A method, system and storage medium for vascular tissue segmentation based on MR images

Abstract

The present invention relates to the field of medical image processing technology, and more specifically to a method, system, and storage medium for vascular tissue segmentation based on MR images. The method comprises the following steps: enhancing vascular information in MR images using a Canny operator to obtain an MR-enhanced image; constructing a first segmentation model and a second segmentation model based on the MR image and the MR-enhanced image, respectively, using a UNet network; guiding the training of the first segmentation model using the second segmentation model, and using the output of the trained first segmentation model as the vascular tissue segmentation result in the MR image. The present invention can guide the segmentation results on low-resolution images to learn from the segmentation results on high-resolution images, thereby obtaining high-precision segmentation results directly on low-resolution images without increasing the data processing process and model complexity of the segmentation process, thereby improving the efficiency of accurate vascular tissue segmentation.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a vascular tissue segmentation method, system and storage medium based on MR images. Background Art

[0002] Vessel segmentation aims to automatically segment vascular tissue regions from multimodal magnetic resonance (MR) images captured by advanced medical imaging equipment. By segmenting blood vessels, we can provide vascular morphology and localization, which plays a vital role in disease diagnosis and monitoring.

[0003] In the existing technology, deep learning algorithms are usually used to segment vascular tissue in MR images. Such vascular tissue segmentation models based on deep learning usually enhance the features of vascular tissue in MR images, improve the image resolution, and then perform segmentation to obtain high-precision segmentation results. Therefore, in the existing technology, high-precision segmentation results need to be based on high-resolution images, and it is difficult to obtain high-precision segmentation results directly on low-resolution images. In order to obtain the same high-precision segmentation results, corresponding resolution improvement processing is required, which correspondingly increases the data processing process and model complexity of the segmentation process, limiting the efficiency of accurate segmentation of vascular tissue. Summary of the Invention

[0004] The purpose of the present invention is to provide a vascular tissue segmentation method, system and storage medium based on MR images to solve the technical problem in the prior art that it is difficult to obtain high-precision resolution results directly on low-resolution images, which limits the efficiency of accurate vascular tissue segmentation.

[0005] In order to solve the above technical problems, the present invention specifically provides the following technical solutions:

[0006] A vascular tissue segmentation method based on MR images comprises the following steps:

[0007] Acquiring MR images containing vascular tissue;

[0008] The Canny operator is used to enhance the vascular information in the MR image to obtain an MR enhanced image;

[0009] Using a UNet network, based on the MR image and the MR enhanced image, respectively construct a first segmentation model for segmenting vascular tissue on the MR image and a second segmentation model for segmenting vascular tissue on the MR enhanced image;

[0010] The second segmentation model is used to guide the first segmentation model for training, and the output result of the first segmentation model after training is used as the vascular tissue segmentation result in the MR image.

[0011] As a preferred embodiment of the present invention, the MR image enhancement method includes:

[0012] The Canny operator is used to perform edge detection on the MR image to obtain an edge detection image of the MR image;

[0013] The edge detection image and the MR image are superimposed and fused to obtain the MR enhanced image.

[0014] As a preferred solution of the present invention, the method for constructing the first segmentation model includes:

[0015] Use the UNet network as the network structure of the first segmentation model;

[0016] Using the MR image as the input of the UNet network and the vascular tissue mask annotation data in the MR image as the output of the UNet network to construct the first segmentation model;

[0017] The first segmentation model is:

[0018] ;

[0019] Where, is the vascular tissue mask of the MR image, is the MR image, and UNet is the UNet network.

[0020] As a preferred embodiment of the present invention, the method for constructing the second segmentation model includes:

[0021] Use the UNet network as the network structure of the second segmentation model;

[0022] Using the MR enhanced image as the input of the UNet network and the vascular tissue mask annotation data in the MR enhanced image as the output of the UNet network to construct the second segmentation model;

[0023] The second segmentation model is:

[0024] ;

[0025] Where, is the vascular tissue mask of the MR enhanced image, is the MR enhanced image, and UNet is the UNet network.

[0026] As a preferred embodiment of the present invention, the method for constructing the loss function for training the first segmentation model includes:

[0027] The prediction loss for building the first segmentation model is:

[0028] ;

[0029] Where, To predict losses, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training data set, is the true value of the vascular tissue mask of the i-th MR image sample in the training dataset, and m is the total number of MR image samples in the training dataset;

[0030] The guided loss of the second segmentation model for the first segmentation model is:

[0031] ;

[0032] ;

[0033] Where, To guide the loss, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training data set, is the vascular tissue mask output by the second segmentation model of the MR enhanced image corresponding to the i-th MR image sample in the training data set, To guide the adjustment coefficient, it has high credibility and great guiding power. is the true value of the vascular tissue mask of the MR enhanced image corresponding to the i-th MR image sample in the training data set, 、 and All are L2 norm forms;

[0034] The loss function for training the first segmentation model is formed by combining the prediction loss and the guidance loss:

[0035] ;

[0036] Where, is the loss function used for training the first segmentation model.

[0037] As a preferred embodiment of the present invention, the training method of the first segmentation model includes:

[0038] Divide the training data set into training set and test set;

[0039] Based on the loss function, the UNet network is trained on the training set to obtain the first segmentation model;

[0040] Based on the evaluation indicators, the performance of the first segmentation model is evaluated on the test set.

[0041] In a preferred embodiment of the present invention, the vascular tissue mask annotation data in the MR enhanced image and the vascular tissue mask annotation data in the MR image are both obtained by manual annotation.

[0042] As a preferred solution of the present invention, the evaluation indicators include Dice indicator, IoU indicator, and MAE indicator.

[0043] As a preferred embodiment of the present invention, the present invention provides a vascular tissue segmentation system based on MR images, which is applied to a vascular tissue segmentation method based on MR images. The system includes:

[0044] a data acquisition unit, configured to acquire an MR image containing vascular tissue;

[0045] A data enhancement unit, configured to enhance the vascular information in the MR image using a Canny operator to obtain an MR enhanced image;

[0046] a model construction unit for constructing, using a UNet network, a first segmentation model for segmenting vascular tissue in the MR image and a second segmentation model for segmenting vascular tissue in the MR enhanced image based on the MR image and the MR enhanced image, respectively; using the second segmentation model to guide the training of the first segmentation model, and using the output result of the trained first segmentation model as the vascular tissue segmentation result in the MR image;

[0047] The segmentation output unit is used to obtain the vascular tissue segmentation result in the MR image using the first segmentation model after training.

[0048] As a preferred embodiment of the present invention, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions. When a processor executes the computer-executable instructions, a vascular tissue segmentation method based on MR images is implemented.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention respectively constructs a first segmentation model for segmenting vascular tissue on MR images and a second segmentation model for segmenting vascular tissue on MR enhanced images, and uses the second segmentation model to guide the training of the first segmentation model, which can guide the segmentation results on low-resolution images to learn from the segmentation results of high-resolution images, so that high-precision segmentation results can be obtained directly on low-resolution images without increasing the data processing process and model complexity of the segmentation process, thereby improving the efficiency of accurate segmentation of vascular tissue. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other implementation drawings based on the provided drawings without inventive effort.

[0052] Figure 1 A flow chart of a vascular tissue segmentation method based on MR images provided in an embodiment of the present invention;

[0053] Figure 2 This is a block diagram of a vascular tissue segmentation system based on MR images provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0055] like Figure 1 As shown, the present invention provides a vascular tissue segmentation method based on MR images, comprising the following steps:

[0056] Acquiring MR images containing vascular tissue;

[0057] The Canny operator is used to enhance the vascular information in the MR image to obtain an MR enhanced image;

[0058] Using a UNet network, based on the MR image and the MR enhanced image, respectively construct a first segmentation model for segmenting vascular tissue on the MR image and a second segmentation model for segmenting vascular tissue on the MR enhanced image;

[0059] The second segmentation model is used to guide the first segmentation model for training, and the output result of the first segmentation model after training is used as the vascular tissue segmentation result in the MR image.

[0060] The present invention first performs edge detection on the MR image using the Canny operator, thereby obtaining an edge detection image in the MR image. The linear attributes of the vascular tissue itself will be detected in the edge detection, and the overall contour of the vascular tissue will be obtained. The edge detection image containing the overall contour of the vascular tissue is superimposed on the original MR image, thereby achieving pixel enhancement of the vascular tissue in the original MR image and obtaining an MR enhanced image. Compared with the original MR image, the resolution of the vascular tissue is improved.

[0061] The present invention constructs a first segmentation model for segmenting vascular tissue on MR images to obtain vascular segmentation results on MR images, and constructs a second segmentation model for segmenting vascular tissue on MR enhanced images to obtain vascular segmentation results on MR images, wherein the first segmentation model performs vascular tissue segmentation on low-resolution MR images, and the second segmentation model performs vascular segmentation on high-resolution MR images. It can be seen that compared with the second segmentation model, the first segmentation model is a low-precision segmentation, and the second segmentation model is a high-precision segmentation. However, the first segmentation model is an efficient segmentation process, and the second segmentation model requires image enhancement first, which is an overall inefficient segmentation process.

[0062] The present invention uses the second segmentation model to guide the training of the first segmentation model, so that during the training process, the output of the first segmentation model approaches the output of the second segmentation model, that is, the low-precision segmentation approaches the high-precision segmentation. After the training is completed, the first segmentation model can obtain high-precision segmentation results on low-resolution MR images. Ultimately, the first segmentation model has both high precision and high efficiency.

[0063] In order to realize the second segmentation model to guide the training of the first segmentation model, the traditional loss function of the first segmentation model predicts the loss (i.e., L mse , quantifying the loss between the output of the first segmentation model and the true value), an additional guided loss part is added. The guided loss measures the degree of difference between the output of the first segmentation model and the output of the second segmentation model. Training with this loss can ensure that the degree of difference between the output of the first segmentation model and the output of the second segmentation model is minimized after training, so that the first segmentation model can achieve high-precision segmentation on low-resolution MR images.

[0064] Furthermore, the present invention also adds a guidance adjustment coefficient to the guidance loss part, which is used to control and adjust the training guidance progress of the second segmentation model to the first segmentation model. The guidance adjustment coefficient is related to the accuracy performance of the second segmentation model. When the accuracy performance of the second segmentation model is higher, it means that the credibility of the second segmentation model is higher, and the training guidance effect with the second segmentation model is better. In this case, a large guidance adjustment coefficient is assigned to increase the weight of the guidance loss part in the total loss function, and the training focus of the first segmentation model is placed on learning from the second segmentation model so that the first segmentation model obtains high-precision performance. Correspondingly, when the accuracy performance of the second segmentation model is low, it means that the credibility of the second segmentation model is low, and the training guidance effect with the second segmentation model is poor. In this case, a small guidance adjustment coefficient is assigned to reduce the weight of the guidance loss part in the total loss function, and the training focus of the first segmentation model is placed on maintaining the current prediction loss to a minimum, so that the first segmentation model maintains the current best accuracy performance, that is, the highest accuracy performance on MR images, to avoid being misguided in training, resulting in a reduction or loss of the existing accuracy (basic base).

[0065] Therefore, the present invention adds a guidance adjustment coefficient to the guidance loss part to achieve adaptive control of the guidance process and ensure effective guidance.

[0066] The present invention first performs edge detection on the MR image using the Canny operator, thereby obtaining an edge detection image in the MR image. The linear attributes of the vascular tissue itself are detected in the edge detection, and the overall contour of the vascular tissue is obtained. The edge detection image containing the overall contour of the vascular tissue is superimposed on the original MR image, thereby achieving pixel enhancement of the vascular tissue in the original MR image to obtain an MR enhanced image. Compared with the original MR image, the resolution of the vascular tissue is improved, as shown below:

[0067] MR image enhancement methods include:

[0068] The Canny operator is used to perform edge detection on the MR image to obtain an edge detection image of the MR image;

[0069] The edge detection image is superimposed and fused with the MR image to obtain the MR enhanced image.

[0070] The present invention constructs a first segmentation model for segmenting vascular tissue on MR images, and obtains vascular segmentation results on MR images, which are specifically as follows:

[0071] The method for constructing the first segmentation model includes:

[0072] Use the UNet network as the network structure of the first segmentation model;

[0073] The MR image is used as the input of the UNet network, and the vascular tissue mask annotation data in the MR image is used as the output of the UNet network to construct the first segmentation model;

[0074] The first segmentation model is:

[0075] ;

[0076] Where, is the vascular tissue mask of the MR image, is the MR image, and UNet is the UNet network.

[0077] The present invention constructs a second segmentation model for segmenting vascular tissue on MR enhanced images, and obtains vascular segmentation results on MR images, which are as follows:

[0078] The method for constructing the second segmentation model includes:

[0079] Use the UNet network as the network structure of the second segmentation model;

[0080] The MR enhanced image is used as the input of the UNet network, and the vascular tissue mask annotation data in the MR enhanced image is used as the output of the UNet network to construct a second segmentation model;

[0081] The second segmentation model is:

[0082] ;

[0083] Where, is the vascular tissue mask of the MR enhanced image, is the MR enhanced image, and UNet is the UNet network.

[0084] The present invention uses a second segmentation model to guide the training of the first segmentation model, so that during the training process, the output of the first segmentation model approaches the output of the second segmentation model, that is, the low-precision segmentation approaches the high-precision segmentation. After the training is completed, the first segmentation model can obtain high-precision segmentation results on low-resolution MR images. Ultimately, the first segmentation model has both high precision and high efficiency. In order to achieve the second segmentation model to guide the training of the first segmentation model, the present invention uses the traditional loss function prediction loss (i.e., L mse , which quantifies the loss between the output of the first segmentation model and the true value, adds an additional guided loss part as follows:

[0085] The loss function methods for constructing the first segmentation model training include:

[0086] The prediction loss for building the first segmentation model is:

[0087] ;

[0088] Where, To predict losses, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training dataset, is the true value of the vascular tissue mask of the i-th MR image sample in the training dataset, and m is the total number of MR image samples in the training dataset;

[0089] The guided loss of the second segmentation model for the first segmentation model is:

[0090] ;

[0091] ;

[0092] Where, To guide the loss, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training dataset, is the vascular tissue mask output by the second segmentation model of the MR enhanced image corresponding to the i-th MR image sample in the training data set, is the boot adjustment coefficient, is the true value of the vascular tissue mask of the MR enhanced image corresponding to the i-th MR image sample in the training data set, 、 and All are L2 norm forms;

[0093] The loss function for training the first segmentation model is composed of the prediction loss and the guidance loss:

[0094] ;

[0095] Where, is the loss function used for training the first segmentation model.

[0096] The present invention also adds a guidance adjustment coefficient to the guidance loss part, which is used to control and adjust the training guidance progress of the second segmentation model to the first segmentation model. The guidance adjustment coefficient is related to the accuracy performance of the second segmentation model. When the accuracy performance of the second segmentation model is higher, the credibility of the second segmentation model is higher, and the training guidance effect with the second segmentation model is better. A larger guidance adjustment coefficient is assigned to make the weight of the guidance loss part in the total loss function larger, and the training focus of the first segmentation model is placed on learning from the second segmentation model, so that the first segmentation model obtains high-precision performance.

[0097] Correspondingly, when the accuracy performance of the second segmentation model is low, it means that the credibility of the second segmentation model is low, and the training guidance effect of the second segmentation model is poor. In this case, a small guidance adjustment coefficient is given to make the weight of the guidance loss part in the total loss function smaller, and the training focus of the first segmentation model is placed on maintaining the current prediction loss to the minimum, so that the first segmentation model maintains the current best accuracy performance, that is, the highest accuracy performance on MR images, to avoid being misguided in training, resulting in a reduction or loss of the existing accuracy (basic plate).

[0098] Therefore, the present invention adds a guidance adjustment coefficient to the guidance loss part to achieve adaptive control of the guidance process and ensure effective guidance.

[0099] The training method of the first segmentation model includes:

[0100] Divide the training data set into training set and test set;

[0101] Based on the loss function, the UNet network is trained on the training set to obtain the first segmentation model;

[0102] Based on the evaluation indicators, the performance of the first segmentation model is evaluated on the test set.

[0103] The vascular tissue mask annotation data in the MR enhanced image and the vascular tissue mask annotation data in the MR image are both obtained by manual annotation.

[0104] Evaluation indicators include Dice indicator, IoU indicator, and MAE indicator.

[0105] like Figure 2 As shown, the present invention provides a vascular tissue segmentation system based on MR images, which is applied to a vascular tissue segmentation method based on MR images. The system includes:

[0106] a data acquisition unit, configured to acquire an MR image containing vascular tissue;

[0107] A data enhancement unit, configured to enhance the vascular information in the MR image using a Canny operator to obtain an MR enhanced image;

[0108] a model construction unit for constructing, using a UNet network, a first segmentation model for segmenting vascular tissue in the MR image and a second segmentation model for segmenting vascular tissue in the MR enhanced image based on the MR image and the MR enhanced image, respectively; using the second segmentation model to guide the training of the first segmentation model, and using the output result of the trained first segmentation model as the vascular tissue segmentation result in the MR image;

[0109] The segmentation output unit is used to obtain the vascular tissue segmentation result in the MR image using the first segmentation model after training.

[0110] The present invention provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, a vascular tissue segmentation method based on MR images is implemented.

[0111] The present invention respectively constructs a first segmentation model for segmenting vascular tissue on MR images and a second segmentation model for segmenting vascular tissue on MR enhanced images, and uses the second segmentation model to guide the training of the first segmentation model, which can guide the segmentation results on low-resolution images to learn from the segmentation results of high-resolution images, so that high-precision segmentation results can be obtained directly on low-resolution images without increasing the data processing process and model complexity of the segmentation process, thereby improving the efficiency of accurate segmentation of vascular tissue.

[0112] The above embodiments are merely exemplary embodiments of the present application and are not intended to limit the scope of the present application. The scope of protection of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and scope of protection of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the scope of protection of the present application.

Claims

1. A vascular tissue segmentation method based on MR images, characterized in that: The following steps are involved: Acquiring MR images containing vascular tissue; The Canny operator is used to enhance the vascular information in the MR image to obtain an MR enhanced image; Using a UNet network, based on the MR image and the MR enhanced image, respectively construct a first segmentation model for segmenting vascular tissue on the MR image and a second segmentation model for segmenting vascular tissue on the MR enhanced image; Using the second segmentation model to guide the first segmentation model for training, and using the output result of the trained first segmentation model as the vascular tissue segmentation result in the MR image; The method for constructing the first segmentation model includes: Use the UNet network as the network structure of the first segmentation model; Using the MR image as the input of the UNet network and the vascular tissue mask annotation data in the MR image as the output of the UNet network to construct the first segmentation model; The first segmentation model is: ； Where, is the vascular tissue mask of the MR image, is the MR image, UNet is the UNet network; The method for constructing the second segmentation model includes: Use the UNet network as the network structure of the second segmentation model; Using the MR enhanced image as the input of the UNet network and the vascular tissue mask annotation data in the MR enhanced image as the output of the UNet network to construct the second segmentation model; The second segmentation model is: ； Where, is the vascular tissue mask of the MR enhanced image, is the MR enhanced image, UNet is the UNet network; The loss function methods for constructing the first segmentation model training include: The prediction loss for building the first segmentation model is: ； Where, To predict losses, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training data set, is the true value of the vascular tissue mask of the i-th MR image sample in the training dataset, and m is the total number of MR image samples in the training dataset; The guided loss of the second segmentation model for the first segmentation model is: ； ； Where, To guide the loss, is the vascular tissue mask output by the first segmentation model for the i-th MR image sample in the training data set, is the vascular tissue mask output by the second segmentation model of the MR enhanced image corresponding to the i-th MR image sample in the training data set, To guide the adjustment coefficient, it has high credibility and great guiding power. is the true value of the vascular tissue mask of the MR enhanced image corresponding to the i-th MR image sample in the training data set, 、 and All are L2 norm forms; The loss function for training the first segmentation model is formed by combining the prediction loss and the guidance loss: ； Where, is the loss function used for training the first segmentation model.

2. The vascular tissue segmentation method based on MR images according to claim 1, characterized in that: The MR image enhancement method includes: The Canny operator is used to perform edge detection on the MR image to obtain an edge detection image of the MR image; The edge detection image and the MR image are superimposed and fused to obtain the MR enhanced image.

3. The vascular tissue segmentation method based on MR images according to claim 1, characterized in that: The training method of the first segmentation model includes: Divide the training data set into training set and test set; Based on the loss function, the UNet network is trained on the training set to obtain the first segmentation model; Based on the evaluation indicators, the performance of the first segmentation model is evaluated on the test set.

4. The vascular tissue segmentation method based on MR images according to claim 3, characterized in that: The vascular tissue mask annotation data in the MR enhanced image and the vascular tissue mask annotation data in the MR image are both obtained by manual annotation.

5. The vascular tissue segmentation method based on MR images according to claim 3, characterized in that: The evaluation indicators include Dice indicator, IoU indicator, and MAE indicator.

6. A vascular tissue segmentation system based on MR images, characterized in that: A method for segmenting vascular tissue based on MR images as described in any one of claims 1 to 5, wherein the system comprises: a data acquisition unit, configured to acquire an MR image containing vascular tissue; A data enhancement unit, configured to enhance the vascular information in the MR image using a Canny operator to obtain an MR enhanced image; a model construction unit for constructing, using a UNet network, a first segmentation model for segmenting vascular tissue in the MR image and a second segmentation model for segmenting vascular tissue in the MR enhanced image based on the MR image and the MR enhanced image, respectively; using the second segmentation model to guide the training of the first segmentation model, and using the output result of the trained first segmentation model as the vascular tissue segmentation result in the MR image; The segmentation output unit is used to obtain the vascular tissue segmentation result in the MR image using the first segmentation model after training.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, and when a processor executes the computer-executable instructions, the method according to any one of claims 1 to 5 is implemented.

Images