CN117765330A - MRI image-based data labeling method and system - Google Patents

Abstract

The present invention relates to the field of data annotation, and proposes a data annotation method and system based on MRI images. The method includes: obtaining an MRI denoised image and an MRI smoothed image through image denoising and smoothing processing, and performing target area segmentation on the MRI smoothed image. And extract the parameters of the area, calculate the edge intensity of the target area, based on the edge intensity, perform corner detection in the remaining area of the MRI smooth image, extract the corner suppression value and information factor, and use it to identify the texture features of the remaining area. According to the texture Features construct an image annotation model of the original MRI image, and use the training set to train the model to obtain a trained image annotation model. Use the trained model to annotate and predict the original MRI image, obtain the predicted annotated image, and remove the overlapping parts. Get the final image annotation result. The invention can improve the accuracy of MRI image data annotation.

Description

Data annotation method and system based on MRI images

Technical field

The present invention relates to the field of data annotation, and in particular to a data annotation method and system based on MRI images.

Background technique

MRI image is a medical imaging technology that obtains images by utilizing the resonance phenomenon of atomic nuclei under the action of strong magnetic fields and radio frequency pulses. MRI images are used to observe the structure and function of internal tissues and organs of the human body, and are of clinical importance in diagnosing diseases. Value, MRI images can also present high-contrast and high-definition anatomical information, while also providing functional information and observation of dynamic processes.

Currently, the method of data annotation in MRI images can be implemented using automated annotation methods, which mainly use computer vision technology and machine learning algorithms to automatically identify and annotate MRI images, and require a large amount of annotated training data to train the model. , to learn the characteristics of specific lesions or structures, so as to automatically label the corresponding areas in the image. However, when processing complex MRI images, due to the differences between different types of lesions and structures, there will be certain errors, so , a data annotation method and system based on MRI images is needed to improve the accuracy of MRI image data annotation.

Contents of the invention

The present invention provides a data annotation method and system based on MRI images, and its main purpose is to improve the accuracy of MRI image data annotation.

In order to achieve the above objectives, the data annotation method based on MRI images provided by the present invention includes:

Obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and smooth the MRI denoised image to obtain a smoothed MRI image;

Segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, query the gradient factor corresponding to the image gradient, and based on the Gradient factor, calculates the edge strength of the target area;

Based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain corner suppression values, information factors corresponding to the corner suppression values are extracted, and based on the information factors, the corresponding remaining areas are identified texture characteristics;

Based on the texture features, construct an image annotation model corresponding to the original MRI image, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to model the image annotation model Train to obtain the trained image annotation model;

The trained model is used to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and the overlap of the predicted annotated image is removed to obtain an image annotation result corresponding to the original MRI image.

Optionally, obtaining an original MRI image, performing image denoising on the original MRI image, and obtaining an MRI denoised image includes:

Use preset medical equipment to perform an MRI scan on the patient and obtain the original MRI image;

Perform noise analysis on the original MRI image to obtain noise data;

Identify the noise direction corresponding to the noise data;

Based on the noise direction, image denoising is performed on the original MRI image to obtain an MRI denoised image.

Optionally, the smoothing process on the MRI denoised image to obtain a smooth MRI image includes:

Identify image requirements corresponding to the MRI denoised image;

Based on the image requirements, set the smoothing parameters corresponding to the MRI denoising image;

Based on the smoothing parameters, a preset smoothing filter is enabled to smooth the MRI denoised image to obtain an MRI smoothed image.

Optionally, segmenting the target area in the MRI smooth image and extracting regional parameters corresponding to the target area includes:

Identify the target factor corresponding to the MIR smooth image;

Based on the target factor, determine the target area corresponding to the MRI smooth image;

Segment the target area in the MRI smooth image using a preset segmentation algorithm;

Query the regional requirements corresponding to the target area, and extract the regional parameters corresponding to the regional requirements.

Optionally, calculating the image gradient corresponding to the target area based on the area parameters includes:

First, use the following formula to calculate the average gray value corresponding to the target area:

Wherein, MG represents the average gray value corresponding to the target area, N represents the number of pixels in the target area, and I(x,y) represents the gray value of the pixel at the coordinate (x, y).

Secondly, based on the average gray value, calculate the sum of squared differences between each pixel in the target area and its neighbor pixels:

SD=∑[I(x,y)-I(x',y')] ²

Among them, SD represents the sum of squared differences between each pixel in the target area and its neighboring pixels, (x', y') represents the neighborhood pixel position of the (x, y) pixel, I(x, y) and I( x', y') respectively represent the pixel gray value in the target area.

Finally, based on the sum of squared differences, the image gradient corresponding to the target area is calculated:

Where, GT represents the image gradient corresponding to the target area, and N represents the number of pixels in the target area.

Optionally, calculating the edge strength of the target area based on the gradient factor includes:

Among them, BQ represents the edge strength of the target area, (i,j) represents the pixel coordinates in the image, MB represents the target area, Gx(i,j) and Gy(i,j) respectively represent the pixel point (i , the gradient values in the horizontal and vertical directions at j), M represents the total number of pixels in the target area.

Optionally, based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain a corner suppression value, including:

Based on the edge intensity, determine the edge image corresponding to the remaining area;

Perform corner point monitoring on the edge image to obtain corner point parameters;

Based on the corner point parameters, calculate the corner point response value corresponding to the remaining area;

Perform non-maximum suppression on the corner point response value to obtain a corner point suppression value.

Optionally, extracting information factors corresponding to the corner point suppression values, and identifying texture features corresponding to the remaining areas based on the information factors includes:

Determine the fixed neighborhood corresponding to the corner point suppression value, and extract the information factor corresponding to the fixed neighborhood;

Determine the texture category corresponding to the information factor;

Based on the texture category, texture features corresponding to the remaining area are identified.

Optionally, using a trained model to perform annotation prediction on the original MRI image to obtain a predicted annotation image includes:

Obtain MRI image samples corresponding to the original MRI images;

Perform equalization processing on the MRI image samples to obtain equalized samples;

Use the trained model to perform image prediction on the balanced sample to obtain a predicted image;

Image annotation is performed on the predicted image to obtain a predicted annotated image.

In order to solve the above problems, the present invention also provides a data annotation system based on MRI images. The system includes:

A smoothing processing module, used to obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and perform smoothing processing on the MRI denoised image to obtain an MRI smoothed image;

An edge calculation module is used to segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, and query the image gradient corresponding to the Gradient factor, based on the gradient factor, calculate the edge strength of the target area;

A texture recognition module, configured to perform corner point detection on the remaining area of the MRI smooth image based on the edge intensity, obtain a corner point suppression value, and extract an information factor corresponding to the corner point suppression value. Based on the information factor, Identify the texture features corresponding to the remaining area;

A model training module, configured to construct an image annotation model corresponding to the original MRI image based on the texture features, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to The above image annotation model is used for model training, and a trained image annotation model is obtained;

The image annotation module is used to use the trained model to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and to remove overlaps on the predicted annotated image to obtain an image annotation result corresponding to the original MRI image.

The present invention obtains original MRI images and performs image denoising on the original MRI images to obtain MRI denoised images, which can improve image quality, improve image analysis and diagnosis effects, reduce misdiagnosis rates, and optimize subsequent image processing processes, which can reduce The influence of noise makes the image clearer and the details clearer, thereby improving the accuracy of detection. The present invention can help accurate positioning and detection by segmenting the target area in the MRI smooth image and extracting the regional parameters corresponding to the target area. Analyze the target area to provide strong support for disease diagnosis, treatment and research, and can realize automated image analysis, thereby saving manpower and material costs and improving efficiency. Based on the edge intensity, the present invention performs an analysis on the remaining area of the MRI smooth image. Corner detection and obtaining corner suppression values can provide richer image features, enhance the accuracy of target positioning, assist image registration and alignment, and assist image segmentation and other applications, thus helping to improve the effect of medical image processing and Quality, the present invention constructs an image annotation model corresponding to the original MRI image based on the texture characteristics, and uses the image annotation model to divide the image training set corresponding to the MRI smooth image, which can automatically perform image annotation and classification, and improve Processing efficiency, in order to better analyze and process different types of images, and help carry out corresponding research and treatment plans in a targeted manner. The present invention obtains the corresponding original MRI image by removing overlaps on the predicted annotated images. Image annotation results can improve image quality, improve readability, promote image analysis and processing effects, and improve lesion detection and diagnostic accuracy, which has important benefits for research and clinical practice in the medical field. Therefore, the present invention proposes a data annotation method and system based on MRI images to improve the accuracy of MRI image data annotation.

Description of the drawings

Figure 1 is a schematic flow chart of a data annotation method based on MRI images provided by an embodiment of the present invention;

Figure 2 is a schematic module diagram of a data annotation system based on MRI images provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of the internal structure of an electronic device according to an MRI image-based data annotation method provided by an embodiment of the present invention.

The realization of the purpose, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present invention and are not intended to limit the present invention.

The embodiment of the present application provides a data annotation method based on MRI images. The execution subject of the MRI image-based data annotation method includes, but is not limited to, at least one of a server, a terminal, and other electronic devices that can be configured to execute the method provided by the embodiments of the present application. In other words, the MRI image-based data annotation method can be executed by software or hardware installed on the terminal device or the server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server may be an independent server, or may provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, and content delivery networks (Content Delivery Network, CDN), and cloud servers for basic cloud computing services such as big data and artificial intelligence platforms.

Refer to FIG. 1 , which is a schematic flow chart of a data annotation method based on MRI images provided by an embodiment of the present invention. In this embodiment, the data annotation method based on MRI images includes:

S1. Obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and smooth the MRI denoised image to obtain a smoothed MRI image.

The present invention obtains original MRI images and performs image denoising on the original MRI images to obtain MRI denoised images, which can improve image quality, improve image analysis and diagnosis effects, reduce misdiagnosis rates, and optimize subsequent image processing processes, which can reduce The influence of noise makes the image clearer and the details clearer, thereby improving the accuracy of detection.

Wherein, the original MRI image refers to an MRI image collected directly from the patient without any processing; the MRI denoising image refers to applying an image processing algorithm to the original MRI image to reduce or remove noise and artifacts in the image. The image obtained after removing undesirable image features such as shadows.

As an embodiment of the present invention, obtaining an original MRI image, performing image denoising on the original MRI image, and obtaining an MRI denoised image includes: using preset medical equipment to perform an MRI scan on the patient to obtain the original MRI image. ; Perform noise analysis on the original MRI image to obtain noise data; identify the noise direction corresponding to the noise data; perform image denoising on the original MRI image based on the noise direction to obtain an MRI denoised image.

Wherein, the preset medical equipment refers to the medical imaging equipment predetermined to perform MRI scanning on the patient, such as a magnetic resonance imaging machine; the noise data refers to the interference signal present in the original MRI image. ; The noise direction refers to determining the directionality of the noise in the image based on the characteristics and distribution of the noise data.

Further, the noise data can be obtained through noise analysis tools, such as Audacity, Adobe Audition and other tools; the noise direction can be obtained through beamforming algorithms, such as Delay-and-Sum, Beamforming and other algorithms.

The present invention obtains an MRI smoothed image by smoothing the MRI denoised image, and can obtain a clearer, more contrasty and accurate image by enhancing the image quality, which helps to obtain more accurate information, thereby avoiding the need for Repeat the scan.

Wherein, the MRI smooth image refers to an MRI image that reduces or eliminates noise in the image while retaining key image features.

As an embodiment of the present invention, smoothing the MRI denoised image to obtain an MRI smoothed image includes: identifying image requirements corresponding to the MRI denoised image; based on the image requirements, setting the MRI Smoothing parameters corresponding to the denoised image; based on the smoothing parameters, enable a preset smoothing filter to smooth the MRI denoised image to obtain an MRI smoothed image.

Wherein, the image requirement refers to the input image required for image processing or analysis, which can be an original image or a preprocessed image; the smoothing parameter refers to a smoothing filter used to control the image smoothing operation. parameters, such as: window size, convolution kernel size, etc.; the preset smoothing filter refers to a filter designed according to a specific mathematical model for image smoothing processing, such as: mean filter, Gaussian filter wait.

Further, the image requirements can be obtained through image processing tools, such as: OpenCV, Pillow, Scikit-image, etc.; the smoothing parameters can be obtained through statistical analysis tools, such as: Matplotlib, NumPy, SciPy, etc.

S2. Segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, query the gradient factor corresponding to the image gradient, based on The gradient factor calculates the edge strength of the target area.

By segmenting the target area in the MRI smooth image and extracting the regional parameters corresponding to the target area, the present invention can help accurately locate and analyze the target area, provide strong support for disease diagnosis, treatment and research, and can realize automated image processing. Analysis, thereby saving human and material costs and improving efficiency.

Wherein, the target area refers to a specific area of interest in the MRI smooth image, such as a tumor area, brain structure, etc.; the area parameters refer to quantitative parameters obtained by analyzing the characteristics of the target area.

As an embodiment of the present invention, segmenting the target area in the MRI smooth image and extracting the regional parameters corresponding to the target area includes: identifying the target factor corresponding to the MIR smooth image; based on the target factor, Determine the target area corresponding to the MRI smooth image; segment the target area in the MRI smooth image using a preset segmentation algorithm; query the regional requirements corresponding to the target area; and extract the regional parameters corresponding to the regional requirements.

Wherein, the target factor refers to a specific factor or feature used to identify the target of interest in the MRI smooth image; the preset segmentation algorithm refers to a specific algorithm that separates the target area and the background area in the MRI smooth image, Commonly used preset segmentation algorithms include: threshold segmentation, edge detection, region growing, graph theory-based methods, etc.; the region requirements refer to specific attributes or parameters that need to be analyzed, measured or extracted in the target region.

Further, the target factors can be obtained through machine learning models, such as support vector machines SVM, random forest RF, deep learning models, etc.; the regional requirements can be obtained through threshold segmentation algorithms, such as Otsu, adaptive threshold and other algorithms.

Optionally, as an embodiment of the present invention, calculating the image gradient corresponding to the target area based on the area parameters includes:

First, use the following formula to calculate the average gray value corresponding to the target area:

Wherein, MG represents the average gray value corresponding to the target area, N represents the number of pixels in the target area, and I(x,y) represents the gray value of the pixel at the coordinate (x, y).

Secondly, based on the average gray value, calculate the sum of squared differences between each pixel in the target area and its neighbor pixels:

SD=∑[I(x,y)-I(x',y')] ²

Among them, SD represents the sum of squared differences between each pixel in the target area and its neighboring pixels, (x',y') represents the neighborhood pixel position of the (x,y) pixel, I(x,y) and I( x', y') respectively represent the pixel gray value in the target area.

Finally, based on the sum of squared differences, the image gradient corresponding to the target area is calculated:

Where, GT represents the image gradient corresponding to the target area, and N represents the number of pixels in the target area.

The present invention queries the gradient factor corresponding to the image gradient and calculates the edge strength of the target area based on the gradient factor, which can separate the target from the background and help identify and classify different objects in the image, making the edges more clear. significantly, thus enhancing the visual effect of the target area.

Wherein, the image gradient refers to the direction and intensity of the fastest changing pixel value in the image; the gradient factor refers to the modulus length of the image gradient, also known as the gradient amplitude or gradient size; the edge strength refers to the edge strength of the image. Edge prominence or intensity.

Optionally, the gradient factor can be obtained through a detection algorithm, such as Prewitt operator and Laplacian operator.

As an embodiment of the present invention, calculating the edge strength of the target area based on the gradient factor includes:

Among them, BQ represents the edge strength of the target area, (i,j) represents the pixel coordinates in the image, MB represents the target area, Gx(i,j) and Gy(i,j) respectively represent the pixel point (i , the gradient values in the horizontal and vertical directions at j), M represents the total number of pixels in the target area.

S3. Based on the edge intensity, perform corner detection on the remaining area of the MRI smooth image to obtain the corner suppression value, extract the information factor corresponding to the corner suppression value, and identify the remaining area based on the information factor. Texture features corresponding to the region.

Based on the edge intensity, the present invention performs corner point detection on the remaining areas of the MRI smooth image to obtain corner point suppression values, which can provide richer image features, enhance the accuracy of target positioning, and assist image registration and alignment. and auxiliary image segmentation and other applications, thus helping to improve the effect and quality of medical image processing.

Wherein, the remaining area refers to the image area that is not covered by smoothing after the smoothing operation; the corner suppression value refers to the position and attribute information of the corner points obtained by performing corner point detection on the remaining area. .

As an embodiment of the present invention, the step of performing corner point detection on the remaining area of the MRI smooth image based on the edge intensity to obtain the corner point suppression value includes: based on the edge intensity, determining the corresponding area of the remaining area. edge image; perform corner point monitoring on the edge image to obtain corner point parameters; calculate the corner point response value corresponding to the remaining area based on the corner point parameter; perform non-maximum processing on the corner point response value Suppress, get the corner suppression value.

Wherein, the edge image refers to a binary image obtained according to the edge intensity of the image (for example, the edge extracted by the Canny algorithm); the corner parameter refers to the calculated value when corner detection is performed on the edge image. Characteristic parameters of each corner point; the corner response value refers to an important value of each corner point in the image calculated based on the corner point parameters.

Further, the edge image can be obtained through edge detection algorithms, such as: Canny algorithm, Sobel operator, Laplacian operator, etc.: The corner point parameters can be obtained through corner point detection algorithms, such as: Harris corner point detection algorithm , Shi-Tomasi corner detection algorithm, FAST corner detection algorithm, etc.: The corner response value can be obtained through the corner response function, such as: functions cv2.cornerHarris(), cv2.goodFeaturesToTrack(), etc. in the OpenCV library function.

By extracting information factors corresponding to the corner suppression values, and identifying texture features corresponding to the remaining areas based on the information factors, the present invention can bring more refined, accurate, and robust texture feature extraction and classification results, thereby Improve application effects and performance in areas such as image analysis, computer vision, and pattern recognition.

Among them, the information factor refers to a quantitative index calculated and analyzed by the corner point suppression value and its surrounding pixels; the texture feature refers to a feature used to describe and distinguish different materials, textures and morphological features.

As an embodiment of the present invention, extracting the information factor corresponding to the corner suppression value, and identifying the texture features corresponding to the remaining area based on the information factor includes: determining the fixed value corresponding to the corner suppression value. Neighborhood; extract information factors corresponding to the fixed neighborhood; determine texture categories corresponding to the information factors; and identify texture features corresponding to the remaining areas based on the texture categories.

Wherein, the fixed neighborhood refers to a fixed-size window or neighborhood defined in order to extract information factors during the corner suppression value calculation process; the texture category refers to the texture feature judged based on the information factor. classification categories.

Further, the fixed neighborhood can be obtained through Python processing library, such as: PIL, scikit-image, etc.; the texture category can be obtained through LBP algorithm.

S4. Based on the texture features, construct an image annotation model corresponding to the original MRI image, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to annotate the image model Carry out model training and obtain the trained image annotation model.

Based on the texture features, the present invention constructs an image annotation model corresponding to the original MRI image, and uses the image annotation model to divide the image training set corresponding to the MRI smooth image, which can automatically perform image annotation and classification and improve processing efficiency. , in order to better analyze and process different types of images, and help carry out corresponding research and treatment plans.

Among them, the image annotation model means that the image can be automatically assigned corresponding annotations or classifications based on the extracted texture features and other information; the image training set refers to a set of known annotations used to train and optimize the image annotation model. or a collection of classified image samples.

Optionally, the image annotation model can be obtained through model building tools, such as TensorFlow, PyTorch and other tools; the image training set can be obtained through partitioning tools, such as LabelImg, RectLabel and other tools.

By using the training set to perform model training on the image annotation model, the present invention obtains a trained image annotation model, which can improve the accuracy of annotation, automate the annotation process, improve the generalization ability and adaptability of the model, and also provide Optimization and improvement of the model provide the basis to make the model more suitable for specific tasks and application scenarios.

Wherein, the trained image annotation model refers to a model that has higher accuracy and generalization ability after being trained and optimized on the training set. Optionally, the trained image annotation model can be used by using the trained image annotation model. The set is obtained by performing model training on the image annotation model.

S5. Use the trained model to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and remove overlaps on the predicted annotated image to obtain an image annotation result corresponding to the original MRI image.

The present invention uses a trained model to annotate and predict the original MRI image to obtain a predicted annotated image, which can intuitively display the structure, lesions or features in the image, helping doctors quickly understand the patient's condition and make corresponding diagnosis. and treatment decisions.

Wherein, the predicted annotated image refers to an image generated by annotating and predicting the original MRI image through the trained model.

As an embodiment of the present invention, using a trained model to annotate and predict the original MRI image to obtain a predicted annotated image includes: obtaining an MRI image sample corresponding to the original MRI image; Perform equalization processing to obtain a balanced sample; use a trained model to perform image prediction on the balanced sample to obtain a predicted image; perform image annotation on the predicted image to obtain a predicted annotated image.

Wherein, the MRI image sample refers to an image sample obtained from a medical database or other sources, including information about tissue type, density, shape, etc.; the balanced sample refers to an image sample obtained after equalizing the MRI image sample. sample.

Further, the MRI image sample can be obtained through an MRI scanner; the equalized sample can be obtained through an equalization algorithm, such as: histogram equalization, adaptive histogram equalization, etc.

The present invention obtains image annotation results corresponding to the original MRI images by removing overlaps on the predicted annotation images, which can improve image quality, improve readability, promote image analysis and processing effects, and improve the accuracy of lesion detection and diagnosis. , which has important benefits for research and clinical practice in the medical field.

The image annotation result refers to the result of labeling and annotating the original MRI image. Optionally, the image annotation result can be obtained through a neural network model, such as U-Net, ResNet, VGG and other models.

The present invention obtains original MRI images and performs image denoising on the original MRI images to obtain MRI denoised images, which can improve image quality, improve image analysis and diagnosis effects, reduce misdiagnosis rates, and optimize subsequent image processing processes, which can reduce The influence of noise makes the image clearer and the details clearer, thereby improving the accuracy of detection. The present invention can help accurate positioning and detection by segmenting the target area in the MRI smooth image and extracting the regional parameters corresponding to the target area. Analyze the target area to provide strong support for disease diagnosis, treatment and research, and can realize automated image analysis, thereby saving manpower and material costs and improving efficiency. Based on the edge intensity, the present invention performs an analysis on the remaining area of the MRI smooth image. Corner detection and obtaining corner suppression values can provide richer image features, enhance the accuracy of target positioning, assist image registration and alignment, and assist image segmentation and other applications, thus helping to improve the effect of medical image processing and Quality, the present invention constructs an image annotation model corresponding to the original MRI image based on the texture characteristics, and uses the image annotation model to divide the image training set corresponding to the MRI smooth image, which can automatically perform image annotation and classification, and improve Processing efficiency, in order to better analyze and process different types of images, and help carry out corresponding research and treatment plans in a targeted manner. The present invention obtains the corresponding original MRI image by removing overlaps on the predicted annotated images. Image annotation results can improve image quality, improve readability, promote image analysis and processing effects, and improve lesion detection and diagnostic accuracy, which has important benefits for research and clinical practice in the medical field. Therefore, the present invention proposes a data annotation method and system based on MRI images to improve the accuracy of MRI image data annotation. As shown in Figure 2, it is a functional module diagram of the MRI image-based data annotation method and system provided by an embodiment of the present invention.

The MRI image-based data annotation system 200 of the present invention can be installed in electronic equipment. According to the implemented functions, the MRI image-based data annotation system 200 may include a smoothing processing module 201, an edge calculation module 202, a texture recognition module 203, a model training module 204, and an image annotation module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of the electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

In this embodiment, the functions of each module/unit are as follows:

The smoothing processing module 201 is used to obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and perform smoothing processing on the MRI denoised image to obtain an MRI smoothed image;

The edge calculation module 202 is used to segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, and query the image The gradient factor corresponding to the gradient, based on the gradient factor, calculates the edge strength of the target area;

The texture recognition module 203 is configured to perform corner point detection on the remaining area of the MRI smooth image based on the edge intensity, obtain the corner point suppression value, extract the information factor corresponding to the corner point suppression value, and based on the Information factor, identifying the texture features corresponding to the remaining area;

The model training module 204 is configured to construct an image annotation model corresponding to the original MRI image based on the texture features, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training Collect and perform model training on the image annotation model to obtain a trained image annotation model;

The image annotation module 205 is used to use a trained model to annotate and predict the original MRI image to obtain a predicted annotated image, and remove overlaps from the predicted annotated image to obtain an image annotation result corresponding to the original MRI image. .

In detail, each module described in the MRI image-based data annotation system 200 described in the embodiment of the present invention adopts the same technical means as the MRI image-based data annotation method described in the accompanying drawings, and can generate The same technical effects will not be repeated here.

As shown in Figure 3, it is a schematic structural diagram of an electronic device that implements the data annotation method based on MRI images according to the present invention.

The electronic device 1 may include a processor 30, a memory 31, a communication bus 32 and a communication interface 33, and may also include a computer program stored in the memory 31 and executable on the processor 30, such as based on artificial intelligence. engineering safety supervision procedures.

The processor 30 may be composed of an integrated circuit in some embodiments, for example, it may be composed of a single packaged integrated circuit, or it may be composed of multiple integrated circuits packaged with the same function or different functions, including one or A combination of multiple central processing units (CPUs), microprocessors, digital processing chips, graphics processors and various control chips, etc. The processor 30 is the control core (ControlUnit) of the electronic device 1, using various interfaces and lines to connect various components of the entire electronic device, by running or executing programs or modules stored in the memory 31 (for example, executing Engineering safety supervision program based on artificial intelligence, etc.), and calls the data stored in the memory 31 to perform various functions of the electronic device and process data.

The memory 31 includes at least one type of readable storage medium. The readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. . In some embodiments, the memory 31 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 31 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), or a secure digital (SD) device equipped on the electronic device. ) card, Flash Card, etc. Further, the memory 31 may also include both an internal storage unit of the electronic device and an external storage device. The memory 31 can not only be used to store application software installed on the electronic device and various types of data, such as codes for database configuration connection programs, etc., but can also be used to temporarily store data that has been output or is to be output.

The communication bus 32 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, or the like. The bus can be divided into address bus, data bus, control bus, etc. The bus is configured to implement connection communication between the memory 31 and at least one processor 30 and the like.

The communication interface 33 is used for communication between the above-mentioned electronic device 1 and other devices, and includes a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which are generally used to establish communication connections between the electronic device and other electronic devices. The user interface may be a display (Display) or an input unit (such as a keyboard). Optionally, the user interface may also be a standard wired interface or a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, or the like. The display may also be appropriately referred to as a display screen or a display unit, and is used for displaying information processed in the electronic device and for displaying a visualized user interface.

FIG. 3 only shows an electronic device with components. Persons skilled in the art can understand that the structure shown in FIG. 3 does not limit the electronic device 1 and may include fewer or more components than shown in the figure. components, or combinations of certain components, or different arrangements of components.

For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) that supplies power to various components. Preferably, the power supply may be logically connected to the at least one processor 30 through a power management device, so that through the power management device The device implements functions such as charging management, discharge management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power failure detection circuits, power converters or inverters, power status indicators and other arbitrary components. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described again here.

It should be understood that the described examples are for illustrative purposes only.

The database configuration connection program stored in the memory 31 of the electronic device 1 is a combination of multiple computer programs. When run in the processor 30, it can realize:

Obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and smooth the MRI denoised image to obtain a smoothed MRI image;

Segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, query the gradient factor corresponding to the image gradient, and based on the Gradient factor, calculates the edge strength of the target area;

Based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain corner suppression values, information factors corresponding to the corner suppression values are extracted, and based on the information factors, the corresponding remaining areas are identified texture characteristics;

Based on the texture features, construct an image annotation model corresponding to the original MRI image, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to model the image annotation model Train to obtain the trained image annotation model;

The trained model is used to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and the overlap of the predicted annotated image is removed to obtain an image annotation result corresponding to the original MRI image.

Specifically, for the specific implementation method of the above computer program by the processor 30, reference can be made to the description of the relevant steps in the corresponding embodiment in Figure 1, which will not be described again here.

Furthermore, if the integrated modules/units of the electronic device 1 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium. The storage medium may be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Memory).

The present invention also provides a storage medium. The readable storage medium stores a computer program. When executed by a processor of an electronic device, the computer program can realize:

Obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and smooth the MRI denoised image to obtain a smoothed MRI image;

Segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, query the gradient factor corresponding to the image gradient, and based on the Gradient factor, calculates the edge strength of the target area;

Based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain corner suppression values, information factors corresponding to the corner suppression values are extracted, and based on the information factors, the corresponding remaining areas are identified texture characteristics;

Based on the texture features, construct an image annotation model corresponding to the original MRI image, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to model the image annotation model Train to obtain the trained image annotation model;

The trained model is used to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and the overlap of the predicted annotated image is removed to obtain an image annotation result corresponding to the original MRI image.

In the several embodiments provided by the present invention, it should be understood that the disclosed equipment, devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional module in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of hardware plus software function modules.

It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

Therefore, the embodiments should be regarded as illustrative and non-restrictive from any point of view, and the scope of the present invention is defined by the appended claims rather than the above description, and it is therefore intended that all claims falling within the claims All changes within the meaning and scope of equivalent elements are included in the present invention.

It should be noted that in this article, relational terms such as “first” and “second” are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these There is no such actual relationship or sequence between entities or operations. Furthermore, the terms "comprises," "comprises," or any other variations thereof are intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also those not expressly listed other elements, or elements inherent to the process, method, article or equipment. Without further limitation, an element defined by the statement "comprises a..." does not exclude the presence of additional identical elements in a process, method, article, or apparatus that includes the stated element.

The above descriptions are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be practiced in other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

1. A data annotation method based on MRI images, characterized in that the method includes: Obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and smooth the MRI denoised image to obtain a smoothed MRI image; Segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, query the gradient factor corresponding to the image gradient, and based on the Gradient factor, calculates the edge strength of the target area; Based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain corner suppression values, information factors corresponding to the corner suppression values are extracted, and based on the information factors, the corresponding remaining areas are identified texture characteristics; Based on the texture features, construct an image annotation model corresponding to the original MRI image, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to model the image annotation model Train to obtain the trained image annotation model; The trained model is used to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and the overlap of the predicted annotated image is removed to obtain an image annotation result corresponding to the original MRI image. 2. The data annotation method based on MRI images as claimed in claim 1, characterized in that said obtaining the original MRI image, performing image denoising on the original MRI image to obtain the MRI denoised image includes: Use preset medical equipment to perform an MRI scan on the patient and obtain the original MRI image; Perform noise analysis on the original MRI image to obtain noise data; Identify the noise direction corresponding to the noise data; Based on the noise direction, image denoising is performed on the original MRI image to obtain an MRI denoised image. 3. The data annotation method based on MRI images as claimed in claim 1, characterized in that, smoothing the MRI denoised image to obtain an MRI smoothed image includes: Identify image requirements corresponding to the MRI denoised image; Based on the image requirements, set the smoothing parameters corresponding to the MRI denoising image; Based on the smoothing parameters, a preset smoothing filter is enabled to smooth the MRI denoised image to obtain an MRI smoothed image. 4. The data annotation method based on MRI images as claimed in claim 1, wherein the segmenting the target area in the MRI smooth image and extracting the regional parameters corresponding to the target area includes: Identify the target factor corresponding to the MIR smooth image; Based on the target factor, determine the target area corresponding to the MRI smooth image; Segment the target area in the MRI smooth image using a preset segmentation algorithm; Query the regional requirements corresponding to the target area, and extract the regional parameters corresponding to the regional requirements. 5. The data annotation method based on MRI images as claimed in claim 1, characterized in that, based on the region parameters, calculating the image gradient corresponding to the target region includes: First, use the following formula to calculate the average gray value corresponding to the target area: Wherein, MG represents the average gray value corresponding to the target area, N represents the number of pixels in the target area, and I(x,y) represents the gray value of the pixel at the coordinate (x, y). Secondly, based on the average gray value, calculate the sum of squared differences between each pixel in the target area and its neighbor pixels: SD=∑[I(x,y)-I(x',y')] ² Among them, SD represents the sum of squared differences between each pixel in the target area and its neighboring pixels, (x', y') represents the neighborhood pixel position of the (x, y) pixel, I(x, y) and I( x', y') respectively represent the pixel gray value in the target area. Finally, based on the sum of squared differences, the image gradient corresponding to the target area is calculated: Where, GT represents the image gradient corresponding to the target area, and N represents the number of pixels in the target area. 6. The method for data annotation based on MRI images according to claim 1, wherein the step of calculating the edge strength of the target area based on the gradient factor comprises: Among them, BQ represents the edge strength of the target area, (i,j) represents the pixel coordinates in the image, MB represents the target area, Gx(i,j) and Gy(i,j) respectively represent the pixel point (i , the gradient values in the horizontal and vertical directions at j), M represents the total number of pixels in the target area. 7. The data annotation method based on MRI images according to claim 1, characterized in that, based on the edge intensity, corner detection is performed on the remaining area of the MRI smooth image to obtain a corner suppression value, including : Based on the edge intensity, determine the edge image corresponding to the remaining area; Perform corner point monitoring on the edge image to obtain corner point parameters; Based on the corner point parameters, calculating the corner point response values corresponding to the remaining area; Perform non-maximum suppression on the corner point response value to obtain a corner point suppression value. 8. The data annotation method based on MRI images according to claim 1, characterized in that: extracting information factors corresponding to the corner point suppression values, and identifying texture features corresponding to the remaining areas based on the information factors. ,include: Determine the fixed neighborhood corresponding to the corner point suppression value, and extract the information factor corresponding to the fixed neighborhood; Determine the texture category corresponding to the information factor; Based on the texture category, texture features corresponding to the remaining area are identified. 9. The data annotation method based on MRI images as claimed in claim 1, characterized in that the use of a trained model to perform annotation prediction on the original MRI image to obtain a predicted annotation image includes: Obtain MRI image samples corresponding to the original MRI images; Perform equalization processing on the MRI image samples to obtain equalized samples; Use the trained model to perform image prediction on the balanced sample to obtain a predicted image; Image annotation is performed on the predicted image to obtain a predicted annotated image. 10. A data annotation system based on MRI images, characterized in that it is used to perform the data annotation method based on MRI images as described in any one of claims 1-9, and the system includes: A smoothing processing module, used to obtain an original MRI image, perform image denoising on the original MRI image to obtain an MRI denoised image, and perform smoothing processing on the MRI denoised image to obtain an MRI smoothed image; An edge calculation module is used to segment the target area in the MRI smooth image, extract the area parameters corresponding to the target area, calculate the image gradient corresponding to the target area based on the area parameters, and query the image gradient corresponding to the Gradient factor, based on the gradient factor, calculate the edge strength of the target area; A texture recognition module, configured to perform corner point detection on the remaining area of the MRI smooth image based on the edge intensity, obtain a corner point suppression value, and extract an information factor corresponding to the corner point suppression value. Based on the information factor, Identify the texture features corresponding to the remaining area; A model training module, configured to construct an image annotation model corresponding to the original MRI image based on the texture features, use the image annotation model to divide the image training set corresponding to the MRI smooth image, and use the training set to The above image annotation model is used for model training, and a trained image annotation model is obtained; The image annotation module is used to use the trained model to perform annotation prediction on the original MRI image to obtain a predicted annotated image, and to remove overlaps on the predicted annotated image to obtain an image annotation result corresponding to the original MRI image.