Medical data processing network determination method and device and medical data processing method

Abstract

The invention relates to a medical data processing network determination method and apparatus, and a medical data processing method.The medical data processing network determination method comprises the steps of obtaining first medical data in response to a training instruction triggered and generated by a user on an application interface, and performing unsupervised training on a data processing network according to the first medical data, obtaining a target data processing network; inputting the second medical data into the target data processing network for testing to obtain a test result; and evaluating the performance of the target data processing network according to the test result to obtain an evaluation result, and storing network parameters of the target data processing network under the condition that the evaluation result shows that the performance of the target data processing network meets requirements. The generalization ability of the target data processing network obtained by using the determination method of the medical data processing network provided by the invention is relatively high.

Description

Technical Field

[0001] The present application relates to the field of medical technology, and in particular to a method and device for determining a medical data processing network and a medical data processing method. Background Art

[0002] With the rapid development of neural network technology, the use of neural networks in the medical field is becoming more and more widespread. Hospitals usually directly use various trained neural networks provided by suppliers to process medical data.

[0003] However, in order to protect user privacy, etc., hospitals have strict data control and suppliers are unable to obtain medical data in a timely manner, resulting in poor generalization capabilities of the neural networks provided by suppliers. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device and medical data processing method for determining a medical data processing network that can improve the generalization capability of the data processing network in order to address the above technical problems.

[0005] In a first aspect, the present application provides a method for determining a medical data processing network, the method comprising:

[0006] In response to a training instruction generated by a user triggering on an application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network;

[0007] Inputting the second medical data into the target data processing network for testing to obtain a test result;

[0008] The performance of the target data processing network is evaluated according to the test result to obtain an evaluation result, and when the evaluation result indicates that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are stored.

[0009] In one embodiment, the method further comprises:

[0010] Display test results on the application interface;

[0011] The performance of the target data processing network is evaluated based on the test results to obtain evaluation results, including:

[0012] If a confirmation instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network meets the requirements; if a retraining instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

[0013] In one embodiment, the method further comprises:

[0014] When the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned to be executed.

[0015] In one embodiment, obtaining first medical data includes:

[0016] Obtaining initial medical data;

[0017] Determining whether the initial medical data meets the training conditions; the training conditions include whether the amount of the initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold;

[0018] If the initial medical data meets the training condition, determining the initial medical data as the first medical data;

[0019] If the initial medical data does not meet the training condition, new initial medical data is acquired, and based on the new initial medical data, the step of determining whether the initial medical data meets the training condition is returned to be executed.

[0020] In one embodiment, storing the network parameters of the target data processing network includes:

[0021] According to the type of the target data processing network, network parameters of the target data processing network are marked to obtain marked network parameters;

[0022] The marked network parameters are stored.

[0023] In a second aspect, an embodiment of the present application provides a medical data processing method, the method comprising:

[0024] Obtaining medical data to be processed;

[0025] In response to a selection instruction triggered by a user on an application interface, a target network parameter indicated by the selection instruction is obtained, and a basic neural network model is obtained, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each network parameter is obtained according to the method provided in the first aspect above;

[0026] The medical data to be processed is input into the target neural network model for data processing to obtain the processing results.

[0027] In a third aspect, an embodiment of the present application provides a device for determining a medical data processing network, the device comprising:

[0028] an acquisition module, configured to acquire first medical data in response to a training instruction generated by a user triggering the training instruction on the application interface, and to perform unsupervised training on the data processing network according to the first medical data to obtain a target data processing network;

[0029] A testing module, used for inputting the second medical data into the target data processing network for testing to obtain a test result;

[0030] The evaluation module is used to evaluate the performance of the target data processing network according to the test results, obtain the evaluation results, and store the network parameters of the target data processing network when the evaluation results indicate that the performance of the target data processing network meets the requirements.

[0031] In a fourth aspect, an embodiment of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the methods provided in the first and second aspects are implemented.

[0032] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the methods provided in the first and second aspects above are implemented.

[0033] In a sixth aspect, an embodiment of the present application further provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the methods provided in the first and second aspects above.

[0034] In the above-mentioned method, device and method for determining the medical data processing network, the user can trigger the operation on the application interface to generate a training instruction, so as to obtain the first medical data to perform unsupervised training on the data processing network and obtain the target data processing network. In other words, the target data processing network can be trained at the user end (the hospital terminal), so that the generalization ability of the target data processing network is stronger. In addition, in the present embodiment, the target data processing network can also be tested based on the second medical data, and the performance of the target data processing network can be evaluated through the test results. In this way, the network parameters of the obtained target data processing network can be guaranteed to be more accurate, so that the method for determining the medical data processing network has higher reliability and practicality. In addition, in the present embodiment, unsupervised training is performed on the data processing network, so there is no need to obtain a large number of training samples (first medical data), no need to annotate the first medical data, and the hardware requirements are low, which can save training costs to a great extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A diagram of an application environment of a method for determining a medical data processing network in one embodiment;

[0036] Figure 2 A schematic diagram of a flow chart of steps of a method for determining a medical data processing network in one embodiment;

[0037] Figure 3 A schematic flow chart of steps of a method for determining a medical data processing network in another embodiment;

[0038] Figure 4 A schematic flow chart of steps of a method for determining a medical data processing network in another embodiment;

[0039] Figure 5 A schematic flow chart of steps of a method for determining a medical data processing network in another embodiment;

[0040] Figure 6 A schematic diagram of a process flow of a medical data processing method in one embodiment;

[0041] Figure 7 is a schematic diagram of the structure of a medical data processing system in one embodiment;

[0042] Figure 8 is a schematic structural diagram of a determination device of a medical data processing network in one embodiment;

[0043] Fig. 9 is a schematic diagram of the structure of a medical data processing device in one embodiment;

[0044] Fig.10 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0046] The serial numbers assigned to the components in this application, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning.

[0047] First of all, before specifically introducing the technical solutions of the embodiments disclosed in the present application, the background technology or technical evolution context on which the embodiments of the present application are based is introduced. With the rapid development of neural network technology, the use of neural networks in the medical field is becoming more and more extensive. Hospitals usually directly use various trained neural networks provided by suppliers to process medical data. However, in clinical use, in order to protect the privacy of users, etc., hospitals have strict data control, and data corresponding to some new clinical scenarios cannot be delivered to suppliers in a timely manner, such as the application of new drugs in hospitals and the establishment of new workflows, resulting in poor generalization capabilities of neural networks provided by suppliers. In this regard, the present application provides a method for determining a medical data processing network.

[0048] The technical solution of the present application and how the technical solution of the present application solves the technical problem are described in detail below with specific embodiments.

[0049] The method for determining a medical data processing network provided in the embodiment of the present application can be applied to a medical system, such as Figure 1 As shown, the medical system 10 includes a terminal 11 arranged at the hospital end, and a terminal or server 12 arranged at the supplier end, and the terminal 11 and the server 12 are communicatively connected. The server 12 at the supplier end transmits the trained neural network to the terminal 11 at the hospital end. The hospital staff can use the neural network to process the medical data based on the terminal 11 at the hospital end. The terminal 11 at the hospital end can be a computer device, which can be, but not limited to, an industrial computer, a laptop computer, and a tablet computer.

[0050] In one embodiment, Figure 2 As shown, a method for determining a medical data processing network is provided. This embodiment uses the method applied to a terminal in the above medical system as an example. In this embodiment, the steps of the method include:

[0051] Step 200: In response to a training instruction generated by a user triggering the generation on the application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network.

[0052] The application interface refers to the interface displayed by the application software for determining the medical data processing network displayed on the display interface of the terminal of the medical system. A training control of the medical data processing network is displayed on the application interface. The user triggers the training control of the medical data processing network on the application interface. The terminal generates a training instruction in response to the trigger operation, and obtains the first medical data based on the training instruction. The first medical data is the training sample required for training the target data processing network. The first medical data can be pre-stored in the storage device of the terminal by the user / user, and the terminal directly obtains the first medical data from the storage device based on the training instruction. The first medical data can also be input into the terminal by the user / user, and the terminal will send a prompt message to the user based on the training instruction to prompt the user / user to enter the first medical data. The first medical data can be a medical image, a medical video, or other medical information. This embodiment does not limit the specific method of obtaining the first medical data, as long as its function can be achieved.

[0053] The above-mentioned users may be internal users of the hospital, such as doctors, technicians, etc.; the above-mentioned users may also be external users of the hospital, such as equipment suppliers, software algorithm engineers, etc. In the case where the user is an external user of the hospital, the external user of the hospital can obtain the first medical data in real time, and use the first medical data to train the deep learning network, thereby improving the generalization ability of the deep learning network of the external institution of the hospital. In the case where the user is an internal user of the hospital, the internal user of the hospital can obtain the first medical data in the case of the application of some new drugs in the hospital and the establishment of new workflows, or other changes in the clinical application environment, and use the first medical data to perform unsupervised training on the data processing network to obtain the target data processing network. The target data processing network obtained in this way can adapt to the decentralized medical data in the hospital and the diversified clinical application scenarios, and has stronger generalization ability.

[0054] In an optional embodiment, the terminal includes application software for training the data processing network. When the user needs to determine the medical data processing network, the user clicks the icon of the application software, and the terminal enters the application interface in response to the click operation. A start control is displayed on the application interface, and the user triggers the start control (click operation or sliding operation, etc.), and the terminal generates a training instruction in response to the trigger operation, and obtains the first medical data based on the training instruction.

[0055] In another optional embodiment, a training control of the data processing network and a usage control of the data processing network are displayed on the application interface of the terminal. The user triggers the training control of the data processing network on the application interface (click operation or sliding operation, etc.), and the terminal generates a training instruction in response to the trigger operation, and obtains the first medical data based on the training instruction. The user triggers the usage control of the data processing network on the application interface (click operation or sliding operation, etc.), and the terminal generates a usage instruction in response to the trigger operation, and displays prompt information to the user based on the usage instruction to prompt the user to input the medical data to be processed; after the terminal receives the medical data to be processed input by the user, the data processing network stored in the terminal is used to process the medical data to be processed, and the data processing network is sent to the terminal by the server of the supplier (equipment manufacturer).

[0056] The data processing network can be an initial neural network pre-stored by the user, that is, a neural network after initialization processing. The data processing network can also be a data processing network obtained by the terminal from a server on the supplier (equipment manufacturer) side. After obtaining the first medical data, the terminal inputs the first medical data into the data processing network, performs unsupervised training on the data processing network, and obtains a target data processing network. In other words, the terminal uses the first medical data as a training sample to perform unsupervised training on the data processing network to obtain a target data processing network.

[0057] Step 210: Input the second medical data into the target data processing network for testing to obtain a test result.

[0058] The second medical data may be pre-stored in a storage device of the terminal by the user, and the terminal directly obtains the second medical data from the storage device when needed. The first medical data may also be input into the terminal by the user, and the terminal will send a prompt message to the user to prompt the user to input the second medical data. This embodiment does not limit the specific method of obtaining the second medical data, as long as its function can be achieved.

[0059] In an optional embodiment, the first medical data and the second medical data may be when the user / user stores the medical data in advance in the storage device of the terminal, and the stored medical data is directly divided into the first medical data and the second medical data. The first medical data and the second medical data may also be when the user / user stores the medical data in advance in the terminal, and the terminal divides the medical data in the storage device into the first medical data and the second medical data according to a preset rule. The preset rule may be a ratio of 7:3, that is, the terminal divides the medical data according to a ratio of 7:3, and after the division, the first medical data accounts for a larger proportion, and the second medical data accounts for a smaller proportion.

[0060] After acquiring the second medical data, the terminal inputs the second medical data into the target data processing network, tests the target data processing network, and obtains a test result, which is a result of the target data processing network processing the second medical data.

[0061] Step 220: Evaluate the performance of the target data processing network according to the test result to obtain an evaluation result, and if the evaluation result indicates that the performance of the target data processing network meets the requirements, store the network parameters of the target data processing network.

[0062] After obtaining the test result output by the target data processing network, the terminal evaluates the performance of the target data processing network according to the test result to obtain an evaluation result. The evaluation result is used to characterize whether the performance of the target data processing network meets the requirements, which are pre-set by the user. If the evaluation result indicates that the performance of the target data processing network meets the requirements, that is, the target data processing network obtained after unsupervised training can be directly used by the user, the network parameters of the target data processing network are stored. The network parameters of the target data processing network can be stored in the storage device of the terminal. This embodiment does not limit the specific method of evaluating the performance of the target data processing network according to the test results, as long as its function can be achieved.

[0063] In an optional embodiment, a trained evaluation model is pre-stored in a storage device of the terminal, and the terminal inputs the test result into the evaluation model, and outputs the evaluation result by analyzing the test result of the evaluation model.

[0064] The method for determining the medical data processing network provided in the embodiment of the present application is that the user can perform a trigger operation on the application interface to generate a training instruction, so as to obtain the first medical data to perform unsupervised training on the data processing network and obtain the target data processing network. In other words, the target data processing network can be trained on the user side, so that the generalization ability of the target data processing network is stronger. And in the present embodiment, the target data processing network can also be tested based on the second medical data, and the performance of the target data processing network can be evaluated by the test results. In this way, the network parameters of the target data processing network obtained can be guaranteed to be more accurate, so that the method for determining the medical data processing network has higher reliability and practicality. In addition, the present embodiment performs unsupervised training on the data processing network, so there is no need to obtain a large number of training samples (first medical data), there is no need to annotate the first medical data, and the hardware requirements are low, which can save training costs to a great extent.

[0065] In one embodiment, the step of determining the medical data processing network further comprises:

[0066] Display the test results on the application interface.

[0067] After the terminal inputs the second medical data into the target data processing network for testing and obtains the test result, the terminal can display the obtained test result on the application interface so that the user can view the test result on the application interface.

[0068] In this case, if Figure 3 As shown, it is related to evaluating the performance of the target data processing network according to the test results to obtain an implementation method of the evaluation results, and the steps of the implementation method include:

[0069] Step 300: If a confirmation instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network meets the requirements.

[0070] When the test results are displayed on the application interface, controls that facilitate user evaluation are displayed for the test results. The controls may include pass controls and fail controls corresponding to the test results, and may also include confirmation controls and negation controls corresponding to the test results. The user observes and evaluates the test results, and triggers the controls corresponding to the test results displayed on the application interface to determine the evaluation results.

[0071] If the user performs a trigger operation on a control (confirm control) on the application interface, the terminal generates a confirmation instruction in response to the trigger operation, and the terminal obtains an evaluation result based on the confirmation instruction, which indicates that the performance of the target data processing network meets the requirements.

[0072] Step 310: If a retraining instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

[0073] If the user triggers the failed control (negative control) on the application interface, the terminal generates a retraining instruction in response to the triggering operation, and the terminal obtains an evaluation result based on the retraining instruction, and the evaluation result indicates that the performance of the target data processing network does not meet the requirements. The terminal obtains new first medical data based on the retraining instruction, and returns to perform the step of performing unsupervised training on the data processing network according to the first medical data to obtain the target data processing network in step 200 based on the new medical data, and executes steps 210 and 220.

[0074] In this embodiment, the test results are displayed on the application interface, and the performance of the target data processing network is evaluated through interaction with the user, which makes the method for determining the medical data processing network more practical and reliable.

[0075] In one embodiment, the step of determining the medical data processing network further comprises:

[0076] When the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned to be executed.

[0077] When the terminal evaluates the performance of the target data processing network according to the test results and the evaluation result indicates that the performance of the target data processing network does not meet the requirements, indicating that the currently trained target data processing network cannot accurately process the medical data, the terminal will reacquire new first medical data and perform unsupervised training on the data processing network based on the new first medical data, that is, return the new first medical data as the first medical data to perform unsupervised training on the data processing network according to the first medical data in step 200 to obtain the target data processing network, and execute steps 210 and 220.

[0078] The first medical data reacquired may be directly acquired from a storage device of the terminal, or the terminal may send a prompt message to the user, and the user may input the message based on the prompt message. When the data processing network is unsupervisedly trained using new medical data, the data processing network may be a new initial neural network, or a target data processing network obtained from the previous training.

[0079] In this embodiment, when the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data will be obtained, and a new target data processing network will be retrained and then evaluated. This cyclic processing method can obtain a target data processing network with performance that meets the requirements, making the method for determining the medical data processing network more practical and reliable.

[0080] In one embodiment, Figure 4 As shown, an implementation method for obtaining first medical data includes the following steps:

[0081] Step 400: Acquire initial medical data.

[0082] The terminal obtains initial medical data. The initial medical data may be pre-stored by the user in a storage device of the terminal, input by the user, or obtained from a PACS (Picture Archiving and Communication Systems) server of the hospital.

[0083] Step 410, determine whether the initial medical data meets the training conditions; the training conditions include whether the amount of the initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold.

[0084] The training conditions are pre-set by the user / subscriber based on actual experience and the input requirements of the data processing network, and stored in the storage device of the terminal. The training conditions may include the amount of medical data required for training the data processing network, that is, whether the amount of initial medical data reaches the preset data threshold; the size of the input medical data required by the input end of the data processing network, that is, whether the size of the initial medical data meets the preset size threshold; whether the medical data obtained by the terminal is the same as the input by the user or stored in the storage device, that is, whether the obtained initial medical data is accurate.

[0085] After acquiring the initial medical data, the terminal determines whether the initial medical data meets the training conditions. That is, after acquiring the initial medical data, the terminal determines the amount of the acquired initial medical data, compares the amount with the preset amount threshold, and determines whether the amount reaches the preset amount threshold; acquires the size of the initial medical data, compares the size with the preset size threshold, and determines whether the size is equal to the preset size threshold; analyzes the initial medical data to determine whether the initial medical data is accurate (complete). If the initial medical data is a medical image, the size of the initial medical data refers to the size of the medical image.

[0086] In an optional embodiment, the initial medical data corresponds to the memory occupied by the initial medical data during the storage process. The terminal determines the memory occupied by the initial medical data by analyzing the initial medical data, and compares the memory occupied by the initial medical data with the memory occupied by the acquired initial medical data to determine whether the acquired initial medical data is accurate.

[0087] Step 420: If the initial medical data meets the training condition, the initial medical data is determined as the first medical data.

[0088] If the terminal determines that the initial medical data meets the training conditions, it means that the initial medical data can be directly used to train the data processing network, and the initial medical data is determined as the first medical data.

[0089] Step 430: If the initial medical data does not meet the training condition, new initial medical data is obtained, and based on the new initial medical data, the step of returning to determine whether the initial medical data meets the training condition is executed.

[0090] If the terminal determines that the initial medical data does not meet the training conditions, indicating that the initial medical data cannot be used to train the data processing network, new initial medical data is obtained, and based on the new initial medical data, the execution of steps 410 to 430 is returned. The description of obtaining new initial medical data can refer to the specific description of obtaining initial medical data above, which will not be repeated here.

[0091] In this embodiment, after the initial medical data is obtained, it is determined whether the initial medical data meets the training conditions. This ensures that the first medical data used in training the data processing network is accurate, thereby ensuring that the target data processing network obtained is more accurate, and thus making the method for determining the medical data processing network more reliable and practical.

[0092] In one embodiment, Figure 5 As shown, an implementation method of storing network parameters of a target data processing network is involved, and the steps of the implementation method include:

[0093] Step 500: Mark the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters.

[0094] The type of the target data processing network may be determined according to the functions that the target data processing network can realize. After determining the target data processing network, the terminal will obtain the network parameters of the target data processing network, and mark the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters. Different types correspond to different marks. This embodiment does not limit the specific method of marking the network parameters of the target data processing network, as long as its function can be realized.

[0095] In an optional embodiment, the terminal may add a recognizable field to the network parameters of the target data processing network to obtain the marked network parameters. The recognizable field may be letters, numbers, etc.

[0096] In another optional embodiment, the terminal may establish an association relationship between the network parameters of the target data processing network and the corresponding labels to obtain the labeled network parameters. The labels may be letters, numbers, etc.

[0097] Assume that the network parameter is represented by B, and the function that the target data processing network can realize is to segment medical images. The type of the target data processing network is a segmentation network, and the type of segmentation network is marked as 1. The marked network parameters can be represented as 1B, or 1-B (there is a correlation between 1 and B).

[0098] Step 510: Store the marked network parameters.

[0099] After obtaining the marked network parameters, the terminal stores the marked network parameters, that is, the marked network parameters are stored in a storage device of the terminal.

[0100] In this embodiment, when storing the network parameters of the target data processing network, the network parameters are marked and the marked network parameters are stored. In this way, when the network parameters are subsequently acquired, they are acquired according to the marks, which facilitates the rapid acquisition of the network parameters and enables the medical data processing network to have higher practicality and reliability.

[0101] In one embodiment, Figure 6 As shown, a medical data processing method is provided. This embodiment uses the method applied to a terminal in the above medical system as an example. In this embodiment, the steps of the method include:

[0102] Step 600: Obtain medical data to be processed.

[0103] The terminal obtains the medical data to be processed. The medical data to be processed may be pre-stored by the user in the storage device of the terminal, and the terminal directly obtains it from the storage device when needed. The medical data to be processed may also be input by the user in real time. The medical data to be processed may be a medical image to be processed, a medical video to be processed, or other medical information to be processed. This embodiment does not limit the type of medical data to be processed and the specific method of obtaining the medical data to be processed, as long as its function can be achieved.

[0104] Step 610: in response to a selection instruction generated by a user triggering on an application interface, a target network parameter indicated by the selection instruction is obtained, and a basic neural network model is obtained, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each network parameter is obtained according to the above-mentioned determination method of the medical data processing network;

[0105] The application interface refers to an interface for processing the medical data to be processed displayed on the display interface of the terminal. A network selection control is displayed on the application interface. The user triggers the network selection control on the application interface to generate a selection instruction, and obtains the target network parameter based on the selection instruction. The target network parameter is the network parameter indicated by the selection instruction, that is, the network parameter selected from multiple network parameters indicated by the selection instruction. The multiple network parameters are obtained using the above-mentioned method for determining the medical data processing network, which will not be repeated here.

[0106] In an optional embodiment, the application interface includes a self-training network usage control, and the user performs a trigger operation (click operation or sliding operation, etc.) on the self-training network usage control on the application interface, and the terminal displays a network parameter display interface in response to the trigger operation. A plurality of network parameters are displayed on the network parameter display interface. The user performs a trigger operation (click operation or sliding operation, etc.) on the target network parameter on the network parameter display interface, and the terminal generates a selection instruction in response to the trigger operation, and obtains the target network parameter based on the selection instruction.

[0107] The basic neural network model may refer to an untrained initial neural network model. The structure of the basic neural network model is the same as the structure of the target data processing network model. After obtaining the basic neural network model, the terminal uses the obtained target network parameters to replace the network parameters in the basic neural network model to obtain the target neural network model.

[0108] Step 620: input the medical data to be processed into the target neural network model for data processing to obtain a processing result.

[0109] After obtaining the target neural network model, the terminal inputs the medical data to be processed into the neural network model, and processes the data through the neural network model to obtain the processing result.

[0110] The medical data processing method provided in the embodiment of the present application obtains the medical data to be processed, responds to the selection instruction generated by the user triggering on the application interface, obtains the target network parameters indicated by the selection instruction, and obtains the basic neural network model, and obtains the target neural network model according to the target network parameters and the basic neural network model; the medical data to be processed is input into the target neural network model for data processing to obtain the processing result. In this embodiment, the target network parameters are obtained from a plurality of network parameters determined by the determination method of the medical data processing network provided in the above embodiment, that is, when processing the medical data to be processed, the target neural network model corresponding to the network parameters obtained by the terminal self-training can be used, and the target neural network model has strong generalization ability and can process the medical data to be processed to obtain more accurate processing results.

[0111] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0112] See also Figure 7 An embodiment of the present application provides a medical data processing system 20, which includes a medical data transmission unit 21, a network training unit 22, a network storage unit 23, and a performance evaluation unit 24.

[0113] The medical data transmission unit 21 is used to transmit the first medical data and the second medical data to the network training unit 22. The network training unit 22 is used to perform unsupervised training on the data processing network based on the first medical data to obtain the target data processing network; and input the second medical data into the target data processing network for testing to obtain the test result; and transmit the test result to the performance evaluation unit 24. The performance evaluation unit 24 is used to evaluate the performance of the target data processing network according to the test result; if it is determined that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are transmitted to the network storage unit 23; if it is determined that the performance of the target data processing network does not meet the requirements, the medical data transmission unit 21 is used to transmit new first medical data and second medical data to the network training unit 22, and the network training unit 22 re-executes the corresponding measures. The network storage unit 23 is used to receive network parameters and mark and store the network parameters.

[0114] The network training unit 22 is also used to transmit the training parameters in the process of unsupervised training of the data processing network based on the first medical data to the network storage unit 23. The network storage unit 23 can also be used to save the training parameters (such as loss function) so as to display the training parameters to the user.

[0115] Based on the same inventive concept, the embodiment of the present application also provides a device for determining a medical data processing network for implementing the method for determining a medical data processing network involved above, and a medical data processing device for the medical data processing method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more devices for determining a medical data processing network and medical data processing devices provided below can refer to the limitations of the method for determining a medical data processing network and the medical data processing method above, and will not be repeated here.

[0116] In one embodiment, Figure 8 As shown, a device 30 for determining a medical data processing network is provided, comprising: an acquisition module 31, a test module 32 and an evaluation module 33, wherein:

[0117] The acquisition module 31 is used to acquire the first medical data in response to the training instruction generated by the user triggering on the application interface, and perform unsupervised training on the data processing network according to the first medical data to obtain the target data processing network.

[0118] The testing module 32 is used to input the second medical data into the target data processing network for testing to obtain a test result.

[0119] The evaluation module 33 is used to evaluate the performance of the target data processing network according to the test results to obtain an evaluation result, and if the evaluation result indicates that the performance of the target data processing network meets the requirements, store the network parameters of the target data processing network.

[0120] In one embodiment, the determination device of the medical data processing network further includes a display module. The display module is used to display the test results on the application interface. The evaluation module is specifically used to determine that the evaluation result indicates that the performance of the target data processing network meets the requirements if a confirmation instruction triggered by the user on the application interface is received; and determine that the evaluation result indicates that the performance of the target data processing network does not meet the requirements if a retraining instruction triggered by the user on the application interface is received.

[0121] In one embodiment, the evaluation module is further used to reacquire new first medical data when the evaluation result indicates that the performance of the target data processing network does not meet the requirements, and return to use the acquisition module to execute corresponding steps based on the new first medical data.

[0122] In one embodiment, the acquisition module 31 includes an acquisition unit, a judgment unit, a first determination unit, and a second determination unit. The acquisition unit is used to acquire initial medical data; the judgment unit is used to determine whether the initial medical data meets the training conditions; the training conditions include whether the amount of initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets the preset size threshold; the first determination unit is used to determine the initial medical data as first medical data if the initial medical data meets the training conditions; the second determination unit is used to acquire new initial medical data if the initial medical data does not meet the training conditions, and use the judgment unit to perform corresponding steps based on the new initial medical data.

[0123] In one embodiment, the evaluation module includes a storage unit, and the storage unit is used to mark the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters; and store the marked network parameters.

[0124] In one embodiment, Fig. 9 As shown, a medical data processing device 40 is provided, comprising: a medical data acquisition module 41, a network model acquisition module 42 and an input module 43, wherein:

[0125] The medical data acquisition module 41 is used to acquire the medical data to be processed.

[0126] The network model acquisition module 42 is used to respond to the selection instruction triggered by the user on the application interface, obtain the target network parameters indicated by the selection instruction, and obtain the basic neural network model, and obtain the target neural network model based on the target network parameters and the basic neural network model; the selection instruction is used to select the target network parameters from multiple network parameters; each network parameter is obtained using the determination device of the above-mentioned medical data processing network.

[0127] The input module 43 is used to input the medical data to be processed into the target neural network model for data processing to obtain processing results.

[0128] The above-mentioned determination device of the medical data processing network and each module in the medical data processing device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0129] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.10As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for determining a medical data processing network and a medical data processing method are implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0130] Those skilled in the art will understand that Fig.10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0131] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0132] In response to a training instruction generated by a user triggering on an application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network;

[0133] Inputting the second medical data into the target data processing network for testing to obtain a test result;

[0134] The performance of the target data processing network is evaluated according to the test result to obtain an evaluation result, and when the evaluation result indicates that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are stored.

[0135] In one embodiment, when the processor executes the computer program, the following steps are also implemented: displaying the test results on the application interface; if a confirmation instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network meets the requirements; if a retraining instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

[0136] In one embodiment, when the processor executes the computer program, the following steps are also implemented: when the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned.

[0137] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining initial medical data; determining whether the initial medical data meets the training conditions; the training conditions include whether the amount of initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold; if the initial medical data meets the training conditions, determining the initial medical data as first medical data; if the initial medical data does not meet the training conditions, obtaining new initial medical data, and returning to the step of determining whether the initial medical data meets the training conditions based on the new initial medical data.

[0138] In one embodiment, when executing the computer program, the processor further implements the following steps: marking the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters; and storing the marked network parameters.

[0139] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0140] Obtaining medical data to be processed;

[0141] In response to a selection instruction triggered by a user on an application interface, a target network parameter indicated by the selection instruction is obtained, as well as a basic neural network model, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each network parameter is obtained according to the above-mentioned determination method of the medical data processing network;

[0142] The medical data to be processed is input into the target neural network model for data processing to obtain the processing results.

[0143] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0144] In response to a training instruction generated by a user triggering on an application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network;

[0145] Inputting the second medical data into the target data processing network for testing to obtain a test result;

[0146] The performance of the target data processing network is evaluated according to the test result to obtain an evaluation result, and when the evaluation result indicates that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are stored.

[0147] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: displaying the test results on the application interface; if a confirmation instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network meets the requirements; if a retraining instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

[0148] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: when the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned.

[0149] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining initial medical data; determining whether the initial medical data meets the training conditions; the training conditions include whether the amount of initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold; if the initial medical data meets the training conditions, determining the initial medical data as first medical data; if the initial medical data does not meet the training conditions, obtaining new initial medical data, and returning to the step of determining whether the initial medical data meets the training conditions based on the new initial medical data.

[0150] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: marking the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters; and storing the marked network parameters.

[0151] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0152] Obtaining medical data to be processed;

[0153] In response to a selection instruction triggered by a user on an application interface, a target network parameter indicated by the selection instruction is obtained, as well as a basic neural network model, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each network parameter is obtained according to the above-mentioned determination method of the medical data processing network;

[0154] The medical data to be processed is input into the target neural network model for data processing to obtain the processing results.

[0155] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0156] In response to a training instruction generated by a user triggering on an application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network;

[0157] Inputting the second medical data into the target data processing network for testing to obtain a test result;

[0158] The performance of the target data processing network is evaluated according to the test result to obtain an evaluation result, and when the evaluation result indicates that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are stored.

[0159] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: displaying the test results on the application interface; if a confirmation instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network meets the requirements; if a retraining instruction triggered by the user on the application interface is received, determining that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

[0160] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: when the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned.

[0161] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining initial medical data; determining whether the initial medical data meets the training conditions; the training conditions include whether the amount of initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold; if the initial medical data meets the training conditions, determining the initial medical data as first medical data; if the initial medical data does not meet the training conditions, obtaining new initial medical data, and returning to the step of determining whether the initial medical data meets the training conditions based on the new initial medical data.

[0162] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: marking the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters; and storing the marked network parameters.

[0163] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0164] Obtaining medical data to be processed;

[0165] In response to a selection instruction triggered by a user on an application interface, a target network parameter indicated by the selection instruction is obtained, as well as a basic neural network model, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each network parameter is obtained according to the above-mentioned determination method of the medical data processing network;

[0166] The medical data to be processed is input into the target neural network model for data processing to obtain the processing results.

[0167] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0168] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0169] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for determining a medical data processing network, characterized in that: The method comprises: In response to a training instruction generated by a user triggering on an application interface, first medical data is acquired, and unsupervised training is performed on a data processing network according to the first medical data to obtain a target data processing network; Inputting the second medical data into the target data processing network for testing to obtain a test result; The performance of the target data processing network is evaluated according to the test result to obtain an evaluation result, and when the evaluation result indicates that the performance of the target data processing network meets the requirements, the network parameters of the target data processing network are stored.

2. The method according to claim 1, characterized in that The method further comprises: Displaying the test results on the application interface; The step of evaluating the performance of the target data processing network according to the test result to obtain an evaluation result includes: If a confirmation instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network meets the requirements; if a retraining instruction generated by the user on the application interface is received, it is determined that the evaluation result indicates that the performance of the target data processing network does not meet the requirements.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: When the evaluation result indicates that the performance of the target data processing network does not meet the requirements, new first medical data is reacquired, and based on the new first medical data, the step of performing unsupervised training on the data processing network according to the first medical data is returned to be executed.

4. The method according to claim 1 or 2, characterized in that: The acquiring of the first medical data comprises: Obtaining initial medical data; Determining whether the initial medical data meets training conditions; the training conditions include whether the amount of the initial medical data reaches a preset amount threshold, whether the initial medical data is accurate, and whether the size of the initial medical data meets a preset size threshold; If the initial medical data meets the training condition, determining the initial medical data as the first medical data; If the initial medical data does not meet the training condition, new initial medical data is acquired, and based on the new initial medical data, the step of determining whether the initial medical data meets the training condition is returned to be executed.

5. The method according to claim 1 or 2, characterized in that: The storing of the network parameters of the target data processing network includes: Marking the network parameters of the target data processing network according to the type of the target data processing network to obtain the marked network parameters; The marked network parameters are stored.

6. A medical data processing method, characterized in that: The method comprises: Obtaining medical data to be processed; In response to a selection instruction triggered by a user on an application interface, a target network parameter indicated by the selection instruction is obtained, as well as a basic neural network model, and a target neural network model is obtained according to the target network parameter and the basic neural network model; the selection instruction is used to select the target network parameter from a plurality of network parameters; each of the network parameters is obtained according to the method described in any one of claims 1 to 5; The medical data to be processed is input into the target neural network model for data processing to obtain a processing result.

7. A device for determining a medical data processing network, characterized in that: The device comprises: an acquisition module, configured to acquire first medical data in response to a training instruction generated by a user triggering the training instruction on the application interface, and to perform unsupervised training on the data processing network according to the first medical data to obtain a target data processing network; A testing module, used for inputting the second medical data into the target data processing network for testing to obtain a test result; An evaluation module is used to evaluate the performance of the target data processing network according to the test result to obtain an evaluation result, and if the evaluation result indicates that the performance of the target data processing network meets the requirements, store the network parameters of the target data processing network.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images