifier is locked.

[0072] Among them, the credit mark is used to represent the functional information of whether the mark is executed during the execution of different medical treatment operations.

[0073] In the embodiment of the present application, the credit mark is used to characterize the functional information of whether different medical consultation operations are executed in the intelligent medical system, that is, it indicates that the admission operation currently executed in the intelligent medical system has a risk of fraud, so the credit mark is retrieved for locking. Specifically, the medical credit database can be created when applying for admission, and is used to store all the information generated by the user after admission, including basic information and medical treatment information prescribed by the doctor. At the same time, a credit mark is marked to determine whether the current user is being evaluated. For example, if the patient's B-ultrasound medical project carries a credit mark, it means that the user is being evaluated. When the medical data evaluation fails, the current execution subject locks the credit mark, so that the doctor cannot change the credit mark, indicating that the B-ultrasound medical project cannot be executed at this time. It should be noted that when the doctor performs a medical operation on a user marked with a credit tag, he needs to change the sign to indicate that the operation is being performed on this user.

[0074] 104. If the length of time for locking the credit identification is greater than a preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported to update the medical information and medical credit parameters.

[0075] In the embodiment of the present application, if the credit identification is locked, it means that the user has failed the assessment. In order to allow the user to continue to perform the steps of the admission process, a certain time interval can be set for the user, and an alarm can be issued to the user during this time, so that the user can pay the fee or fill in the information to unfreeze the credit identification. At this time, the admission process will not be intercepted. However, if the length of time the credit identification is locked is greater than the time threshold, it means that the user is lazy in dealing with the admission assessment failure factor. At this time, it can be determined that the user is likely to be a fraudster, and then the thread executed in step 101 is intercepted and suspended, and the request for admission is reported to the supervisor, so that the user can update the information in the admission process, or instruct the supervisor to intervene manually. Effectively achieve the purpose of preventing fraudulent hospitalization.

[0076] For further explanation and limitation, in an embodiment of the present application, before obtaining the medical information and medical credit parameters of the admission request, the method of this embodiment also includes: obtaining a medical assessment data training sample set; constructing a credit assessment processing model composed of two tower networks, and performing model training on the credit assessment processing model based on the medical assessment data training sample set to obtain a trained credit assessment processing model, and the weight between the two tower networks is configured based on the weight value calculated based on the additional information.

[0077] Among them, the medical assessment data training sample set includes structured and unstructured medical sample data and medical credit sample data.

[0078] Preferably, the network structure of the credit assessment processing model is as follows: Figure 2 As shown in FIG. 1 , it is composed of two tower networks on the left and right. For example, it can be composed of network1 and network2. The structures are TextCNN+FC, where FC (fully connected layers, FC) is a fully connected layer. Each neuron in the layer is fully connected to all neurons in the previous layer, which can be realized by a 1*1 convolution operation. The tower structure on the left inputs the structured information and unstructured information of the user's basic information, and the tower on the right inputs the user's medical information and medical credit identification. The processing results of the left tower and the right tower are calculated by combining the weight value as the prediction result of the hospital admission declaration. Specifically, in order to complete the model training of the tower network, the similarity of the training errors of the two towers is calculated during the training process. Specifically, the network can extract the features of the input information, calculate the errors of the extracted features, and thus calculate the similarity. If the similarity value of the left and right towers is greater than the threshold, the training is completed. If it is less than, the training is iterated again. Correspondingly, the prediction result is that the last layer of the network is a softmax layer, and the predicted probability value of fraud can be directly obtained. According to the comparison between the probability value and the manually set threshold value, the value greater than the threshold is 1, and the value less than the threshold is 0, and the prediction result of the hospital admission declaration is finally determined.

[0079] For example, corpus is collected from the user's medical information and hospitalization information, including all structured and unstructured information input from the left tower, which may include user basic information and diagnosis information, and the information input from the right tower includes medical information such as doctor's prescription information and hospitalization information. As a pair, the fraud label (0, 1) corresponding to the training data is used as the prediction result, where 1 represents fraudulent hospitalization and 0 represents normal hospitalization. The data set is randomly divided into training set, validation set and test set in a ratio of 8:1:1.

[0080] It should be noted that feature processing is performed before training the model. Discrete features such as gender, education, symptoms, and drugs are directly represented by one-hot. Continuous features need to be discretized. For example, for age, the continuous value is discretized into buckets using equal frequency. For unstructured and semi-structured text information such as diagnosis information and hospitalization information, the word embedding is directly input and feature extraction is performed using TextCNN. The word embedding information can come from the offline trained word2vec model.

[0081] In this process, structured text information is input into TextCNN to extract features and then concat with the user's unstructured basic information and pass through the FC layer to obtain a fixed dimension of 128 on the left tower. In the same way, medical information and medical credit parameters are input into the TextCNN model of the right tower, concat and pass through the FC layer to output a 128-dimensional vector for model processing.

[0082] In order to further limit the weight distribution of the two tower networks during the model training process, in an embodiment of the present application, model training of the credit assessment processing model based on the medical assessment data training sample set includes: determining whether the medical credit parameter is greater than a preset credit threshold; if greater than, obtaining a first additional function, and calculating a weight value matching the medical credit parameter based on the first additional function, so as to configure the weight between the two tower networks based on the weight value; if less than, obtaining a second additional function, and calculating a weight value matching the medical credit parameter based on the second additional function, so as to configure the weight between the two tower networks based on the weight value.

[0083] Specifically, when weights of two tower-shaped networks are configured, the configuration can be performed based on the additional information of the medical credit parameter, wherein the additional information is a calculation function for calculating weight values ​​of different medical credit parameters. For example, if the medical credit parameter is greater than 1, the additional information is Formula 1, i.e., f1 = (medical credit parameter - 1) / b, wherein b is a weight ratio parameter, and the determined f1 and 1-f1 correspond to the weights of the two tower-shaped networks respectively; if the medical credit parameter is less than 1, the additional information is Formula 2, i.e., f2 = medical credit parameter / b, and the determined f2 and 1-f2 correspond to the weights of the two tower-shaped networks respectively.

[0084] Since medical credit parameters are important parameters in credit assessment processing and are updated at any time according to historical medical treatment situations, in an embodiment of the present application, before obtaining the medical information and medical credit parameters of the admission request, the method of this embodiment also includes: obtaining historical medical information matching the admission request, and parsing historical claims information and historical payment data based on the historical medical information; based on the difference between the historical claims information and the historical payment information, searching for medical credit parameters matching the medical information from a preset payment evaluation relationship.

[0085] Specifically, in the smart medical system as the current execution end, there are records of the medical information of a large number of users, as well as the medical treatment information and payment status of the users. Therefore, when obtaining the medical credit parameter, the payment amount paid for the historical medical information is first obtained to illustrate the user's payment ability, and the historical claim information is obtained at the same time to illustrate that the user has the claim ability, and the safety of hospitalization can be expected based on the claim situation. Further, a payment evaluation relationship can be pre-stored in the current execution end, that is, based on the payment situation as the result of the credit evaluation, and the evaluation parameter value corresponding to the difference range between different payment amounts and historical claim information can be included as a medical credit parameter. It can be understood that the difference between the payment amount and the historical claim information can be positive or negative. When it is a positive value, it means that the payment amount is greater than the claim amount, so the corresponding evaluation parameter is small. When it is a negative value, it means that the payment amount is less than the claim amount, so the corresponding evaluation parameter is large.

[0086] Furthermore, in an embodiment of the present application, before retrieving the credit identifier that matches the admission request and locking the credit identifier, the method of this embodiment also includes: determining at least one diagnosis and treatment operation that matches the medical information, and configuring an initialized credit identifier for the diagnosis and treatment operation to determine whether to lock the credit identifier based on the medical assessment data.

[0087] Among them, the diagnosis and treatment operation includes but is not limited to the treatment operation that the doctor has already performed on the user at any stage. When generating the reservation information corresponding to the diagnosis and treatment operation, an initialized credit mark is configured for the diagnosis and treatment operation, that is, a mark that can be changed, so that the doctor can change the signature before performing the operation. When the current execution end determines that the medical data evaluation result fails, the credit mark is locked, and the doctor can be informed of the risk.

[0088] In the embodiment of the present application, another situation corresponding to step 103 also includes: if the medical assessment data is a successful assessment result, the credit identifier corresponding to the diagnosis and treatment operation is cleared, and claim processing information for the admission request is generated to indicate the approval of the admission request.

[0089] Specifically, when the medical assessment data shows a successful assessment result, it means that the risk of fraud of the user currently being assessed is low and the user is a normal hospitalized user. In this way, the credit identification corresponding to the medical treatment operation in the user's account will be cleared, and the claim processing information of the admission request will be generated to indicate that the admission request is approved.

[0090] After the user is admitted to the hospital, the doctor may make a secondary diagnosis based on the user's physical condition, and will regenerate medical information and medical credit parameters accordingly. At this time, it is necessary to re-perform credit assessment processing based on the updated content. Accordingly, in the embodiment of the present application, the method also includes: if an update operation on the medical information is detected, re-acquiring the medical credit parameters; re-executing the step of predicting and processing the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data for the admission request.

[0091] The present application provides a method for processing medical evaluation data based on machine learning. Compared with the prior art, the embodiment of the present application obtains the medical information and medical credit parameters of the admission request, and calls the execution thread of the admission operation to perform the admission processing, and the medical credit parameter is an evaluation parameter for the historical medical payment status; the medical information and medical credit parameter are predicted and processed based on the trained credit evaluation processing model to obtain the medical evaluation data of the admission request, and the credit evaluation processing model is constructed based on two tower networks and the tower networks are weighted and configured based on the additional information of the medical credit parameter to complete the model training; if the medical evaluation data is an evaluation failure result, the credit identifier matching the admission request is called, and the credit identifier is locked, and the credit identifier is used to characterize the functional information of marking whether to execute in the process of executing different medical treatment operations; if the time length of locking the credit identifier is greater than the preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported to update the medical information and the medical credit parameter. It avoids the occurrence of fraudulent hospitalization claims and improves the processing efficiency of real hospitalization claims and the business processing capabilities of the intelligent medical system.

[0092] Furthermore, as a response to the above Figure 1 The present application provides a device for processing medical evaluation data based on machine learning, such as Figure 3 As shown, the device includes: a calling module 21, a prediction module 22, a locking module 23, and a reporting module 24.

[0093] The retrieval module 21 is used to obtain the medical information and medical credit parameters of the admission request, and to retrieve the execution thread of the admission operation to perform the admission process. The medical credit parameter is an evaluation parameter for the payment status of the historical medical treatment;

[0094] A prediction module 22 is used to perform prediction processing on the medical information and medical credit parameters based on a trained credit assessment processing model to obtain medical assessment data of the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted based on additional information of the medical credit parameters to complete model training;

[0095] A locking module 23, for retrieving a credit identifier that matches the hospital admission request and locking the credit identifier if the medical assessment data is an assessment failure result, wherein the credit identifier is used to indicate functional information of marking whether to execute different medical consultation operations;

[0096] The reporting module 24 is used to intercept and suspend the execution thread if the length of time for locking the credit identification is greater than a preset time threshold, and report the admission request to update the medical information and the medical credit parameter.

[0097] In a specific application scenario, before the calling module 21, the device further includes:

[0098] An acquisition module, used to acquire a medical evaluation data training sample set, wherein the medical evaluation data training sample set includes structured and unstructured medical sample data and medical credit sample data;

[0099] A construction module is used to construct a credit assessment processing model consisting of two tower networks, and to perform model training on the credit assessment processing model based on the medical assessment data training sample set to obtain a trained credit assessment processing model, wherein the weight between the two tower networks is configured based on the weight value calculated based on the additional information.

[0100] In a specific application scenario, the building blocks include:

[0101] A judging unit, used for judging whether the medical credit parameter is greater than a preset credit threshold;

[0102] a configuration unit, configured to obtain a first additional function if it is greater than , and calculate a weight value matching the medical credit parameter based on the first additional function, so as to configure the weight between the two tower networks based on the weight value;

[0103] The configuration unit is further configured to obtain a second additional function if it is less than , and calculate a weight value matching the medical credit parameter based on the second additional function, so as to configure the weight between the two tower networks based on the weight value.

[0104] In a specific application scenario, before the calling module 21, the device further includes:

[0105] A parsing module, used for obtaining historical medical information matching the admission request, and parsing historical claim information and historical payment data based on the historical medical information;

[0106] A search module is used to search for medical credit parameters matching the medical information from a preset payment evaluation relationship based on the difference between the historical claim information and the historical payment information.

[0107] In a specific application scenario, before the locking module 23, the device further includes:

[0108] A configuration module is used to determine at least one diagnosis and treatment operation that matches the medical information, and configure an initialized credit identifier for the diagnosis and treatment operation to determine whether to lock the credit identifier based on the medical evaluation data.

[0109] In a specific application scenario, the device further includes:

[0110] A clearing module is used to clear the credit identification corresponding to the diagnosis and treatment operation if the medical assessment data is a successful assessment result, and generate claim processing information for the admission request to indicate the approval operation of the admission request.

[0111] In a specific application scenario, the device further includes:

[0112] A detection module, configured to reacquire the medical credit parameter if an update operation on the medical information is detected;

[0113] The updating module is used to re-execute the step of predicting and processing the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data of the admission request.

[0114] The present application provides a processing device for medical evaluation data based on machine learning. Compared with the prior art, the embodiment of the present application obtains the medical information and medical credit parameters of the admission request, and calls the execution thread of the admission operation to perform the admission processing, the medical credit parameter is an evaluation parameter for the historical medical payment status; the medical information and medical credit parameters are predicted and processed based on the trained credit evaluation processing model to obtain the medical evaluation data of the admission request, the credit evaluation processing model is constructed based on two tower networks and the tower networks are weighted and configured based on the additional information of the medical credit parameters to complete the model training; if the medical evaluation data is an evaluation failure result, the credit identifier matching the admission request is called and the credit identifier is locked, the credit identifier is used to characterize the functional information of marking whether to execute in the process of executing different medical treatment operations; if the time length of locking the credit identifier is greater than the preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported to update the medical information and the medical credit parameters. It avoids the occurrence of fraudulent hospitalization claims and improves the processing efficiency of real hospitalization claims and the business processing capabilities of the intelligent medical system.

[0115] According to one embodiment of the present application, a storage medium is provided, wherein the storage medium stores at least one executable instruction, and the computer executable instruction can execute the method for processing medical assessment data based on machine learning in any of the above method embodiments.

[0116] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0117] Figure 4 A schematic diagram of the structure of a computer device provided according to an embodiment of the present application is shown. The specific embodiment of the present application does not limit the specific implementation of the computer device.

[0118] like Figure 4 As shown, the computer device may include: a processor (processor) 302 , a communication interface (Communications Interface) 304 , a memory (memory) 306 , and a communication bus 308 .

[0119] The processor 302 , the communication interface 304 , and the memory 306 communicate with each other via the communication bus 308 .

[0120] The communication interface 304 is used to communicate with other devices such as clients or other servers.

[0121] The processor 302 is used to execute the program 310, and specifically can execute the relevant steps in the above-mentioned embodiment of the method for processing medical assessment data based on machine learning.

[0122] Specifically, the program 310 may include program codes, which include computer operation instructions.

[0123] The processor 302 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in the computer device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0124] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0125] The program 310 may be specifically configured to enable the processor 302 to perform the following operations:

[0126] Obtain the medical information and medical credit parameters of the admission request, and call the execution thread of the admission operation to perform the admission process. The medical credit parameters are evaluation parameters for the historical medical payment status;

[0127] The medical information and medical credit parameters are predicted and processed based on the trained credit assessment processing model to obtain the medical assessment data of the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted and trained based on the additional information of the medical credit parameters;

[0128] If the medical assessment data is an assessment failure result, the credit identifier matching the admission request is retrieved and the credit identifier is locked, wherein the credit identifier is used to represent the functional information of marking whether to execute in the process of executing different medical consultation operations;

[0129] If the length of time for locking the credit identifier is greater than a preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported so that the medical information and the medical credit parameter are updated.

[0130] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the entity device for business data processing based on the multimodal hybrid model, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to realize communication between components within the storage medium, as well as communication with other hardware and software in the information processing entity device.

[0131] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0132] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for processing medical assessment data based on machine learning, characterized in that: include: Obtain the medical information and medical credit parameters of the admission request, and call the execution thread of the admission operation to perform the admission process. The medical credit parameters are evaluation parameters for the historical medical payment status; The medical information and medical credit parameters are predicted and processed based on the trained credit assessment processing model to obtain the medical assessment data of the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted and trained based on the additional information of the medical credit parameters, wherein the additional information is a calculation function for calculating weight values ​​of different medical credit parameters; If the medical assessment data is an assessment failure result, the credit identifier matching the admission request is retrieved and the credit identifier is locked, wherein the credit identifier is used to represent the functional information of marking whether to execute in the process of executing different medical consultation operations; If the length of time for locking the credit identifier is greater than a preset time threshold, the execution thread is intercepted and suspended, and the admission request is reported to update the medical information and the medical credit parameter; Before obtaining the medical information and medical credit parameters of the admission request, the method further includes: Acquire a medical assessment data training sample set, wherein the medical assessment data training sample set includes structured and unstructured medical sample data and medical credit sample data; Constructing a credit assessment processing model consisting of two tower-shaped networks, and performing model training on the credit assessment processing model based on the medical assessment data training sample set to obtain a trained credit assessment processing model, wherein the weight between the two tower-shaped networks is configured based on the weight value calculated based on the additional information; Model training of the credit assessment processing model based on the medical assessment data training sample set includes: Determining whether the medical credit parameter is greater than a preset credit threshold; If it is greater than, a first additional function is obtained, and a weight value matching the medical credit parameter is calculated based on the first additional function, so as to configure the weight between the two tower networks based on the weight value, wherein the first additional function is expressed as f1=(medical credit parameter-1) / b, where b is a weight ratio parameter; If it is less than, a second additional function is obtained, and a weight value matching the medical credit parameter is calculated based on the second additional function to configure the weight between the two tower networks based on the weight value, wherein the second additional function is expressed as f2=medical credit parameter / b.

2. The method according to claim 1, characterized in that Before obtaining the medical information and medical credit parameters of the admission request, the method further includes: Acquire historical medical information matching the admission request, and parse historical claim information and historical payment data based on the historical medical information; Based on the difference between the historical claim information and the historical payment data, a medical credit parameter matching the medical information is searched from a preset payment evaluation relationship.

3. The method according to any one of claims 1 to 2, characterized in that: Before retrieving the credit identifier matching the admission request and locking the credit identifier, the method further includes: At least one diagnosis and treatment operation matching the medical information is determined, and an initialized credit identifier is configured for the diagnosis and treatment operation to determine whether to lock the credit identifier according to the medical evaluation data.

4. The method according to claim 3, characterized in that: The method further comprises: If the medical assessment data is a successful assessment result, the credit identification corresponding to the diagnosis and treatment operation is cleared, and claim processing information for the admission request is generated to indicate the approval operation of the admission request.

5. The method according to claim 1, characterized in that The method further comprises: If an update operation on the medical information is detected, re-acquiring the medical credit parameter; The step of re-executing the step of predicting the medical information and medical credit parameters based on the trained credit assessment processing model to obtain the medical assessment data of the admission request.

6. A device for processing medical evaluation data based on machine learning, characterized in that: include: The retrieval module is used to obtain the medical information and medical credit parameters of the admission request, and to retrieve the execution thread of the admission operation to perform the admission process. The medical credit parameters are evaluation parameters for the historical medical payment status; A prediction module, used to perform prediction processing on the medical information and medical credit parameters based on a trained credit assessment processing model to obtain medical assessment data for the admission request, wherein the credit assessment processing model is constructed based on two tower networks and the tower networks are weighted based on additional information of the medical credit parameters to complete model training, wherein the additional information is a calculation function for calculating weight values ​​of different medical credit parameters; A locking module, used for retrieving a credit identifier that matches the hospital admission request and locking the credit identifier if the medical assessment data is an assessment failure result, wherein the credit identifier is used to represent functional information of marking whether to execute in the process of executing different medical consultation operations; A reporting module, configured to intercept and suspend the execution thread if the length of time for locking the credit identifier is greater than a preset time threshold, and report the admission request to update the medical information and the medical credit parameter; Wherein, before the calling module, the device further includes: An acquisition module, used to acquire a medical evaluation data training sample set, wherein the medical evaluation data training sample set includes structured and unstructured medical sample data and medical credit sample data; A construction module is used to construct a credit assessment processing model consisting of two tower networks, and to perform model training on the credit assessment processing model based on the medical assessment data training sample set to obtain a trained credit assessment processing model, wherein the weight between the two tower networks is configured based on the weight value calculated based on the additional information; The building blocks include: A judging unit, used for judging whether the medical credit parameter is greater than a preset credit threshold; a configuration unit, configured to obtain a first additional function if it is greater than, and calculate a weight value matching the medical credit parameter based on the first additional function, so as to configure the weight between the two tower networks based on the weight value, wherein the first additional function is expressed as f1=(medical credit parameter-1) / b, and b is a weight ratio parameter; The configuration unit is also used to obtain a second additional function if it is less than, and calculate a weight value matching the medical credit parameter based on the second additional function to configure the weight between the two tower networks based on the weight value, and the second additional function is expressed as f2=medical credit parameter / b.

7. A storage medium storing at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the method for processing medical assessment data based on machine learning as described in any one of claims 1-5.

8. A computer device comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the method for processing medical assessment data based on machine learning as described in any one of claims 1-5.

Images