Hyperuricemia and gouty nephropathy classification method and apparatus, and electronic device

Through the method based on the classification prediction model of hyperuricemia and gouty nephropathy, the embedded layer blocking and coding layer fusion characteristics of B-type ultrasound images, combined with the attention mechanism, the early accurate diagnosis and staging of gouty nephropathy are achieved. The problem of difficulty in early diagnosis in the prior art has been solved, and screening efficiency and accuracy have been improved.

WO2024239701A1PCT designated stage expired Publication Date: 2024-11-28CHINA JAPAN FRIENDSHIP HOSPITAL
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2024/073969
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-23
Filing Date
2024-01-25
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

The prior art is difficult to accurately diagnose gouty nephropathy in the early stage, resulting in progress and increased treatment difficulty. There is a lack of a convenient and fast screening method to determine the hyperuricemia population and gouty nephropathy classification.

Method used

A classification method based on the classification prediction model of hyperuricemia and gouty nephropathy is proposed. By obtaining B-type ultrasound images of the kidney area, using embedded layer block processing, encoding layer fusion characteristics, combined with attention mechanism, input into the classification layer To obtain classification results for hyperuricemia and gouty nephropathy.

Benefits of technology

It realizes the rapid and accurate classification of B-type ultrasound images of the subjects to be tested, facilitates follow-up diagnosis and treatment, and improves the early diagnosis and staging accuracy of gouty nephropathy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2024073969_28112024_PF_FP_ABST
    Figure CN2024073969_28112024_PF_FP_ABST
Patent Text Reader

Abstract

A hyperuricemia and gouty nephropathy classification method and apparatus, and an electronic device. The method comprises: acquiring a B-type ultrasound image of a kidney area of an object to be predicted, and performing gout nephropathy prediction on the B-type ultrasound image by using a hyperuricemia and gout nephropathy classification prediction model to obtain a hyperuricemia and gout nephropathy classification result. In this way, the B-type ultrasound image is processed by the hyperuricemia and gouty nephropathy classification prediction model, so that a hyperuricemia patient can be conveniently and quickly screened, and the gouty nephropathy of the diagnosed patient is classified, thereby facilitating individualized diagnosis and treatment based on the classification result of the hyperuricemia and the gouty nephropathy classification prediction model in the subsequent staging diagnosis and treatment of the gouty nephropathy.
Need to check novelty before this filing date? Find Prior Art

Description

Method, device and electronic device for classifying hyperuricemia and gouty nephropathy Technical Field

[0001] The present application relates to the field of medical imaging, and specifically to a method, device and electronic equipment for classifying hyperuricemia and gouty nephropathy. Background Art

[0002] Gouty nephropathy, also known as uric acid nephropathy, is a kidney damage caused by hyperuricemia, a condition characterized by excessive uric acid production or decreased excretion. Clinical manifestations of gouty nephropathy may include uric acid stones, low-molecular-weight proteinuria, edema, nocturia, hypertension, elevated uric acid levels in the blood and urine, and renal tubular impairment.

[0003] Currently, the symptoms of early-stage gouty nephropathy are not obvious. If patients seek medical attention only after experiencing significant fatigue, anemia, or other symptoms, the disease may have already passed its early stages, making later treatment more difficult. After obtaining a B-mode ultrasound image of a patient's kidney area, how to conveniently and quickly screen for hyperuricemia to confirm and classify gouty nephropathy is crucial for the precise diagnosis and treatment of gouty nephropathy.

[0004] Summary of the Invention

[0005] To this end, one purpose of the present application is to propose a classification method, device and electronic equipment based on a classification prediction model for hyperuricemia and gouty nephropathy.

[0006] According to the first aspect of the present application, an embodiment of a classification method based on a hyperuricemia and gouty nephropathy classification prediction model is proposed, comprising: obtaining a B-type ultrasound image of the kidney area of ​​the object to be predicted; inputting the B-type ultrasound image into an embedding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform block processing on the B-type ultrasound image through the embedding layer, and determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block processing, and respectively perform splicing processing on the image feature vectors and position vectors corresponding to each of the multiple image blocks to obtain a spliced ​​feature vector corresponding to each of the multiple image blocks; inputting the spliced ​​feature vectors corresponding to each of the multiple image blocks into the encoding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform fusion processing on the image feature vectors corresponding to each of the multiple image blocks through the encoding layer to obtain a feature image of the B-type ultrasound image; and inputting the feature image into the classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain a hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image.

[0007] Optionally, as a first possible implementation method of the first aspect, the encoding layer includes: an attention sublayer and a coding sublayer, and the splicing feature vectors corresponding to each of the multiple image blocks are input into the coding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the encoding layer to obtain the characteristic image of the B-type ultrasound image, including: inputting the splicing feature vectors corresponding to each of the multiple image blocks into the attention sublayer, so as to determine the attention weight corresponding to each image block through the attention sublayer; in the coding sublayer, the splicing feature vectors corresponding to the corresponding image blocks are weightedly fused based on the attention weight corresponding to each image block to obtain the characteristic image of the B-type ultrasound image.

[0008] Optionally, as a second possible implementation method of the first aspect, the splicing feature vectors corresponding to each of the multiple image blocks are input into the attention sublayer to determine the attention weights corresponding to each of the image blocks through the attention sublayer, including: in the attention sublayer, according to the splicing feature vectors corresponding to each of the multiple image blocks, determining the correlation between each of the multiple image blocks; and according to the correlation, determining the attention weights corresponding to each of the image blocks.

[0009] Optionally, as a third possible implementation of the first aspect, the method also includes: obtaining a first image size supported by the hyperuricemia and gouty nephropathy classification prediction model; determining a second image size of the B-type ultrasound image; and when the first image size and the second image size are inconsistent, scaling the B-type ultrasound image so that the image size of the scaled B-type ultrasound image is the same as the first image size.

[0010] Optionally, as a fourth possible implementation method of the first aspect, the training method of the hyperuricemia and gouty nephropathy classification prediction model is: obtaining a B-type ultrasound image of the kidney area of ​​the sample object and the corresponding hyperuricemia and gouty nephropathy classification result; using the B-type ultrasound image of the kidney area of ​​the sample object as the input of the hyperuricemia and gouty nephropathy classification prediction model, and using the hyperuricemia and gouty nephropathy classification result corresponding to the sample object as the output of the hyperuricemia and gouty nephropathy classification prediction model, and training the hyperuricemia and gouty nephropathy classification prediction model.

[0011] According to the classification method based on the hyperuricemia and gouty nephropathy classification prediction model proposed in the embodiment of the present application, a B-type ultrasound image of the kidney area of ​​the object to be predicted is first obtained, and the B-type ultrasound image is input into the embedding layer of the hyperuricemia and gouty nephropathy classification prediction model. The B-type ultrasound image is divided into blocks through the embedding layer to determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block division, and the image feature vectors and position vectors corresponding to each of the multiple image blocks are respectively spliced ​​to obtain the spliced ​​feature vectors corresponding to each of the multiple image blocks. The spliced ​​feature vectors are then input into the encoding layer of the hyperuricemia and gouty nephropathy classification prediction model, and the image feature vectors corresponding to each of the multiple image blocks are fused through the encoding layer to obtain a feature image of the B-type ultrasound image. Finally, the feature image is input into the classification layer of the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image. Therefore, by processing the B-type ultrasound image through the hyperuricemia and gouty nephropathy classification prediction model, the hyperuricemia and gouty nephropathy classification result corresponding to the B-type ultrasound image of the object to be tested can be conveniently and quickly determined, thereby facilitating the subsequent application of the hyperuricemia and gouty nephropathy classification results of the hyperuricemia and gouty nephropathy classification prediction model in the subsequent staging diagnosis and treatment of gouty nephropathy.

[0012] According to the second aspect of the present application, an embodiment of a hyperuricemia and gouty nephropathy classification device is proposed, the device comprising: an acquisition module for acquiring a B-type ultrasound image of the kidney area of ​​the object to be predicted; an embedding layer module for inputting the B-type ultrasound image into the embedding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform block processing on the B-type ultrasound image through the embedding layer, and determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block processing, and respectively perform splicing processing on the image feature vectors and position vectors corresponding to each of the multiple image blocks. processing to obtain the splicing feature vectors corresponding to each of the multiple image blocks; a coding layer module, used to input the splicing feature vectors corresponding to each of the multiple image blocks into the coding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the coding layer to obtain the feature image of the B-type ultrasound image; a classification layer module, used to input the feature image into the classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image.

[0013] The embodiment of the present application provides a gouty nephropathy classification device based on a hyperuricemia and gouty nephropathy classification prediction model. First, a B-type ultrasound image of the kidney area of ​​the object to be predicted is obtained, and the B-type ultrasound image is input into an embedding layer in the hyperuricemia and gouty nephropathy classification prediction model. The B-type ultrasound image is divided into blocks by the embedding layer to determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block division, and the image feature vectors and position vectors corresponding to each of the multiple image blocks are spliced ​​to obtain spliced ​​feature vectors corresponding to each of the multiple image blocks. The spliced ​​feature vectors are then input into an encoding layer in the hyperuricemia and gouty nephropathy classification prediction model, and the image feature vectors corresponding to each of the multiple image blocks are fused through the encoding layer to obtain a feature image of the B-type ultrasound image. Finally, the feature image is input into a classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain a hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image. Therefore, by processing the B-type ultrasound image through the hyperuricemia and gouty nephropathy classification prediction model, the hyperuricemia and gouty nephropathy classification result corresponding to the B-type ultrasound image of the object to be tested can be conveniently and quickly determined, thereby facilitating the subsequent application of the hyperuricemia and gouty nephropathy classification results of the hyperuricemia and gouty nephropathy classification prediction model in the subsequent staging diagnosis and treatment of gouty nephropathy.

[0014] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0016] FIG1 is a flow chart of a classification method based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0017] FIG2 is a flow chart of another classification method based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0018] FIG3 is a flow chart of the training process of the hyperuricemia and gouty nephropathy classification prediction model provided in the embodiment of the present application.

[0019] FIG4 is a flow chart of another method for training a classification prediction model for hyperuricemia and gouty nephropathy according to an embodiment of the present application.

[0020] FIG5 is a schematic structural diagram of a classification device based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0021] FIG6 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0022] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0023] The following describes, with reference to the accompanying drawings, a classification method, device, and electronic device based on a classification prediction model for hyperuricemia and gouty nephropathy according to an embodiment of the present application.

[0024] FIG1 is a flow chart of a classification method based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0025] As shown in FIG1 , the method includes the following steps:

[0026] Step 101: Acquire a B-mode ultrasound image of the kidney region of the object to be predicted.

[0027] It should be noted that the classification method based on the hyperuricemia and gouty nephropathy classification prediction model in this embodiment is executed by a classification device based on the hyperuricemia and gouty nephropathy classification prediction model, and the classification device can be implemented by software and / or hardware, wherein the classification device can be an electronic device or can be configured in an electronic device.

[0028] Among them, the electronic device can be any device with computing capabilities, such as a personal computer, mobile terminal, server, etc. The mobile terminal can be, for example, a vehicle-mounted device, a mobile phone, a tablet computer, a personal digital assistant, a wearable device, and other hardware devices with various operating systems, touch screens and / or display screens.

[0029] Among them, the classification device in this example is used to classify and predict hyperuricemia and gouty nephropathy.

[0030] The object to be predicted is a virtual object to be predicted. A virtual object is a virtual object created for a real-world medical object. The medical object can be a patient (human) or other animal, such as a cat or dog. The virtual object can be represented by a patient identifier. The patient identifier can be a patient account number, a patient ID (number or identification card number), or a patient name.

[0031] In this example, the medical object is a patient as an example for illustrative description.

[0032] In some examples, an ultrasound image acquisition device may be used to inspect and detect the kidney region of a patient corresponding to the object to be predicted, so as to obtain a B-mode ultrasound image of the kidney region corresponding to the object to be predicted.

[0033] As an exemplary embodiment, the hyperuricemia and gouty nephropathy classification prediction model has size requirements for the input image. In order to make the input image meet the requirements of the hyperuricemia and gouty nephropathy classification prediction model, after obtaining the B-type ultrasound image of the kidney area of ​​the object to be predicted, the first image size supported by the hyperuricemia and gouty nephropathy classification prediction model is obtained; the second image size of the B-type ultrasound image is determined, and when the first image size and the second image size are inconsistent, the B-type ultrasound image is scaled so that the image size of the scaled B-type ultrasound image is the same as the first image size.

[0034] It is understood that the training of the hyperuricemia and gouty nephropathy classification prediction model has been completed before the hyperuricemia and gouty nephropathy classification prediction model is officially used. In other words, the hyperuricemia and gouty nephropathy classification prediction model is pre-trained.

[0035] The hyperuricemia and gouty nephropathy classification prediction model in this example can be established based on a vision transformer (VIT) model.

[0036] Step 102: Input the B-type ultrasound image into the embedding layer of the hyperuricemia and gouty nephropathy classification prediction model, so as to perform block processing on the B-type ultrasound image through the embedding layer, determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block processing, and respectively perform splicing processing on the image feature vectors and position vectors corresponding to each of the multiple image blocks to obtain spliced ​​feature vectors corresponding to each of the multiple image blocks.

[0037] It should be noted that, in the embodiment of the present application, the embedding layer performs block processing on the B-type ultrasound image based on a preset size to obtain a plurality of image blocks.

[0038] The preset size is a size pre-set in the embedding layer, for example, the preset size may be 16 x 16. It is understood that the preset size may be pre-set in the embedding layer according to actual needs, and this embodiment does not specifically limit this.

[0039] For example, the preset size is 16x16, and the size of the B-type ultrasound image of the embedding layer in the classification prediction model of hyperuricemia and gouty nephropathy is 224x224. Correspondingly, the embedding layer can divide the B-type ultrasound image into blocks according to the size of 16x16 to obtain 196 image blocks.

[0040] In some examples, each image block may be represented by a vector to obtain an image feature vector corresponding to each image block. Specifically, each image feature vector may be determined based on all pixel values ​​in each channel of each image block. For example, for each image block, all pixel values ​​in each channel of the image block may be integrated into a vector to obtain the image feature vector corresponding to the image block.

[0041] In this example, the position vector is obtained by performing vector representation on the position coding of each image block in the B-mode ultrasound image.

[0042] In step 103, the concatenated feature vectors corresponding to the multiple image blocks are input into the coding layer of the hyperuricemia and gouty nephropathy classification prediction model, so as to fuse the image feature vectors corresponding to the multiple image blocks through the coding layer to obtain a feature image of the B-type ultrasound image.

[0043] Step 104 : input the feature image into the classification layer of the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification results of the B-type ultrasound image.

[0044] Among them, the classification results of hyperuricemia in this example include: with hyperuricemia and without hyperuricemia; the classification results of gouty nephropathy include: without gouty nephropathy and with gouty nephropathy; the presence of gouty nephropathy can also include the following classification results: early stage gouty nephropathy, middle stage gouty nephropathy, and end stage gouty nephropathy; or the presence of gouty nephropathy can include the following classification results: gouty nephropathy stage 1, stage 2, stage 3, stage 4, and stage 5.

[0045] Among them, it is understandable that currently professional doctors are unable to directly diagnose the staging of gouty nephropathy based on B-type ultrasound images. This embodiment processes the B-type ultrasound image through the hyperuricemia and gouty nephropathy classification prediction model to determine the gouty nephropathy staging corresponding to the B-type ultrasound image, which is convenient for subsequent professional doctors to determine the final diagnosis result of the subject to be tested based on the results output by the hyperuricemia and gouty nephropathy classification prediction model and other medical examination results.

[0046] The hyperuricemia and gouty nephropathy classification method provided in the embodiment of the present application first obtains a B-type ultrasound image of the kidney area of ​​the object to be predicted, inputs the B-type ultrasound image into an embedding layer in a gouty nephropathy classification prediction model, blocks the B-type ultrasound image through the embedding layer to determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block division, and splices the image feature vectors and position vectors corresponding to each of the multiple image blocks to obtain a spliced ​​feature vector corresponding to each of the multiple image blocks, and then inputs the spliced ​​feature vector into an encoding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the encoding layer to obtain a feature image of the B-type ultrasound image, and finally inputs the feature image into a classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain a hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image. Therefore, by processing the B-type ultrasound image through the hyperuricemia and gouty nephropathy classification prediction model, the hyperuricemia and gouty nephropathy classification result corresponding to the B-type ultrasound image of the object to be tested can be conveniently and quickly determined, which facilitates the subsequent application of the hyperuricemia and gouty nephropathy classification results of the hyperuricemia and gouty nephropathy classification prediction model in the subsequent staging diagnosis and treatment of hyperuricemia and gouty nephropathy.

[0047] In some exemplary embodiments, in order to clearly understand how the encoding layer in the hyperuricemia and gouty nephropathy classification prediction model obtains the characteristic image of the B-type ultrasound image, the classification method based on the hyperuricemia and gouty nephropathy classification prediction model of the embodiment is further exemplarily described in combination with Figure 2.

[0048] FIG2 is a flow chart of another classification method based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0049] As shown in Figure 2, it may include:

[0050] Step 201: Input the concatenated feature vectors corresponding to the multiple image blocks into the attention sub-layer.

[0051] It can be understood that the encoding layer of the hyperuricemia and gouty nephropathy classification prediction model in this example includes an attention sublayer and an encoding sublayer.

[0052] Step 202: Determine the attention weight corresponding to each image block through the attention sub-layer.

[0053] In some exemplary embodiments, in the attention sublayer, the correlation between the multiple image blocks is determined based on the spliced ​​feature vectors corresponding to each of the multiple image blocks, and the attention weight corresponding to each image block is determined based on the correlation.

[0054] In some examples, for any two image blocks among the multiple image blocks, a correlation degree calculation may be performed based on the splicing feature vectors corresponding to the two image blocks to obtain the correlation degree between the two image blocks.

[0055] Step 203: In the coding sublayer, weighted fusion processing is performed on the spliced ​​feature vectors corresponding to the corresponding image blocks based on the attention weights corresponding to the respective image blocks to obtain a feature image of the B-type ultrasound image.

[0056] In some examples, the splicing feature vectors corresponding to the corresponding image blocks are weighted based on the attention weights corresponding to the respective image blocks to obtain the weighted splicing feature vectors, and the weighted splicing feature vectors corresponding to the respective image blocks are processed according to the position encoding of the respective image blocks in the B-type ultrasound image to obtain the feature image of the B-type ultrasound image.

[0057] In this example, the attention sublayer accurately determines the attention weights corresponding to each image block. In the coding sublayer, weighted fusion processing is performed on the concatenated feature vectors corresponding to each image block based on the attention weights corresponding to each image block to obtain the feature image of the B-type ultrasound image. This allows the feature image of the B-type ultrasound image to be accurately determined.

[0058] In some exemplary embodiments, in order to clearly understand how the hyperuricemia and gouty nephropathy classification prediction model is trained, the training of the hyperuricemia and gouty nephropathy classification prediction model of the embodiment is further exemplified below in conjunction with Figure 3.

[0059] FIG3 is a schematic diagram of a process for training a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application;

[0060] As shown in Figure 3, the process of training the hyperuricemia and gouty nephropathy classification prediction model may include:

[0061] Step 301 : Acquire a B-mode ultrasound image of the kidney region of a sample object and a classification result of hyperuricemia and gouty nephropathy corresponding to the sample object.

[0062] The sample object is a virtual object selected as a sample. A virtual object is a virtual object created for a real-world medical subject. The medical subject can be a patient (human) or other animal, such as a cat or dog. The virtual object can represent a patient identifier. The patient identifier can be a patient account number, patient ID (number or identification card number), or patient name.

[0063] Step 302: Using the B-mode ultrasound image of the kidney region of the sample subject as input to a classification prediction model for hyperuricemia and gouty nephropathy.

[0064] Step 303: The classification results of hyperuricemia and gouty nephropathy corresponding to the sample object are used as the output of the hyperuricemia and gouty nephropathy classification prediction model.

[0065] Step 304 , completing the training of the hyperuricemia and gouty nephropathy classification prediction model based on the B-mode ultrasound image of the kidney region of the sample object and the hyperuricemia and gouty nephropathy classification results corresponding to the sample object.

[0066] In this example, the hyperuricemia and gouty nephropathy classification prediction model is trained multiple times using a large number of sample objects and the hyperuricemia and gouty nephropathy classification results corresponding to the sample objects to obtain a trained hyperuricemia and gouty nephropathy classification prediction model, which can facilitate the subsequent classification of gouty nephropathy in the B-type ultrasound image of the kidney area corresponding to the test sample based on the trained hyperuricemia and gouty nephropathy classification prediction model.

[0067] In order to clearly understand the present application, the process of training the hyperuricemia and gouty nephropathy classification prediction model of this embodiment is exemplarily described below with reference to FIG4 .

[0068] FIG4 is a flow chart of another method for training a classification prediction model for hyperuricemia and gouty nephropathy according to an embodiment of the present application.

[0069] As shown in Figure 4, the process of training the hyperuricemia and gouty nephropathy classification prediction model is as follows:

[0070] Step 401: Acquire a preset number of virtual objects at different stages of the five stages of confirmed hyperuricemia and gouty nephropathy, acquire a preset number of virtual objects in a healthy kidney state, and acquire B-mode ultrasound images of the kidney areas of all the acquired virtual objects.

[0071] Among them, the preset number is the number set according to demand. For example, the above preset number can be 100, or 200, etc. This embodiment does not specifically limit the value of the preset number. The value of the preset number can be set according to actual needs. This embodiment does not specifically limit this.

[0072] Step 402 : annotate the B-mode ultrasound images of the kidney regions of all virtual objects, and generate a data set based on the B-mode ultrasound images of the kidney regions of all virtual objects and the corresponding hyperuricemia and gouty nephropathy classification annotation results.

[0073] Step 403: Use a first preset proportion of data in the data set as a training data set, and use a second preset proportion of data in the data set as a test data set.

[0074] It can be understood that the value obtained by adding the first preset ratio and the second preset ratio is equal to 1. The first preset ratio and the second preset ratio are preset ratios when training the hyperuricemia and gouty nephropathy classification prediction model. For example, the first preset ratio can be 80%, and the second preset ratio can be 20%. For another example, the first preset ratio can be 75%, and the second preset ratio can be 25%. It can be understood that the values ​​of the first preset ratio and the second preset ratio can be preset according to actual needs, and this embodiment does not specifically limit this.

[0075] Step 404 : training the initial VIT model based on the training data set to obtain a classification prediction model for hyperuricemia and gouty nephropathy.

[0076] Step 405 : Input the B-type ultrasound images in the test data set into the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification prediction results corresponding to the B-type ultrasound images.

[0077] Step 406 , based on the hyperuricemia and gouty nephropathy classification prediction results of the corresponding B-type ultrasound images and the corresponding hyperuricemia and gouty nephropathy classification labeling results, determine the preset performance indicators of the hyperuricemia and gouty nephropathy classification prediction model.

[0078] Among them, the preset performance indicators may include classification accuracy index, specificity index, sensitivity index, etc.

[0079] The calculation method of the index value corresponding to the classification accuracy index is as follows: The index value corresponding to the classification accuracy index = (TP+TN) / (TP+FP+FN+TN)

[0080] The calculation method of the sensitivity index value is as follows: Sensitivity = TP / (TP+FN)

[0081] The calculation method of the index value of the specificity index is as follows: Specificity = TN / (TN+FP)

[0082] Where TP represents the number of samples in the test data set that are actually positive and are classified as positive by the hyperuricemia and gouty nephropathy classification prediction model; TN represents the number of samples in the test data set that are actually negative and are classified as negative by the hyperuricemia and gouty nephropathy classification prediction model; FP represents the number of samples in the test data set that are actually negative and are classified as positive by the hyperuricemia and gouty nephropathy classification prediction model; FN represents the number of samples in the test data set that are actually positive but are classified as negative by the hyperuricemia and gouty nephropathy classification prediction model; and N represents the total number of test samples in the test data set. The closer each indicator is to 1, the better the performance of the method.

[0083] In some examples, the prediction performance indicators of various gouty nephropathy stages obtained by testing the hyperuricemia and gouty nephropathy classification prediction models based on the test data set are shown in Table 1.

[0084] Table 1 Prediction performance of the example for gouty chronic kidney disease stage

[0085] Based on the above table, it can be seen that the hyperuricemia and gouty nephropathy classification prediction model trained in this example can complete the prediction of gouty nephropathy stages, and the accuracy rate is above 80%.

[0086] It should be noted that the gouty nephropathy in FIG4 refers to chronic gouty nephropathy.

[0087] Step 407 : When it is determined that the preset performance index of the hyperuricemia and gouty nephropathy classification prediction model meets the preset requirements, the hyperuricemia and gouty nephropathy classification prediction model is determined to be a valid model.

[0088] In some examples, when the preset performance indicators include the above three performance indicators (classification accuracy index, specificity index, sensitivity index), it can be determined whether the index values ​​corresponding to the above three performance indicators are all greater than their respective corresponding preset thresholds. If the index values ​​of the above three performance indicators are all greater than their respective corresponding preset thresholds, the hyperuricemia and gouty nephropathy classification prediction model is determined to be a valid model.

[0089] The preset thresholds corresponding to the three performance indicators may be the same or different, and this embodiment does not impose any specific limitation on this.

[0090] Among them, it can be understood that when the hyperuricemia and gouty nephropathy classification prediction model is a valid model, it means that the B-type ultrasound image of the object to be predicted can be processed by the hyperuricemia and gouty nephropathy classification prediction model, and the hyperuricemia and gouty nephropathy classification results of the B-type ultrasound image of the object to be predicted can be accurately determined.

[0091] FIG5 is a schematic structural diagram of a classification device based on a classification prediction model for hyperuricemia and gouty nephropathy provided in an embodiment of the present application.

[0092] As shown in FIG5 , the classification device based on the hyperuricemia and gouty nephropathy classification prediction model includes: an acquisition module 51 , an embedding layer module 52 , a coding layer module 53 , and a classification layer module 54 , wherein:

[0093] The acquisition module 51 is used to acquire a B-mode ultrasound image of the kidney region of the object to be predicted.

[0094] The embedding layer module 52 is used to input the B-type ultrasound image into the embedding layer of the hyperuricemia and gouty nephropathy classification prediction model, so as to block the B-type ultrasound image through the embedding layer, determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block division, and respectively splice the image feature vectors and position vectors corresponding to each of the multiple image blocks to obtain the spliced ​​feature vectors corresponding to each of the multiple image blocks.

[0095] The coding layer module 53 is used to input the spliced ​​feature vectors corresponding to each of the multiple image blocks into the coding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the coding layer to obtain a feature image of the B-type ultrasound image.

[0096] The classification layer module 54 is used to input the feature image into the classification layer of the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification results of the B-type ultrasound image.

[0097] Furthermore, the encoding layer module 53 includes: an attention sub-layer unit 531 and an encoding sub-layer unit 532.

[0098] The attention sub-layer unit 531 is used to input the spliced ​​feature vectors corresponding to the multiple image blocks into the attention sub-layer, so as to determine the attention weight corresponding to each image block through the attention sub-layer.

[0099] The coding sublayer unit 532 is configured to perform weighted fusion processing on the spliced ​​feature vectors corresponding to the corresponding image blocks based on the attention weights corresponding to the respective image blocks, so as to obtain a feature image of the B-type ultrasound image.

[0100] In one embodiment of the present application, the attention sub-layer unit 531 is specifically configured to:

[0101] In the attention sublayer, the correlation between multiple image blocks is determined based on the concatenated feature vectors corresponding to each of the multiple image blocks.

[0102] According to the correlation degree, the attention weight corresponding to each image block is determined.

[0103] In one embodiment of the present application, the classification device based on the hyperuricemia and gouty nephropathy classification prediction model may further include:

[0104] An image size processing module is used to obtain a first image size supported by the classification prediction model for hyperuricemia and gouty nephropathy; determine a second image size of the B-type ultrasound image; and, when the first image size and the second image size are inconsistent, scale the B-type ultrasound image so that the image size of the scaled B-type ultrasound image is the same as the first image size.

[0105] In one embodiment of the present application, the training method of the hyperuricemia and gouty nephropathy classification prediction model is as follows: obtain a B-type ultrasound image of the kidney area of ​​the sample object and the corresponding actual classification results of hyperuricemia and gouty nephropathy; use the B-type ultrasound image of the kidney area of ​​the sample object as the input of the hyperuricemia and gouty nephropathy classification prediction model, and use the actual classification results of gouty nephropathy as the output of the hyperuricemia and gouty nephropathy classification prediction model to train the gouty nephropathy classification prediction model.

[0106] It should be noted that the above explanation of the method embodiment is also applicable to the device of this embodiment and will not be repeated here.

[0107] The hyperuricemia and gouty nephropathy classification device provided in an embodiment of the present application first obtains a B-type ultrasound image of the kidney area of ​​the object to be predicted, inputs the B-type ultrasound image into an embedding layer in a hyperuricemia and gouty nephropathy classification prediction model, blocks the B-type ultrasound image through the embedding layer to determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block division, and splices the image feature vectors and position vectors corresponding to each of the multiple image blocks to obtain a spliced ​​feature vector corresponding to each of the multiple image blocks, then inputs the spliced ​​feature vector into an encoding layer in the hyperuricemia and gouty nephropathy classification prediction model, fuses the image feature vectors corresponding to each of the multiple image blocks through the encoding layer to obtain a feature image of the B-type ultrasound image, and finally inputs the feature image into a classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain a hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image. Therefore, by processing the B-type ultrasound image through the hyperuricemia and gouty nephropathy classification prediction model, the hyperuricemia and gouty nephropathy classification result corresponding to the B-type ultrasound image of the object to be tested can be conveniently and quickly determined, thereby facilitating the subsequent application of the hyperuricemia and gouty nephropathy classification results of the hyperuricemia and gouty nephropathy classification prediction model in the subsequent staging diagnosis and treatment of hyperuricemia and gouty nephropathy.

[0108] An embodiment of the present application also provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements the classification method based on the hyperuricemia and gouty nephropathy classification prediction model of any of the aforementioned embodiments.

[0109] FIG6 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application.

[0110] The present application is described in more detail below using embodiments. The present application can be embodied in a variety of different forms and should not be construed as being limited to the exemplary embodiments described herein.

[0111] As shown in FIG6 , the electronic device 600 may include a memory 601, a processor 602, a communication interface 603, and a bus architecture 604. The memory 601 is used to store computer-executable instructions for executing the solution of the present application, and the execution is controlled by the processor 602. The processor 602 is used to execute the computer-executable instructions stored in the memory 601, thereby implementing a method for classifying hyperuricemia and gouty nephropathy provided in the above embodiment of the present application.

[0112] Communication interface 603, processor 602, and memory 601 can be interconnected via a bus architecture 604. Bus architecture 604 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. Bus architecture 604 can be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, FIG6 shows only one thick line, but this does not indicate that there is only one bus or only one type of bus.

[0113] The processor 602 may be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the program of the present application.

[0114] The communication interface 603 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), wired access network, etc.

[0115] The memory 601 can be a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, an optical disc storage (including a compact disc, a laser disc, an optical disc, a digital versatile disc, a Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory can exist independently and be connected to the processor via the bus architecture 804. The memory can also be integrated with the processor.

[0116] This electronic device exists in many forms, including but not limited to:

[0117] (1) Mobile communication devices: These devices are characterized by their mobile communication capabilities and are primarily designed to provide voice and data communications. These terminals include smartphones (e.g., iPhones), multimedia phones, feature phones, and low-end phones.

[0118] (2) Ultra-mobile personal computer devices: These devices fall under the category of personal computers, have computing and processing capabilities, and generally also have mobile Internet access. These terminals include PDAs, MIDs, and UMPCs, such as the iPad.

[0119] (3) Portable entertainment devices: These devices can display and play multimedia content. These devices include audio and video players (such as iPods), handheld game consoles, e-books, smart toys, and portable car navigation devices.

[0120] (4) Server: A device that provides computing services. The server consists of a processor, hard disk, memory, system bus, etc. The server is similar to a general computer architecture, but because it needs to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.

[0121] (5) Other electronic devices with data interaction functions.

[0122] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0123] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0124] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0125] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing it in a suitable manner if necessary, and then storing it in a computer memory.

[0126] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0127] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0128] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0129] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A classification method based on a classification prediction model for hyperuricemia and gouty nephropathy, characterized in that: The method comprises: Acquire a B-mode ultrasound image of the kidney region of the object to be predicted; Inputting the B-type ultrasound image into the embedding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform block processing on the B-type ultrasound image through the embedding layer, and determining image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block processing, and respectively performing splicing processing on the image feature vectors and position vectors corresponding to each of the multiple image blocks, so as to obtain splicing feature vectors corresponding to each of the multiple image blocks; Inputting the spliced ​​feature vectors corresponding to each of the multiple image blocks into the coding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform fusion processing on the image feature vectors corresponding to each of the multiple image blocks through the coding layer, so as to obtain a feature image of the B-type ultrasound image; The characteristic image is input into the classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification result of the B-type ultrasound image.

2. The method according to claim 1, characterized in that: The coding layer includes: an attention sublayer and a coding sublayer, and the splicing feature vectors corresponding to each of the multiple image blocks are input into the coding layer in the classification prediction model of hyperuricemia and gouty nephropathy, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the coding layer to obtain the feature image of the B-type ultrasound image, including: Inputting the concatenated feature vectors corresponding to the multiple image blocks into the attention sublayer, so as to determine the attention weights corresponding to the respective image blocks through the attention sublayer; In the coding sublayer, weighted fusion processing is performed on the spliced ​​feature vectors corresponding to the corresponding image blocks based on the attention weights corresponding to the respective image blocks, so as to obtain a feature image of the B-type ultrasound image.

3. The method according to claim 2, characterized in that Inputting the concatenated feature vectors corresponding to the plurality of image blocks to the attention sublayer to determine the attention weights corresponding to the respective image blocks through the attention sublayer includes: In the attention sublayer, determining the correlation between the plurality of image blocks according to the splicing feature vectors corresponding to the plurality of image blocks; According to the correlation degree, the attention weight corresponding to each of the image blocks is determined.

4. The method according to claim 1, characterized in that: The method further comprises: Acquire a first image size supported by the hyperuricemia and gouty nephropathy classification prediction model; determining a second image size of the B-mode ultrasound image; In the case that the first image size and the second image size are inconsistent, scaling processing is performed on the B-type ultrasound image so that the image size of the B-type ultrasound image after the scaling processing is the same as the first image size.

5. The method according to any one of claims 1 to 4, characterized in that The training method of the hyperuricemia and gouty nephropathy classification prediction model is: Obtaining a B-mode ultrasound image of the kidney region of the sample object and corresponding classification results of hyperuricemia and gouty nephropathy; The B-mode ultrasound image of the kidney area of ​​the sample object is used as the input of the hyperuricemia and gouty nephropathy classification prediction model, and the hyperuricemia and gouty nephropathy corresponding to the sample object is The kidney disease classification result is used as the output of the hyperuricemia and gouty nephropathy classification prediction model, and the hyperuricemia and gouty nephropathy classification prediction model is trained.

6. A classification device based on a classification prediction model for hyperuricemia and gouty nephropathy, characterized in that: The device comprises: An acquisition module, used for acquiring a B-mode ultrasound image of a kidney region of an object to be predicted; An embedding layer module, used for inputting the B-type ultrasound image into the embedding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform block processing on the B-type ultrasound image through the embedding layer, and determine the image feature vectors and position vectors corresponding to each of the multiple image blocks obtained by the block processing, and respectively perform splicing processing on the image feature vectors and position vectors corresponding to each of the multiple image blocks, so as to obtain the splicing feature vectors corresponding to each of the multiple image blocks; A coding layer module, used for inputting the spliced ​​feature vectors corresponding to each of the multiple image blocks into the coding layer in the hyperuricemia and gouty nephropathy classification prediction model, so as to perform fusion processing on the image feature vectors corresponding to each of the multiple image blocks through the coding layer, so as to obtain a feature image of the B-type ultrasound image; A classification layer module is used to input the feature image into the classification layer in the hyperuricemia and gouty nephropathy classification prediction model to obtain the hyperuricemia and gouty nephropathy classification results of the B-type ultrasound image.

7. The device according to claim 6, characterized in that The coding layer includes: an attention sublayer and a coding sublayer, and the splicing feature vectors corresponding to each of the multiple image blocks are input into the coding layer in the classification prediction model of hyperuricemia and gouty nephropathy, so as to fuse the image feature vectors corresponding to each of the multiple image blocks through the coding layer to obtain the feature image of the B-type ultrasound image, including:

8. The device according to any one of claims 6 to 7, characterized in that: The training method of the hyperuricemia and gouty nephropathy classification prediction model is: The B-mode ultrasound image of the kidney area of ​​the sample object is used as the input of the hyperuricemia and gouty nephropathy classification prediction model, and the hyperuricemia and gouty nephropathy classification result is used as the output of the hyperuricemia and gouty nephropathy classification prediction model to train the hyperuricemia and gouty nephropathy classification prediction model.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 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 5 are implemented.

Citation Information

Patent Citations

  • Network training method and device, image processing method and device, electronic equipment and storage medium

    CN111310764A

  • Medical ultrasonic image processing method, device and system

    CN112107332A

  • Ultrasonic breast lesion classification method and device and storage medium

    CN112699948A

  • Ultrasonic breast focus classification method and device and storage medium

    CN113239951A

  • Image classification method, and training method and device of image classification model

    CN114418030A