CN115564763A - Thyroid ultrasound image processing method, device, medium and electronic equipment - Google Patents
Thyroid ultrasound image processing method, device, medium and electronic equipment Download PDFInfo
- Publication number
- CN115564763A CN115564763A CN202211358249.0A CN202211358249A CN115564763A CN 115564763 A CN115564763 A CN 115564763A CN 202211358249 A CN202211358249 A CN 202211358249A CN 115564763 A CN115564763 A CN 115564763A
- Authority
- CN
- China
- Prior art keywords
- thyroid
- ultrasound image
- nodule
- image
- processing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10132—Ultrasound image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Multimedia (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Quality & Reliability (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The application provides a thyroid ultrasound image processing method, a thyroid ultrasound image processing device, a thyroid ultrasound image processing medium and electronic equipment. The method comprises the following steps: obtaining an ultrasonic thyroid image; performing first processing on the thyroid ultrasound image to acquire the position of a thyroid nodule; predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule; performing second processing on the thyroid ultrasound image to acquire an identification of a thyroid nodule; and fusing detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, wherein the detection information corresponding to the thyroid ultrasound image comprises the predicted position and the thyroid nodule identifier. The thyroid ultrasonic image processing method is beneficial to improving the accuracy of thyroid nodule detection.
Description
Technical Field
The application belongs to the technical field of medical image processing, relates to an ultrasonic image processing method, and particularly relates to a thyroid ultrasonic image processing method, device, medium and electronic equipment.
Background
The medical ultrasonic examination is a medical imaging diagnosis technology based on ultrasonic waves, and the size, the structure, the pathological focus and the like of muscles and internal organs can be visualized through the ultrasonic medical examination, so that a basis is provided for diagnosis and treatment work of medical staff. The thyroid gland is the largest endocrine gland of a human body, and the size, volume and blood flow of the thyroid gland can be qualitatively and quantitatively estimated through ultrasonic examination, so that qualitative or semi-quantitative diagnosis can be performed on benign and malignant tumors, and the ultrasonic examination becomes a preferred method for image examination of thyroid diseases.
In recent years, with the continuous development of medical artificial intelligence technology, the nodule detection can be performed on the ultrasonic image output by the ultrasonic machine by means of the scientific and technological means of machine learning and deep learning algorithm modeling. For example, the information of thyroid nodules can be automatically identified and detected by an algorithm by means of machine learning and deep learning technologies and a big data modeling method. However, the ultrasound image output by the ultrasound machine is a video stream image, and often only depends on the image information of the current frame in the process of performing traditional deep learning detection, segmentation and classification, and the inter-frame historical information is not correlated, which results in that the thyroid nodule detection result obtained by the prior art is not high in accuracy.
Disclosure of Invention
An object of the application is to provide a thyroid ultrasound image processing method, device, medium and electronic device, which are used for solving the problem that the existing thyroid nodule detection result is not high in accuracy.
In a first aspect, the present application provides a thyroid ultrasound image processing method, including: obtaining a thyroid gland ultrasonic image; performing first processing on the thyroid ultrasonic image to acquire the position of a thyroid nodule; predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule; performing second processing on the thyroid ultrasound image to acquire an identification of a thyroid nodule; and fusing detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, wherein the detection information corresponding to the thyroid ultrasound image comprises the predicted position and the thyroid nodule identification.
In one implementation of the first aspect, the first processing the thyroid ultrasound image to obtain the location of the thyroid nodule comprises: detecting the thyroid ultrasound image by using a deep learning target detection method to obtain an ultrasound image area; and acquiring the position of the thyroid nodule by using a two-level target detection method according to the image in the ultrasonic image area.
In one implementation of the first aspect, the thyroid ultrasound image includes at least one thyroid nodule, each thyroid nodule corresponds to one kalman tracker, and predicting a thyroid nodule position in a next frame ultrasound image as the predicted position includes: processing the position of the thyroid nodule and the historical position of the thyroid nodule with the kalman tracker to predict a thyroid nodule position in a next frame of ultrasound image as the predicted position.
In one implementation manner of the first aspect, the thyroid ultrasound image processing method further includes: updating and maintaining each of the Kalman trackers based on the location of the thyroid nodule.
In one implementation manner of the first aspect, the thyroid ultrasound image processing method further includes: initializing the Kalman tracker.
In one implementation of the first aspect, the second processing of the thyroid ultrasound image to obtain an identification of a thyroid nodule comprises: and processing the ultrasonic video stream containing the thyroid ultrasonic image by utilizing a nodule heavy identification model to acquire the identification of the thyroid nodule, wherein the nodule heavy identification model is a trained neural network model.
In one implementation manner of the first aspect, fusing the detection information corresponding to the thyroid ultrasound image and the historical detection information includes: filtering detection information corresponding to the thyroid ultrasound image and historical detection information; and fusing the detection information corresponding to the thyroid ultrasonic image and the historical detection information to generate a structural detection result of the thyroid nodule.
In a second aspect, the present application provides a thyroid ultrasound image processing apparatus, including: the image acquisition module is used for acquiring a thyroid ultrasound image; the first processing module is used for carrying out first processing on the thyroid ultrasonic image so as to acquire the position of a thyroid nodule; the prediction module is used for predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule; the second processing module is used for carrying out second processing on the thyroid ultrasonic image so as to acquire the identification of the thyroid nodule; and the information fusion module is used for fusing detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, wherein the detection information corresponding to the thyroid ultrasound image comprises the predicted position and the thyroid nodule identifier.
In a third aspect, the present application provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the thyroid ultrasound image processing method according to any one of the first aspect of the present application.
In a fourth aspect, the present application provides an electronic device, comprising: a memory storing a computer program; and the processor is in communication connection with the memory and executes the thyroid ultrasound image processing method according to any one of the first aspects of the application when the computer program is called.
As described above, according to the thyroid ultrasound image processing method provided in the implementation manner of the present application, the detection result of the thyroid ultrasound image is obtained according to the detection information corresponding to the thyroid ultrasound image and the historical detection information. The method fully considers the correlation of the inter-frame historical information, and is favorable for improving the accuracy of thyroid nodule detection.
Drawings
Fig. 1 is a diagram illustrating an application scenario of the method for processing an ultrasound image of a thyroid gland according to an embodiment of the present application.
Fig. 2 is a flowchart illustrating a method for processing an ultrasound image of a thyroid gland according to an embodiment of the present application.
FIG. 3A is a flow chart showing the acquisition of the location of a thyroid nodule in an embodiment of the present application.
Fig. 3B and 3C are diagrams showing examples of ultrasound images of thyroid gland in embodiments of the present application.
Fig. 4 is a flowchart illustrating training of a nodule re-identification model in an embodiment of the present application.
Fig. 5 is a flow chart showing the generation of a structured detection result of thyroid nodules in an embodiment of the present application.
Fig. 6 is a flowchart illustrating real-time detection of a thyroid ultrasound image according to an embodiment of the present application.
Fig. 7 is a schematic structural diagram of a thyroid ultrasound image processing apparatus according to an embodiment of the present application.
Fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Description of the element reference numerals
1. Image processing system
11. Ultrasonic image acquisition device
12. General purpose processor
13. Memory device
14. Display device
700. Thyroid ultrasonic image processing device
710. Image acquisition module
720. First processing module
730. Prediction module
740. Second processing module
750. Information fusion module
800. Electronic device
810. Memory device
820. Processor with a memory having a plurality of memory cells
830. Display device
S21 to S25
S31-S32
S41 to S45
S51 to S52 steps
S61 to S67
Detailed Description
The following description of the embodiments of the present application is provided by way of specific examples, and other advantages and effects of the present application will be readily apparent to those skilled in the art from the disclosure herein. The present application is capable of other and different embodiments and its several details are capable of modifications and/or changes in various respects, all without departing from the spirit of the present application. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present application, and the drawings only show the components related to the present application and are not drawn according to the number, shape and size of the components in actual implementation, and the type, number and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
At present, most of thyroid nodule detection, classification and measurement depend on doctor experience for observation, and nodules appearing in thyroid glands of patients are calibrated and measured manually according to experience, and structured reports are output. In addition, some schemes adopt a machine learning and deep learning method, and an algorithm is enabled to automatically identify and detect thyroid nodules through big data modeling. However, during the real-time ultrasound scanning process, the position of the probe on the hand of the operator is constantly changed, so that the thyroid nodules scanned are different in representation form in different frames of images. However, the existing processes of machine learning, deep learning, detection, segmentation and classification only depend on the image information of the current frame, and do not correlate the historical information between frames, which results in that the thyroid nodule detection result obtained by the prior art is not high in accuracy.
At least in view of the above problems, the present application provides a thyroid ultrasound image processing method. The method obtains the detection result of the thyroid ultrasound image according to the detection information corresponding to the thyroid ultrasound image and the historical detection information. The method fully considers the correlation of the inter-frame historical information, and is favorable for improving the accuracy of thyroid nodule detection.
Fig. 1 is a diagram illustrating an application scenario of a thyroid ultrasound image processing method according to an embodiment of the present application. As shown in fig. 1, the thyroid ultrasound image processing method provided in the embodiment of the present application is applied to an image processing system 1. The image processing system 1 includes an ultrasound image acquisition device 11, a general-purpose processor 12, a memory 13, and a display device 14.
The ultrasonic image acquiring apparatus 11 is an apparatus for diagnosing diseases by displaying a tomographic image of internal organs or lesions of a human body by using physical characteristics of ultrasonic waves propagated in the human body, and includes an ultrasonic probe and the like.
The general purpose processor 12 may be any type of device capable of Processing electronic instructions, and the image Processing system 1 may include one or more general purpose processors 13, including, for example, one or both of a Central Processing Unit (CPU) and a Neural-Network Processing Unit (NPU).
The Memory 13 may include a Volatile Memory (Volatile Memory), such as a Random Access Memory (RAM), a cache Memory, and a Non-Volatile Memory (Non-Volatile Memory), such as a Read-Only Memory (ROM), a Flash Memory (Flash Memory), a Hard Disk Drive (Hard Disk Drive, HDD), or a Solid-State Drive (SSD).
A display device 14 for displaying information that requires user manipulation or is provided to a user, and various graphical user interfaces in the image processing system 1, which may be composed of graphics, text, icons, video, and any combination thereof.
The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application.
Fig. 2 is a flowchart illustrating a thyroid ultrasound image processing method according to an embodiment of the present application. As shown in fig. 2, the thyroid ultrasound image processing method in the embodiment of the present application includes the following steps S21 to S25.
And S21, obtaining a thyroid ultrasound image. The thyroid ultrasound image can be obtained, for example, by the ultrasound image acquisition apparatus 1 in fig. 1.
And S22, carrying out first processing on the thyroid ultrasonic image to acquire the position of a thyroid nodule.
And S23, predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule. Wherein, the historical position of the thyroid nodule can be obtained by a plurality of frames of ultrasonic images before the thyroid ultrasonic image.
And S24, carrying out second processing on the thyroid ultrasonic image to acquire an identifier of the thyroid nodule, wherein the identifier is used for distinguishing different thyroid nodules.
And S25, fusing detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, wherein the detection information corresponding to the thyroid ultrasound image comprises a predicted position and an identification of a thyroid nodule.
According to the above description, the thyroid ultrasound image processing method provided in the embodiment of the present application can obtain the detection result of the thyroid ultrasound image according to the detection information corresponding to the thyroid ultrasound image and the historical detection information. The method fully considers the correlation of the inter-frame historical information, and is beneficial to improving the accuracy of thyroid nodule detection.
Fig. 3A is a flowchart illustrating a first processing performed on a thyroid ultrasound image to obtain a position of a thyroid nodule according to an embodiment of the present application. As shown in fig. 3A, acquiring the position of the thyroid nodule in the embodiment of the present application includes the following steps S31 and S32.
And S31, detecting the thyroid ultrasound image by using a deep learning target detection method to obtain an ultrasound image area. Specifically, as shown in fig. 3B, the thyroid ultrasound image includes many other information, such as patient information, ultrasound detection time, and the like, in addition to the ultrasound image area. In order to improve the processing efficiency and the processing accuracy, in step S31, the ultrasound image region in the thyroid ultrasound image may be detected by using a deep learning object detection method.
And S32, acquiring the position of the thyroid nodule by using a two-level target detection method according to the image in the ultrasonic image area. Fig. 3C is a diagram illustrating an example of a thyroid nodule region acquired in an embodiment of the present application. The thyroid nodule region is represented by a rectangular box, and the position of the rectangular box is the position of the thyroid nodule.
It should be noted that the target detection method and the secondary target detection method adopted in the embodiments of the present application may be implemented by a model such as a neural network, but the present application is not limited thereto.
In one embodiment of the present application, the thyroid ultrasound image includes at least one thyroid nodule, and each thyroid nodule corresponds to one kalman tracker. In the embodiment of the present application, predicting the position of the thyroid nodule in the next frame of ultrasound image as the predicted position includes: the position of the thyroid nodule and the historical position of the thyroid nodule are processed using a kalman tracker to predict the thyroid nodule position in the next frame of ultrasound image as a predicted position. The Kalman tracker is realized based on a Kalman filter and is used for tracking the position of a thyroid nodule in an ultrasonic image. The kalman filter is a highly efficient recursive filter (autoregressive filter) that can estimate the state of a dynamic system from a series of incomplete and noisy measurements.
Optionally, the thyroid ultrasound image processing method provided in this embodiment of the present application may further include: each kalman tracker is updated and maintained according to the location of the thyroid nodule. Specifically, for any thyroid nodule a, its corresponding kalman tracker is a. In step S22, the position of the thyroid nodule a in the thyroid ultrasound image can be obtained, and the kalman tracker a is maintained and updated based on the position, so as to improve the accuracy of the model processing result.
Optionally, the thyroid ultrasound image processing method provided in this embodiment of the present application may further include: the kalman tracker is initialized. Specifically, for any thyroid nodule B, when the thyroid nodule B is first detected in a frame of ultrasound images, a kalman tracker B is configured and initialized for the thyroid nodule B. In the subsequently acquired ultrasound images, the kalman tracker B only needs to be maintained and updated according to the position of the thyroid nodule B in each frame of ultrasound image.
In an embodiment of the present application, the second processing the thyroid ultrasound image to obtain an identifier of the thyroid nodule includes: and processing the ultrasonic video stream containing the thyroid ultrasonic image by using a nodule heavy identification model to acquire the identification of thyroid nodules, wherein the nodule heavy identification model is a trained neural network model. In particular, the nodule re-identification model may employ a re-identification technique, for each thyroid nodule in the thyroid ultrasound image, matching the unique nodule and configuring a unique identifier for it according to the historical information.
Optionally, fig. 4 is a flowchart illustrating training of a nodule re-recognition model in an embodiment of the present application. As shown in fig. 4, the training of the nodule re-recognition model in the embodiment of the present application includes the following steps S41 to S45.
S41, a nodule re-identification model is constructed, for example, the nodule re-identification model may be constructed by using a deep learning image classification model based on resnet.
And S42, obtaining training data, wherein the training data comprises a plurality of thyroid images and classification labels of thyroid nodules in the thyroid images.
And S43, training the nodule weight recognition model by using the training data. In this process, the classifier can be trained by metric learning using different nodules.
And S44, carrying out algorithm optimization processing on the nodule re-recognition model.
And S45, testing the trained nodule weight recognition model.
Fig. 5 is a flowchart illustrating detection information and historical detection information corresponding to a fused thyroid ultrasound image according to an embodiment of the present disclosure. As shown in fig. 5, the detection information and the historical detection information corresponding to the fused thyroid ultrasound image in the embodiment of the present application include the following steps S51 and S52.
And S51, filtering the detection information corresponding to the thyroid ultrasound image and the historical detection information. For example, a threshold value may be set, and detection information with unreliable results and low stability may be filtered according to the threshold value.
And S52, fusing the detection information corresponding to the thyroid ultrasound image and the historical detection information to generate a structural detection result of the thyroid nodule. For example, stable and reliable structured detection results can be output through a voting and clustering mechanism.
Optionally, the thyroid ultrasound image processing method provided in the embodiment of the present application may be used to process an ultrasound image in real time. Fig. 6 shows a flow chart for real-time processing of ultrasound images. As shown in fig. 6, the real-time processing of the ultrasound image includes the following steps S61 to S67.
And S61, acquiring the thyroid gland ultrasonic image acquired by the ultrasonic image acquisition device in real time as a current frame image.
And S62, processing the current frame image to acquire the current frame position of the thyroid nodule. The current frame position of the thyroid nodule refers to the position of the thyroid nodule in the current frame image.
And S63, predicting the position of the thyroid nodule in the next frame of image by using a Kalman tracker and combining the current frame position and the historical frame position of the thyroid nodule. The historical frame position of the thyroid nodule refers to the position of the thyroid nodule in a historical frame, and the historical frame can be a plurality of frames of images acquired before the current frame.
And S64, processing the ultrasonic video stream of the patient by using the nodule re-identification model to configure the identification of each thyroid nodule in the current frame image. And for each thyroid nodule in the current frame image, matching a unique nodule according to the historical information and configuring a unique identifier for the unique nodule.
And S65, updating and maintaining the Kalman tracker by using the current frame position of the thyroid nodule acquired in the step S62.
And S66, fusing the detection information corresponding to the current frame image and the detection information corresponding to the historical frame image to generate a structural detection result of the thyroid nodule. For example, the detection information corresponding to the current frame image and the detection information corresponding to the historical frame image may be subjected to a packing process to obtain a structured detection result. The detection information corresponding to the current frame image includes: and predicting the position of the thyroid nodule in the next frame image according to the current frame image and the identification of each thyroid nodule in the current frame image. The detection information corresponding to any historical frame comprises: and predicting the position of the thyroid nodule in the next frame image according to the historical frame, and predicting the position of the thyroid nodule in the historical frame image.
And S67, filtering out detection information with low confidence in the structured detection result according to a set threshold value, and outputting stable structured report data.
Optionally, the real-time processing of the ultrasound image when a new thyroid nodule is detected in step S62 may further include: a kalman tracker is initialized for the new thyroid nodule.
Through the steps S61 to S67, the current frame image acquired by the ultrasound image acquisition device can be processed in real time to obtain the structured report data corresponding to the current frame image, so as to implement real-time detection and processing of the ultrasound image.
The scope of the method for processing an ultrasound image of a thyroid gland according to the embodiment of the present application is not limited to the order of executing the steps listed in this embodiment, and all the solutions implemented by adding, subtracting, and replacing steps in the prior art according to the principles of the present application are included in the scope of the present application.
The embodiment of the present application further provides a thyroid ultrasound image processing device, which can implement the thyroid ultrasound image processing method of the present application, but the implementation device of the thyroid ultrasound image processing method of the present application includes but is not limited to the structure of the thyroid ultrasound image processing device recited in the embodiment of the present application, and all structural modifications and substitutions in the prior art made according to the principles of the present application are included in the scope of protection of the present application.
Fig. 7 is a schematic structural diagram of a thyroid ultrasound image processing apparatus 700 according to an embodiment of the present application. As shown in fig. 7, the thyroid ultrasound image processing apparatus 700 provided in the embodiment of the present application includes an image acquisition module 710, a first processing module 720, a prediction module 730, a second processing module 740, and an information fusion module 750. The image acquisition module 710 is used for acquiring an ultrasound image of the thyroid gland. The first processing module 720 is used for performing a first processing on the thyroid ultrasound image to acquire the position of the thyroid nodule. The prediction module 730 is configured to predict a thyroid nodule position in the next frame of ultrasound image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule. The second processing module 740 is configured to perform a second processing on the thyroid ultrasound image to obtain an identification of a thyroid nodule. The information fusion module 750 is configured to fuse detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, where the detection information corresponding to the thyroid ultrasound image includes the predicted position and the thyroid nodule identifier.
The modules in the thyroid ultrasound image processing apparatus 700 correspond to the steps S21 to S25 in the thyroid ultrasound image processing method shown in fig. 2 one by one, and are not described herein again for saving the description.
The present application also provides a computer readable storage medium having a computer program stored thereon. The computer program is used for realizing the thyroid ultrasound image processing method according to any embodiment of the application when being executed by a processor.
Any combination of one or more storage media may be employed herein. The storage medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a RAM, a ROM, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The application also provides an electronic device. Fig. 8 is a schematic structural diagram of an electronic device 800 according to an embodiment of the application. As shown in fig. 8, the electronic device 800 in this embodiment includes a memory 810 and a processor 820.
The memory 810 is used to store computer programs; preferably, the memory 810 includes: various media that can store program codes, such as ROM, RAM, magnetic disk, U-disk, memory card, or optical disk.
In particular, memory 810 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) and/or cache memory. The electronic device 800 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. Memory 810 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the application.
Alternatively, the Processor 820 may be a general-purpose Processor including a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components.
Optionally, in this embodiment, the electronic device 800 may further include a display 830. The display 830 is communicatively coupled to the memory 810 and the processor 820 for displaying a GUI interactive interface associated with the thyroid ultrasound image processing method.
In summary, according to the thyroid ultrasound image processing method provided in the embodiment of the present application, the detection result of the thyroid ultrasound image is obtained according to the detection information corresponding to the thyroid ultrasound image and the historical detection information. The method fully considers the correlation of the inter-frame historical information, and is beneficial to improving the accuracy of thyroid nodule detection. In addition, based on the thyroid ultrasound image processing method provided by the embodiment of the application, a doctor can lock the thyroid nodule through a target tracking method in the process of scanning the thyroid nodule, so that a result with high reliability and high stability is output. Therefore, the application effectively overcomes various defects in the prior art and has high industrial utilization value.
The above embodiments are merely illustrative of the principles and utilities of the present application and are not intended to limit the present application. Any person skilled in the art can modify or change the above-described embodiments without departing from the spirit and scope of the present application. Accordingly, it is intended that all equivalent modifications or changes which may be made by those skilled in the art without departing from the spirit and technical spirit of the present disclosure be covered by the claims of the present application.
Claims (10)
1. A thyroid ultrasound image processing method is characterized by comprising the following steps:
obtaining an ultrasonic thyroid image;
performing first processing on the thyroid ultrasonic image to acquire the position of a thyroid nodule;
predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule;
performing second processing on the thyroid ultrasound image to acquire an identification of a thyroid nodule;
and fusing detection information corresponding to the thyroid ultrasound image and historical detection information to obtain a detection result of the thyroid ultrasound image, wherein the detection information corresponding to the thyroid ultrasound image comprises the predicted position and the thyroid nodule identifier.
2. The method of claim 1, wherein the first processing of the thyroid ultrasound image to obtain the location of the thyroid nodule comprises:
detecting the thyroid ultrasound image by using a deep learning target detection method to obtain an ultrasound image area;
and acquiring the position of the thyroid nodule by using a two-level target detection method according to the image in the ultrasonic image area.
3. The method of claim 1, wherein the thyroid ultrasound image comprises at least one thyroid nodule, each thyroid nodule corresponding to a kalman tracker; the predicting the thyroid nodule position in the next frame of ultrasound image as the predicted position comprises:
processing the position of the thyroid nodule and the historical position of the thyroid nodule with the kalman tracker to predict a thyroid nodule position in a next frame of ultrasound image as the predicted position.
4. The thyroid ultrasound image processing method according to claim 3, further comprising: updating and maintaining each of the Kalman trackers based on the location of the thyroid nodule.
5. The thyroid ultrasound image processing method according to claim 3, further comprising: initializing the Kalman tracker.
6. The method of claim 1, wherein the second processing of the thyroid ultrasound image to obtain an identification of thyroid nodule comprises:
and processing the ultrasonic video stream containing the thyroid ultrasonic image by using a nodule weight recognition model to obtain the identification of the thyroid nodule, wherein the nodule weight recognition model is a trained neural network model.
7. The method for processing the thyroid ultrasound image according to claim 1, wherein fusing the detection information corresponding to the thyroid ultrasound image and the historical detection information comprises:
filtering detection information corresponding to the thyroid ultrasound image and historical detection information;
and fusing the detection information corresponding to the thyroid ultrasonic image and the historical detection information to generate a structural detection result of the thyroid nodule.
8. A thyroid ultrasound image processing apparatus, comprising:
the image acquisition module is used for acquiring a thyroid ultrasound image;
the first processing module is used for carrying out first processing on the thyroid ultrasonic image so as to acquire the position of a thyroid nodule;
the prediction module is used for predicting the position of the thyroid nodule in the next frame of ultrasonic image as a predicted position according to the position of the thyroid nodule and the historical position of the thyroid nodule;
the second processing module is used for carrying out second processing on the thyroid ultrasonic image so as to acquire the identification of the thyroid nodule;
and the information fusion module is used for fusing detection information and historical detection information corresponding to the thyroid ultrasonic image to obtain a detection result of the thyroid ultrasonic image, wherein the detection information corresponding to the thyroid ultrasonic image comprises the predicted position and the thyroid nodule identification.
9. A computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program, when executed by a processor, implements the thyroid ultrasound image processing method of any one of claims 1 to 7.
10. An electronic device, characterized in that the electronic device comprises:
a memory storing a computer program;
a processor, communicatively coupled to the memory, that executes the method of processing a thyroid ultrasound image according to any one of claims 1 to 7 when the computer program is invoked.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211358249.0A CN115564763A (en) | 2022-11-01 | 2022-11-01 | Thyroid ultrasound image processing method, device, medium and electronic equipment |
| PCT/CN2023/083851 WO2024093099A1 (en) | 2022-11-01 | 2023-03-24 | Thyroid ultrasound image processing method and apparatus, medium and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211358249.0A CN115564763A (en) | 2022-11-01 | 2022-11-01 | Thyroid ultrasound image processing method, device, medium and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115564763A true CN115564763A (en) | 2023-01-03 |
Family
ID=84769025
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211358249.0A Pending CN115564763A (en) | 2022-11-01 | 2022-11-01 | Thyroid ultrasound image processing method, device, medium and electronic equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115564763A (en) |
| WO (1) | WO2024093099A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024093099A1 (en) * | 2022-11-01 | 2024-05-10 | 上海杏脉信息科技有限公司 | Thyroid ultrasound image processing method and apparatus, medium and electronic device |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103505288B (en) * | 2012-06-29 | 2017-11-17 | 通用电气公司 | Ultrasonic imaging method and supersonic imaging apparatus |
| US10993653B1 (en) * | 2018-07-13 | 2021-05-04 | Johnson Thomas | Machine learning based non-invasive diagnosis of thyroid disease |
| CN111461202B (en) * | 2020-03-30 | 2023-12-05 | 上海深至信息科技有限公司 | Thyroid nodule ultrasonic image real-time identification method and device |
| CN111862044B (en) * | 2020-07-21 | 2024-06-18 | 深圳蓝湘智影科技有限公司 | Ultrasonic image processing method, ultrasonic image processing device, computer equipment and storage medium |
| CN114693640A (en) * | 2022-03-31 | 2022-07-01 | 合肥合滨智能机器人有限公司 | Ultrasonic image focus attribute classification processing method, system, electronic equipment and storage medium based on video sequence |
| CN115564763A (en) * | 2022-11-01 | 2023-01-03 | 上海杏脉信息科技有限公司 | Thyroid ultrasound image processing method, device, medium and electronic equipment |
-
2022
- 2022-11-01 CN CN202211358249.0A patent/CN115564763A/en active Pending
-
2023
- 2023-03-24 WO PCT/CN2023/083851 patent/WO2024093099A1/en unknown
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024093099A1 (en) * | 2022-11-01 | 2024-05-10 | 上海杏脉信息科技有限公司 | Thyroid ultrasound image processing method and apparatus, medium and electronic device |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024093099A1 (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111428709B (en) | Image processing method, device, computer equipment and storage medium | |
| Tilve et al. | Pneumonia detection using deep learning approaches | |
| US10691980B1 (en) | Multi-task learning for chest X-ray abnormality classification | |
| CN109102490B (en) | Automatic image registration quality assessment | |
| CN109741312B (en) | Lung nodule identification method, device, equipment and medium | |
| US11996198B2 (en) | Determination of a growth rate of an object in 3D data sets using deep learning | |
| Cong et al. | BCS-Net: Boundary, context, and semantic for automatic COVID-19 lung infection segmentation from CT images | |
| CN111325714A (en) | Region-of-interest processing method, computer device and readable storage medium | |
| CN113313699B (en) | X-ray chest disease classification and positioning method based on weak supervision learning and electronic equipment | |
| CN110648318A (en) | Auxiliary analysis method and device for skin diseases, electronic equipment and storage medium | |
| CN114494215A (en) | Transformer-based thyroid nodule detection method | |
| CN115564763A (en) | Thyroid ultrasound image processing method, device, medium and electronic equipment | |
| Xu et al. | Lung segmentation in chest X‐ray image using multi‐interaction feature fusion network | |
| Chaisangmongkon et al. | External validation of deep learning algorithms for cardiothoracic ratio measurement | |
| CN116051538A (en) | Left ventricle segmentation method based on transducer echocardiography | |
| CN114360695B (en) | Auxiliary system, medium and equipment for breast ultrasonic scanning and analyzing | |
| CN112530554B (en) | Scanning positioning method and device, storage medium and electronic equipment | |
| CN114010227B (en) | Right ventricle characteristic information identification method and device | |
| CN113222989B (en) | Image grading method and device, storage medium and electronic equipment | |
| CN111768367B (en) | Data processing method, device and storage medium | |
| KR20230040484A (en) | Chest medical image object detection system and method thereof | |
| Zhang et al. | CXR-Net: an encoder-decoder-encoder multitask deep neural network for explainable and accurate diagnosis of COVID-19 pneumonia with chest X-ray images | |
| EP4022561A1 (en) | Methods for analyzing and reducing inter/intra site variability using reduced reference images and improving radiologist diagnostic accuracy and consistency | |
| Guo et al. | A Benchmark and Transformer-based Approach for Automated Hyperparathyroidism Detection | |
| CN113223104B (en) | Cardiac MR image interpolation method and system based on causal relationship |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |