Ultrasonic scanning method, device, equipment and storage medium

By combining machine identification and user feedback, the final section type of ultrasonic scanning is determined, and the user is allowed to select the optimal section during the ultrasonic scanning process, which solves the problem of inaccurate section type in ultrasonic scanning and improves the accuracy of ultrasonic sections and reliability of subsequent use.

CN112529083BActive Publication Date: 2024-11-12SONOSCAPE MEDICAL CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011473795.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-12-15
Publication Date
2024-11-12
Estimated Expiration
2040-12-15

AI Technical Summary

Technical Problem

During the ultrasound scanning process, the ultrasound scanning area automatically determined by the device may have deviations, resulting in inaccurate ultrasound sections obtained and affect subsequent use.

Method used

By using the first classification model to identify the initial image, determine the initial section type, and combine user feedback information to determine the final section type, and enable ultrasonic scanning according to the sectional protocol of the final section type to ensure the accuracy of the section type. If the initial section type is inaccurate, adjust the model parameters to improve the recognition accuracy, and during the ultrasound scanning process, a second classification model is used to identify multiple standard sections, allowing the user to select the optimal section and enhance the accuracy of the ultrasound section.

Benefits of technology

The accuracy of ultrasonic sections obtained by ultrasonic scanning is improved, and the reliability and accuracy of subsequent use is ensured. Through the combination of machine recognition and user feedback, the recognition accuracy of the model is enhanced, and the problem of automatic identification deviation of the equipment is solved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112529083B_ABST
    Figure CN112529083B_ABST
Patent Text Reader

Abstract

The present application discloses an ultrasonic scanning method, comprising the following steps: obtaining an initial image for ultrasonic scanning; using a first classification model to identify the initial image and determine the initial section type of the initial image, the initial section type being the target part targeted by the current ultrasonic scanning; receiving user feedback on the accuracy of the initial section type; determining the final section type of the initial image based on the accuracy feedback information; and starting ultrasonic scanning according to the section protocol corresponding to the final section type. Applying the technical solution provided by the present application, the machine recognition and the user's feedback are combined to determine the final section type of the initial image, and the accuracy of the final section type is ensured. Only by starting ultrasonic scanning based on the section protocol corresponding to the accurately determined final section type can a more accurate ultrasonic section be obtained, which is convenient for subsequent use. The present application also discloses an ultrasonic scanning device, equipment and storage medium, which have corresponding technical effects.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of computer application technology, and in particular to an ultrasonic scanning method, device, equipment and storage medium. Background Art

[0002] With the rapid development of ultrasound medicine, ultrasound is being used more and more widely. Ultrasound has good resolution for human soft tissues and can obtain useful signals with a high dynamic range, which is conducive to identifying tiny lesions in biological tissues. Ultrasound sections can be obtained through ultrasound scanning, and ultrasound sections are gradually becoming an important basis for doctors' diagnosis and treatment, and their value in preventing, diagnosing, and treating diseases is becoming increasingly higher.

[0003] At present, during the ultrasound scanning process, the current ultrasound scanning area is mostly automatically determined by the equipment. The ultrasound scanning area is the ultrasound scanning type. If there is a deviation, it will cause a large deviation in the final ultrasound section, affecting subsequent use and causing inconvenience to subsequent use. Summary of the invention

[0004] The purpose of this application is to provide an ultrasonic scanning method, device, equipment and storage medium to accurately determine the type of section targeted by the current ultrasonic scanning, enhance the accuracy of the ultrasonic section obtained by the ultrasonic scanning, and facilitate subsequent use.

[0005] In order to solve the above technical problems, this application provides the following technical solutions:

[0006] An ultrasonic scanning method, comprising:

[0007] Obtaining initial ultrasound scan images;

[0008] Using a first classification model to identify the initial image, determine an initial section type of the initial image, and the initial section type is a target part targeted by the current ultrasound scan;

[0009] Receiving user feedback on the accuracy of the initial section type;

[0010] Determining a final section type of the initial image according to the accuracy feedback information;

[0011] Start ultrasound scanning according to the section protocol corresponding to the final section type.

[0012] In a specific implementation of the present application, determining the final section type of the initial image according to the accuracy feedback information includes:

[0013] Determining whether the initial section type is accurate according to the accuracy feedback information;

[0014] If accurate, determining the initial section type as the final section type of the initial image;

[0015] If it is inaccurate, a correction result in the accuracy feedback information is obtained, and the section type corresponding to the correction result is determined as the final section type of the initial image.

[0016] In a specific implementation of the present application, when it is determined that the initial section type is inaccurate, after obtaining the correction result in the accuracy feedback information, the method further includes:

[0017] The model parameters of the first classification model are adjusted based on the correction result so that the adjusted first classification model matches the correction result.

[0018] In a specific implementation of the present application, the accuracy feedback information is information fed back by the user through voice, gesture or input box.

[0019] In a specific embodiment of the present application, after starting the ultrasonic scanning according to the slice protocol corresponding to the final slice type, the method further includes:

[0020] During the ultrasound scanning process, the ultrasound image data is identified using the second classification model to determine a plurality of standard sections under a target section protocol corresponding to the final section type, wherein the target section protocol is a section protocol corresponding to the final section type;

[0021] Outputting the plurality of standard sections so that the user can select an optimal section from the plurality of standard sections;

[0022] The optimal section selected by the user is determined as the ultrasound section under the target section protocol.

[0023] In a specific embodiment of the present application, after outputting the multiple standard sections and before determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method further includes:

[0024] If the user returns a message indicating that there is no suitable section, the step of starting ultrasound scanning according to the section protocol corresponding to the final section type is repeated until it is determined that the user has selected the optimal section from the multiple standard sections.

[0025] In a specific implementation of the present application, the optimal section selected by the user from the multiple standard sections is determined by the following steps:

[0026] Recognize the received voice, gesture or click command of the user;

[0027] The optimal section selected by the user from the multiple standard sections is determined according to the recognition result.

[0028] In a specific embodiment of the present application, after determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method further includes:

[0029] Adding the optimal section to a training library;

[0030] Incrementally train the second classification model based on the training library.

[0031] An ultrasonic scanning device, comprising:

[0032] An initial image acquisition module, used to obtain an initial image of an ultrasound scan;

[0033] An initial section type determination module, used to identify the initial image using a first classification model to determine an initial section type of the initial image, wherein the initial section type is a target part targeted by the current ultrasound scan;

[0034] A feedback information receiving module, used to receive user feedback on the accuracy of the initial section type;

[0035] A final section type determination module, used to determine the final section type of the initial image according to the accuracy feedback information;

[0036] The ultrasonic scanning module is used to start ultrasonic scanning according to the section protocol corresponding to the final section type.

[0037] An ultrasonic scanning device, comprising:

[0038] Ultrasound probe, used to acquire images;

[0039] Memory for storing computer programs;

[0040] A processor is communicatively connected to the ultrasound probe and is used to implement the steps of any one of the above-mentioned ultrasound scanning methods when executing the computer program.

[0041] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the above-mentioned ultrasonic scanning methods.

[0042] By applying the technical solution provided in the embodiment of the present application, after obtaining the initial image of the ultrasound scan, the initial image can be identified using the first classification model to determine the initial section type of the initial image, and the final section type of the initial image can be determined based on the user's feedback information on the accuracy of the initial section type, and the ultrasound scan can be started according to the section protocol corresponding to the final section type. After the initial section type of the initial image is determined by machine recognition, the ultrasound scan is not performed directly according to the section protocol corresponding to the initial section type. Instead, the machine recognition and user feedback are combined to determine the final section type of the initial image. Only in this way can the accuracy of the final section type be guaranteed. Only by starting the ultrasound scan based on the section protocol corresponding to the accurately determined final section type can a more accurate ultrasound section be obtained, which is convenient for subsequent use. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 This is a flowchart of an implementation of the ultrasonic scanning method in the embodiment of the present application;

[0045] Figure 2 Another implementation flow chart of the ultrasonic scanning method in the embodiment of the present application;

[0046] Figure 3 This is a schematic diagram of the structure of an ultrasonic scanning device in an embodiment of the present application;

[0047] Figure 4 This is a schematic diagram of the structure of an ultrasonic scanning device in an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0049] See also Figure 1 FIG. 1 is a flowchart of an ultrasonic scanning method provided in an embodiment of the present application. The method may include the following steps:

[0050] S110: Obtaining an initial image of ultrasound scanning.

[0051] In the embodiment of the present application, the ultrasonic scanning process may include a pre-scan and a formal scan, and the scanning areas of the pre-scan and the formal scan are the same. Before the formal ultrasonic scan is to be performed, an initial image of the pre-scan may be obtained. For example, when the user triggers the ultrasonic scanning operation, the user may be prompted to pre-scan the target area first to obtain an initial image of the ultrasonic scan.

[0052] S120: Using the first classification model to identify the initial image, and determining the initial section type of the initial image.

[0053] The initial section type is the target area of ​​the current ultrasound scan.

[0054] In the embodiment of the present application, the first classification model can be obtained by pre-training through a training library. The training library can contain multiple training data, each of which includes image features and section type labels.

[0055] A first initial classification model can be pre-built, each training data is input into the first initial classification model, and the training data is classified based on its image features to obtain an output result of its corresponding section type, and then the model parameters are adjusted according to the output result and the section type label included in the training data. When the classification accuracy reaches a set first threshold, the first classification model can be obtained.

[0056] After obtaining the initial image of the ultrasound scan, the first classification model can be used to identify the initial image, and the initial section type of the initial image can be determined according to the output result of the first classification model. The initial section type of the initial image is the target part targeted by the current ultrasound scan, and the target part can be the gallbladder, kidney, liver, etc. This is an initial judgment of the actual target part through machine recognition.

[0057] S130: Receive user feedback on the accuracy of the initial section type.

[0058] After obtaining an initial image of the ultrasound scan, the initial image is identified using the first classification model, and the initial section type of the initial image is determined, the initial section type can be output and displayed to the user. Because the initial section type is automatically identified by the machine, there may be deviations. If there are deviations, then the ultrasound section obtained by directly performing ultrasound scanning according to the section protocol corresponding to the initial section type will inevitably have deviations. Therefore, after the initial section type of the initial image is determined by automatic machine identification, the application outputs the initial section type and displays it to the user, so that the user can provide feedback on the accuracy of the initial section type.

[0059] In actual applications, the user knows which part to be scanned with ultrasound. After the initial section type of the initial image determined by the first classification model is output to the user, the user can determine the accuracy of the initial section type according to actual conditions and provide feedback, thereby obtaining accurate feedback information.

[0060] Accuracy feedback information can be information provided by the user through voice, gesture or input box, and can include information on whether the initial section type is accurate and information on the section type given when the user believes it is inaccurate. That is, after the initial section type is output and displayed, the user can answer "correct" or "wrong" through voice, or make correct or wrong gesture operations through gestures, or enter correct or wrong text information in the input box provided to the user, or select the correct or wrong button from the buttons provided to the user. At the same time, if the user believes that the initial section type is incorrectly identified, the user can also provide feedback on the correction result, that is, the correct section type that the user believes is correct, through voice, gesture or input box.

[0061] The following describes the communication using voice and gesture as examples.

[0062] By voice:

[0063] The built-in voice unit can be used to communicate with the user by voice, recognize the voice input by the user, and determine the specific feedback information. If the voice input by the user is recognized as "correct" or "yes", it can be determined that the initial section type fed back by the user is accurate. If the voice input by the user is recognized as "wrong" or "no", it can be determined that the initial section type fed back by the user is inaccurate. The user can input the voice of the correction result at the same time, and the correction result can be obtained after it is recognized. Alternatively, if it is determined that the initial section type fed back by the user is inaccurate, the user is prompted to input the accurate section type, and then the voice information input by the user is recognized to obtain the correction result.

[0064] By gesture:

[0065] The user's gesture operation can be collected by the image acquisition unit, the gesture operation can be identified, and the specific feedback information can be determined. For example, when the user nods multiple times, waves vertically up and down, and other gestures are collected, it is determined that the initial section type of the user feedback is accurate. For example, when the user shakes his head, shakes his hands left and right, and other gestures are collected, it is determined that the initial section type of the user feedback is inaccurate. In the case where it is determined that the initial section type of the user feedback is inaccurate, other section type names can continue to be output, and the user's feedback information can still be determined through gesture recognition until it is consistent with the user's cognition. The correspondence between gesture operations and gesture meanings can be pre-set and made known to the user.

[0066] The voice mode and the gesture mode can be used alone or in combination, and can also be used in combination with text input and other modes.

[0067] Providing information feedback through voice or gestures makes it more convenient for users to operate and can save interaction time.

[0068] After receiving the user's feedback information on the accuracy of the initial section type, the operation of step S140 may be continued.

[0069] S140: Determine the final section type of the initial image according to the accuracy feedback information.

[0070] After receiving the accuracy feedback information of the user on the initial section type, the final section type of the initial image can be determined according to the accuracy feedback information. That is, the final section type of the initial image can be determined by combining machine recognition and user feedback. In this way, the section type of the initial image can be determined more accurately.

[0071] S150: Start ultrasound scanning according to the section protocol corresponding to the final section type.

[0072] After determining the final section type of the initial image, the ultrasound scan can be started according to the section protocol corresponding to the final section type to perform the formal scanning process. During the formal scanning process, the ultrasound section can be obtained.

[0073] By applying the method provided in the embodiment of the present application, after obtaining the initial image of the ultrasound scan, the initial image can be identified using the first classification model to determine the initial section type of the initial image, and the final section type of the initial image can be determined based on the user's feedback information on the accuracy of the initial section type, and the ultrasound scan can be started according to the section protocol corresponding to the final section type. After the initial section type of the initial image is determined by machine recognition, the ultrasound scan is not performed directly according to the section protocol corresponding to the initial section type. Instead, the machine recognition and user feedback are combined to determine the final section type of the initial image. Only in this way can the accuracy of the final section type be guaranteed. Only by starting the ultrasound scan based on the section protocol corresponding to the accurately determined final section type can a more accurate ultrasound section be obtained, which is convenient for subsequent use.

[0074] In one embodiment of the present application, step S140 may include the following steps:

[0075] Step 1: Determine whether the initial section type is accurate based on the accuracy feedback information; if it is accurate, proceed to step 2; if it is inaccurate, proceed to step 3;

[0076] Step 2: Determine the initial section type as the final section type of the initial image;

[0077] Step 3: Obtain the correction result in the accuracy feedback information, and determine the section type corresponding to the correction result as the final section type of the initial image.

[0078] For ease of description, the above three steps are combined for explanation.

[0079] After obtaining an initial image of an ultrasound scan, the initial image is identified using a first classification model to determine an initial section type of the initial image, and after receiving feedback information from the user on the accuracy of the initial section type, it is possible to determine whether the initial section type is accurate based on the accuracy feedback information received from the user.

[0080] In actual applications, after the initial section type is output and displayed, the user can return feedback information that the initial section type is accurate based on the initial section type displayed, if the user determines that it is accurate, indicating that the recognition result of the first classification model is consistent with the user's cognition, and the initial section type can be directly determined as the final section type of the initial image. In the case where the user determines that the initial section type is inaccurate, feedback information that the initial section type is inaccurate can be returned, and the correction result can be carried in the feedback information, so that after receiving the accuracy feedback information, the correction result carried therein can be obtained, and the section type corresponding to the correction result can be determined as the final section type of the initial image, so that the final section type finally determined is consistent with the user's cognition.

[0081] By combining machine recognition with user feedback, the final section type targeted by the current ultrasound scan can be accurately determined, laying the foundation for the subsequent accuracy determination of the ultrasound section.

[0082] In one embodiment of the present application, when it is determined that the initial section type is inaccurate, after obtaining the correction result in the accuracy feedback information, the model parameters of the first classification model can also be adjusted based on the correction result so that the adjusted first classification model matches the correction result.

[0083] It is understandable that if the user determines that the initial section type is inaccurate, it can be considered that the current recognition result of the first classification model is biased. The accuracy feedback information returned by the user can carry a correction result, and the model parameters of the first classification model are adjusted based on the correction result, so that the recognition result obtained after the adjusted first classification model recognizes the initial image can match the correction result. This can improve the recognition accuracy of the first classification model and facilitate subsequent use.

[0084] After the first classification model completes the parameter adjustment, it can provide feedback to the user through voice or interface display.

[0085] In one embodiment of the present application, after starting the ultrasound scan according to the slice protocol corresponding to the final slice type, the method may further include the following steps:

[0086] The first step: during the ultrasound scanning process, the ultrasound image data is identified using the second classification model to determine a plurality of standard sections under a target section protocol corresponding to the final section type, where the target section protocol is a section protocol corresponding to the final section type;

[0087] The second step: output multiple standard sections so that the user can select the optimal section from the multiple standard sections;

[0088] The third step: determining the optimal section selected by the user as the ultrasound section under the target section protocol.

[0089] For ease of description, the above three steps are combined for explanation.

[0090] After the final section type of the initial image is accurately determined, ultrasound scanning can be started according to the section protocol corresponding to the final section type.

[0091] During the ultrasound scanning process, ultrasound image data can be obtained. The second classification model can be used to identify the obtained ultrasound image data and determine multiple standard sections under the target section protocol corresponding to the final section type. The target section protocol is a section protocol corresponding to the final section type, which can be any one of them.

[0092] For a section type, one or more section protocols can be set, and the section protocols of different section types are different. If the final section type has only one section protocol, then the section protocol is the target section protocol corresponding to the final section type. If the final section type has multiple section protocols, each section protocol can be determined as the target section protocol in turn, and multiple standard sections under the target section protocol can be determined one by one.

[0093] The second classification model can be obtained by pre-training based on a training library, and the training library can include multiple training data, each training data includes image features and standard section labels.

[0094] A second initial classification model can be constructed in advance, and each training data can be input into the second initial classification model. The training data can be classified based on its image features to obtain an output result of whether it is a standard section. The model parameters can then be adjusted based on the output result and the standard section label included in the training data. When the classification accuracy reaches a set second threshold, the second classification model can be obtained.

[0095] After the second classification model is used to identify the ultrasound image data and multiple standard sections under the target section protocol corresponding to the final section type are determined, the multiple standard sections can be output and displayed to the user. The user can select the optimal section from the multiple standard sections according to actual conditions.

[0096] In actual applications, after determining multiple standard slices under the target slice protocol, the multiple standard slices can be directly output and displayed to the user on the same page, and the user can select the optimal slice from them, or the multiple standard slices can be output and displayed to the user on multiple pages, and the user can select the optimal slice by flipping through them.

[0097] After the user selects the optimal section from multiple standard sections, the selection result fed back by the user can be obtained, and the optimal section selected by the user is determined as the ultrasound section under the target section protocol.

[0098] In practical applications, if the final section type has multiple section protocols, each section protocol is determined as the target section protocol one by one. Through the above steps, multiple standard sections under each section protocol corresponding to the final section type can be determined respectively, thereby obtaining the optimal ultrasound section corresponding to each section protocol.

[0099] The embodiment of the present application first determines the final section type targeted by the current ultrasound scan, then starts the ultrasound scan according to the section protocol corresponding to the final section type, and then during the ultrasound scan, uses the second classification model to identify the ultrasound image data. After determining multiple standard sections under the target section protocol, the final ultrasound section is not directly randomly selected from the multiple standard sections. Instead, the determined multiple standard sections are output to the user, and the user selects the optimal section from the multiple standard sections, thereby increasing the interaction with the user. Through interaction with the user, the recognition accuracy of the ultrasound section is further enhanced, which is convenient for subsequent use.

[0100] It should be noted that the second classification model and the first classification model can share the same training library, or use different training libraries respectively.

[0101] In one embodiment of the present application, after outputting a plurality of standard sections and before determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method further includes the following steps:

[0102] If the user returns a message indicating that there is no suitable section, the step of starting the ultrasound scan according to the section protocol corresponding to the final section type is repeated until it is determined that the user has selected the optimal section from multiple standard sections.

[0103] In the embodiment of the present application, during the ultrasound scanning process, the ultrasound image data is identified using the second classification model, and after determining multiple standard sections under the target section protocol corresponding to the final section type, the multiple standard sections are output to the user. The user can select the optimal section from the multiple standard sections. If the user believes that there is no suitable section to choose from among the multiple standard sections, information that there is no suitable section can be returned.

[0104] In the case of receiving information from the user that there is no suitable section, the step of starting the ultrasound scan according to the section protocol corresponding to the final section type can be repeated to obtain more ultrasound image data, and then the ultrasound image data is identified using the second classification model to determine and output multiple standard sections under the target section protocol. Multiple standard sections are then output to the user. If the user still receives information that there is no suitable section, the step of starting the ultrasound scan according to the section protocol corresponding to the final section type can be repeated until it is determined that the user has selected the optimal section from the multiple standard sections. The optimal section selected by the user is determined as the ultrasound section corresponding to the target section protocol.

[0105] Through the combination of automatic machine recognition and user feedback, the ultrasound section obtained under the target section protocol will be more accurate and more convenient for subsequent use.

[0106] In one embodiment of the present application, the optimal section selected by the user from multiple standard sections can be determined by the following steps:

[0107] The first step: recognize the user's voice, gesture or click command received;

[0108] The second step: based on the recognition results, determine the optimal section selected by the user from multiple standard sections.

[0109] For ease of description, the above two steps are combined for illustration.

[0110] In an embodiment of the present application, after a plurality of standard slices under the target slice protocol are determined and output to the user, the user may provide feedback through voice, gesture, or click.

[0111] In actual applications, when multiple standard sections are output, each standard section displayed can be numbered, and the user can input the number of the optimal section selected by voice, and the optimal section selected by the user from multiple standard sections can be determined by voice recognition. Alternatively, each standard section can be output one by one, and the user can input whether the currently displayed standard section is the optimal section one by one through voice or gesture. Alternatively, it can be determined whether the currently displayed standard section is the optimal section by clicking the recognition button.

[0112] Communicate with users through voice, gestures, or click commands to facilitate user operation.

[0113] In one embodiment of the present application, after determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method may further include the following steps:

[0114] Step 1: Add the optimal slice to the training library;

[0115] Step 2: Incrementally train the second classification model based on the training library.

[0116] For ease of description, the above two steps are combined for illustration.

[0117] In the embodiment of the present application, after the optimal section selected by the user is determined as the ultrasound section under the target section protocol, the optimal section can be added to the training library, the training library is updated, and then the second classification model is incrementally trained based on the updated training library. This facilitates this operation and subsequent similar application scenarios, and improves the recognition accuracy of the ultrasound section.

[0118] For ease of understanding, Figure 2 The embodiment of the present application is described by taking the specific interaction process between the user and the ultrasonic scanning device as an example, and the ultrasonic scanning device is used to perform the operations in the above steps.

[0119] When an ultrasonic scan is to be performed, the ultrasonic scan device can obtain an initial image of the ultrasonic scan;

[0120] The ultrasonic scanning device uses the first classification model to identify the initial image and determine the initial section type of the initial image;

[0121] The ultrasonic scanning device outputs the initial section type, and the user provides feedback on the accuracy of the initial section type;

[0122] The ultrasonic scanning device determines the final section type of the initial image according to the accuracy feedback information received from the user, and starts the ultrasonic scanning according to the section protocol corresponding to the final section type;

[0123] During the ultrasound scanning process, the ultrasound scanning device uses the second classification model to identify the ultrasound image data and determine a plurality of standard sections under the target section protocol corresponding to the final section type;

[0124] The ultrasonic scanning device outputs multiple standard sections, and the user selects the optimal section from the multiple standard sections;

[0125] The ultrasound scanning device determines whether the user selects the optimal section from the multiple standard sections. If yes, the ultrasound scanning device determines the optimal section selected by the user as the ultrasound section under the target section protocol. If no, the ultrasound scanning device repeats the step of starting ultrasound scanning according to the section protocol corresponding to the final section type.

[0126] The ultrasound scanning device adds the optimal section to the training library and performs incremental training on the second classification model based on the training library.

[0127] In related technologies, each classification model is obtained by training a single, static image. In actual use, it is easy to have low recognition accuracy for data in non-training libraries and overfitting of data in training libraries. In the process of using the classification model, the present application expands the training library and performs online continuous learning and incremental learning on the classification model, which can improve the recognition accuracy and solve problems such as overfitting of data in the training library. At the same time, the present application introduces context and communication feedback mechanisms to limit the recognition scope of the classification model, which can solve problems such as recognition errors caused by similarities of individual static images in different diagnostic parts.

[0128] Corresponding to the above method embodiment, the embodiment of the present application further provides an ultrasonic scanning device. The ultrasonic scanning device described below and the ultrasonic scanning method described above can refer to each other.

[0129] See also Figure 3 As shown, the device may include the following modules:

[0130] An initial image acquisition module 310 is used to obtain an initial image of an ultrasound scan;

[0131] An initial section type determination module 320 is used to identify the initial image using the first classification model to determine the initial section type of the initial image, where the initial section type is the target part targeted by the current ultrasound scan;

[0132] A feedback information receiving module 330 is used to receive user feedback on the accuracy of the initial section type;

[0133] A final section type determination module 340 is used to determine the final section type of the initial image according to the accuracy feedback information;

[0134] The ultrasound scanning module 350 is used to start ultrasound scanning according to the section protocol corresponding to the final section type.

[0135] By using the device provided in the embodiment of the present application, after obtaining the initial image of the ultrasound scan, the initial image can be identified using the first classification model to determine the initial section type of the initial image, and the final section type of the initial image can be determined based on the user's feedback information on the accuracy of the initial section type, and the ultrasound scan can be started according to the section protocol corresponding to the final section type. After the initial section type of the initial image is determined by machine recognition, the ultrasound scan is not performed directly according to the section protocol corresponding to the initial section type. Instead, the machine recognition and user feedback are combined to determine the final section type of the initial image. Only in this way can the accuracy of the final section type be guaranteed. Only by starting the ultrasound scan based on the section protocol corresponding to the accurately determined final section type can a more accurate ultrasound section be obtained, which is convenient for subsequent use.

[0136] In a specific implementation of the present application, the final section type determination module 340 is used to:

[0137] Determine whether the initial section type is accurate based on the accuracy feedback information;

[0138] If accurate, the initial slice type is determined as the final slice type of the initial image;

[0139] If it is inaccurate, a correction result in the accuracy feedback information is obtained, and the section type corresponding to the correction result is determined as the final section type of the initial image.

[0140] In a specific implementation of the present application, a model parameter adjustment module is also included, which is used to:

[0141] In the case where it is determined that the initial section type is inaccurate, after obtaining the correction result in the accuracy feedback information, the model parameters of the first classification model are adjusted based on the correction result so that the adjusted first classification model matches the correction result.

[0142] In a specific implementation of the present application, the accuracy feedback information is information provided by the user through voice, gesture or input box.

[0143] In a specific embodiment of the present application, an ultrasound section determination module is also included, which is used to:

[0144] After starting the ultrasound scan according to the slice protocol corresponding to the final slice type, during the ultrasound scan, the ultrasound image data is identified using the second classification model to determine a plurality of standard slices under the target slice protocol corresponding to the final slice type, where the target slice protocol is the slice protocol corresponding to the final slice type;

[0145] Output multiple standard sections so that users can select the best section from multiple standard sections;

[0146] The optimal section selected by the user is determined as the ultrasound section under the target section protocol.

[0147] In a specific implementation of the present application, the ultrasound section determination module is further used to:

[0148] After outputting a plurality of standard slices and before determining the optimal slice selected by the user as the ultrasound slice under the target slice protocol, the method further includes:

[0149] If the user returns a message indicating that there is no suitable section, the step of starting the ultrasound scan according to the section protocol corresponding to the final section type is repeated until it is determined that the user has selected the optimal section from multiple standard sections.

[0150] In a specific implementation of the present application, the ultrasound section determination module is further used to determine the optimal section selected by the user from multiple standard sections through the following steps:

[0151] Recognize the user's voice, gesture or click command received;

[0152] According to the recognition result, the optimal section selected by the user from multiple standard sections is determined.

[0153] In a specific implementation of the present application, a model incremental training module is also included, which is used to:

[0154] After determining the optimal section selected by the user as the ultrasound section under the target section protocol, adding the optimal section to the training library;

[0155] The second classification model is incrementally trained based on the training library.

[0156] Corresponding to the above method embodiment, the present application embodiment further provides an ultrasonic scanning device, including:

[0157] Ultrasound probe, used to acquire images;

[0158] Memory for storing computer programs;

[0159] The processor is connected to the ultrasound probe for implementing the steps of the ultrasound scanning method when executing the computer program.

[0160] like Figure 4 , which is a schematic diagram of the composition structure of an ultrasonic scanning device, the ultrasonic scanning device may include: a processor 10, a memory 11, a communication interface 12, a communication bus 13 and an ultrasonic probe 14. The processor 10, the memory 11, the communication interface 12 and the ultrasonic probe 14 all communicate with each other through the communication bus 13.

[0161] In the embodiment of the present application, the processor 10 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic devices, etc.

[0162] The processor 10 may call a program stored in the memory 11. Specifically, the processor 10 may execute operations in an embodiment of the method for obtaining an ultrasonic section.

[0163] The memory 11 is used to store one or more programs, which may include program codes, and the program codes include computer operation instructions. In the embodiment of the present application, the memory 11 at least stores programs for implementing the following functions:

[0164] Obtaining initial ultrasound scan images;

[0165] Using the first classification model to identify the initial image, determine the initial section type of the initial image, the initial section type is the target part targeted by the current ultrasound scan;

[0166] Receive user feedback on the accuracy of the initial slice type;

[0167] Determine the final section type of the initial image based on the accuracy feedback information;

[0168] Start ultrasound scanning according to the section protocol corresponding to the final section type.

[0169] In one possible implementation, the memory 11 may include a program storage area and a data storage area, wherein the program storage area can store an operating system and application programs required for at least one function (such as voice recognition function, image recognition function), etc.; the data storage area can store data created during use, such as ultrasound image data, interaction data, etc.

[0170] In addition, the memory 11 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0171] The communication interface 13 may be an interface of a communication module, and is used to connect to other devices or systems.

[0172] Of course, it should be noted that Figure 4 The structure shown does not constitute a limitation on the ultrasonic scanning device in the embodiment of the present application. In actual applications, the ultrasonic scanning device may include Figure 4 More or fewer components than shown, or combinations of certain components.

[0173] Corresponding to the above method embodiment, the embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned ultrasonic scanning method are implemented.

[0174] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0175] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0176] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0177] Specific examples are used herein to illustrate the principles and implementation methods of the present application, and the description of the above embodiments is only used to help understand the technical solution and core ideas of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. An ultrasonic scanning method, characterized in that: include: Obtaining initial ultrasound pre-scan images; Using a first classification model to identify the initial image, determine an initial section type of the initial image, and the initial section type is a target part targeted by the current ultrasound scan; Receiving user feedback on the accuracy of the initial section type; Determining a final section type of the initial image according to the accuracy feedback information; Starting the ultrasound formal scan according to the section protocol corresponding to the final section type; the scanning area of ​​the pre-scan is consistent with that of the formal scan; Wherein, after starting the ultrasound formal scanning according to the section protocol corresponding to the final section type, it also includes: During the ultrasound formal scanning process, the ultrasound image data is identified using the second classification model to determine a plurality of standard sections under the target section protocol corresponding to the final section type, wherein the target section protocol is the section protocol corresponding to the final section type; Outputting the plurality of standard sections so that the user can select an optimal section from the plurality of standard sections; The optimal section selected by the user is determined as the ultrasound section under the target section protocol.

2. The method according to claim 1, characterized in that The determining, according to the accuracy feedback information, the final section type of the initial image comprises: Determining whether the initial section type is accurate according to the accuracy feedback information; If accurate, determining the initial section type as the final section type of the initial image; If it is inaccurate, a correction result in the accuracy feedback information is obtained, and the section type corresponding to the correction result is determined as the final section type of the initial image.

3. The method according to claim 2, characterized in that In the case where it is determined that the initial section type is inaccurate, after obtaining the correction result in the accuracy feedback information, the method further includes: The model parameters of the first classification model are adjusted based on the correction result so that the adjusted first classification model matches the correction result.

4. The method according to claim 1, characterized in that The accuracy feedback information is information fed back by the user through voice, gesture or input box.

5. The method according to claim 1, characterized in that After outputting the plurality of standard sections and before determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method further includes: If the user returns a message indicating that there is no suitable section, the step of starting ultrasound scanning according to the section protocol corresponding to the final section type is repeated until it is determined that the user has selected the optimal section from the multiple standard sections.

6. The method according to claim 1, characterized in that The optimal section selected by the user from the multiple standard sections is determined by the following steps: Recognize the received voice, gesture or click command of the user; The optimal section selected by the user from the multiple standard sections is determined according to the recognition result.

7. The method according to any one of claims 1 to 6, characterized in that After determining the optimal section selected by the user as the ultrasound section under the target section protocol, the method further includes: Adding the optimal section to a training library; Incrementally train the second classification model based on the training library.

8. An ultrasonic scanning device, characterized in that: include: An initial image acquisition module, used to obtain an initial image of an ultrasound pre-scan; An initial section type determination module, used to identify the initial image using a first classification model to determine an initial section type of the initial image, wherein the initial section type is a target part targeted by the current ultrasound scan; A feedback information receiving module, used to receive user feedback on the accuracy of the initial section type; A final section type determination module, used to determine the final section type of the initial image according to the accuracy feedback information; An ultrasound scanning module, used to start a formal ultrasound scan according to a section protocol corresponding to the final section type; the scanning area of ​​the pre-scan is consistent with that of the formal scan; The ultrasonic scanning device is further used to: during the ultrasonic formal scanning process, use the second classification model to identify the ultrasonic image data, and determine a plurality of standard sections under the target section protocol corresponding to the final section type, wherein the target section protocol is the section protocol corresponding to the final section type; The multiple standard sections are outputted so that the user can select an optimal section from the multiple standard sections; and the optimal section selected by the user is determined as the ultrasound section under the target section protocol.

9. An ultrasonic scanning device, characterized in that: include: Ultrasound probe, used to acquire images; Memory for storing computer programs; A processor is communicatively connected to the ultrasound probe, and is used to implement the steps of the ultrasound scanning method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the ultrasonic scanning method according to any one of claims 1 to 7 are implemented.