CN110211166B - Optic nerve dividing method and device in magnetic resonance image - Google Patents
Optic nerve dividing method and device in magnetic resonance image Download PDFInfo
- Publication number
- CN110211166B CN110211166B CN201910509876.1A CN201910509876A CN110211166B CN 110211166 B CN110211166 B CN 110211166B CN 201910509876 A CN201910509876 A CN 201910509876A CN 110211166 B CN110211166 B CN 110211166B
- Authority
- CN
- China
- Prior art keywords
- magnetic resonance
- probability distribution
- resonance image
- visual
- distribution information
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10084—Hybrid tomography; Concurrent acquisition with multiple different tomographic modalities
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30041—Eye; Retina; Ophthalmic
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a method and a device for optic nerve segmentation in a magnetic resonance image, which comprises the following steps: carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage; predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information. According to the embodiment of the invention, the spatial probability distribution information of the visual passage in the magnetic resonance image is acquired, and the visual passage in the magnetic resonance image is segmented according to the shape information and the position information of the spatial probability distribution information, so that the problem of fuzzy boundary of the visual passage is effectively overcome, and the accurate segmentation of the visual passage is realized.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a device for optic nerve segmentation in a magnetic resonance image.
Background
Magnetic Resonance (MR) imaging is an imaging technique for image reconstruction by using signals generated by the Resonance of atomic nuclei in a strong Magnetic field, and nowadays, image-guided surgical navigation based on Magnetic Resonance images has become an important auxiliary mode for sinus and skull base surgical operations.
Due to the complex anatomy of the paranasal sinuses, the skull base and the adjacent orbital region, the anatomy of the paranasal sinuses, the skull base and the adjacent orbital region contains important nerve and blood vessel structures, such as a visual pathway, the slender anatomy of the visual pathway, the low contrast and the fuzzy boundary presented in the magnetic resonance image, the target segmentation method based on threshold or edge detection cannot achieve the effect at all, and the existing segmentation method of the visual pathway has low accuracy due to the fact that the visual pathway occupies a small proportion of the whole head volume data.
Therefore, there is a need for a method and apparatus for optic nerve segmentation in magnetic resonance images to solve the above problems.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a method and a device for optic nerve segmentation in a magnetic resonance image.
In a first aspect, an embodiment of the present invention provides a method for optic nerve segmentation in a magnetic resonance image, including:
carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage;
predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
In a second aspect, an embodiment of the present invention provides an apparatus for optic nerve segmentation in a magnetic resonance image, including:
the image registration module is used for carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage;
the visual path acquisition module is used for predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to acquire a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the method provided in the first aspect when executing the program.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the method as provided in the first aspect.
According to the method and the device for segmenting the optic nerve in the magnetic resonance image, provided by the embodiment of the invention, the visual passage in the magnetic resonance image is segmented by acquiring the spatial probability distribution information of the visual passage in the magnetic resonance image and according to the shape information and the position information of the spatial probability distribution information, so that the problem of fuzzy boundary of the visual passage is effectively overcome, and the accurate segmentation of the visual passage is realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic flowchart of a method for optic nerve segmentation in a magnetic resonance image according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a 3D full convolution neural network according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an optic nerve segmentation apparatus in a magnetic resonance image according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The current visual pathway segmentation methods are mainly classified into a graph-based segmentation method, a statistical model-based segmentation method, and a learning-based segmentation method. The atlas-based method generally includes calculating a deformation field between a reference atlas and an image to be segmented through image registration, and then transmitting labels in the atlas to the image to be segmented through the deformation field to complete segmentation of a target image. The multi-atlas based segmentation is to obtain a plurality of labels after projection by using an atlas, and then obtain a final segmentation result by using a label fusion strategy. Experimental results show that the map-based approach is not robust. The statistical model-based segmentation method generates a final segmented surface mesh by incorporating shape or appearance model prior information and performing iterative optimization according to image information and shape information, wherein an initial contour may need to be manually selected. The learning-based segmentation method trains a classifier by extracting a feature vector of a voxel so as to realize classification at a voxel level, features with discrimination and the performance of the classifier are key factors influencing the result of the learning-based segmentation method, but the accuracy of the obtained segmentation result is low due to low contrast of a visual pathway and surrounding tissues.
Fig. 1 is a schematic flow chart of a method for segmenting an optic nerve in a magnetic resonance image according to an embodiment of the present invention, and as shown in fig. 1, the embodiment of the present invention provides a method for segmenting an optic nerve in a magnetic resonance image, including:
In the embodiment Of the present invention, before segmenting the visual pathway on the target magnetic resonance image, image registration is performed on the target magnetic resonance image to obtain a spatial probability distribution map about the visual pathway in the target magnetic resonance image, and shape information and position information Of the visual pathway, that is, spatial probability distribution information, are obtained according to the spatial probability distribution map, so as to position a Region Of Interest (ROI) Of the visual pathway. In an embodiment of the invention, the localization of the visual pathway ROI is achieved by computing the starting and ending slice positions of the spatial probability distribution image in the transverse, coronal and sagittal planes and dilating them outward to form a bounding box of size 90 x 54 x 90.
102, predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
In the embodiment of the invention, the target magnetic resonance image and the spatial probability distribution information corresponding to the target magnetic resonance image are input into a trained visual pathway segmentation model, the visual pathway segmentation model carries out prediction identification on the visual pathway on the target magnetic resonance image according to the shape information and the position information of the visual pathway in the spatial probability distribution information, and the visual pathway image is segmented on the target magnetic resonance image by adding a label to the identified visual pathway region.
According to the optic nerve segmentation method in the magnetic resonance image, provided by the embodiment of the invention, the visual passage in the magnetic resonance image is segmented by acquiring the spatial probability distribution information of the visual passage in the magnetic resonance image and according to the shape information and the position information of the spatial probability distribution information, so that the problem of fuzzy boundary of the visual passage is effectively overcome, and the accurate segmentation of the visual passage is realized.
On the basis of the above embodiment, the trained visual pathway segmentation model is obtained by training through the following steps:
adding a label to a visual passage on a sample magnetic resonance image to obtain a first sample training set;
carrying out image registration on the first sample training set to obtain sample space probability distribution information;
and training a visual access segmentation model according to the first sample training set and the sample spatial probability distribution information to obtain the trained visual access segmentation model.
In the embodiment of the invention, a visual access region on a sample magnetic resonance image is labeled, the labeled sample magnetic resonance image is used as a first magnetic resonance sample image to construct a first sample training set, then a spatial probability distribution map is constructed according to the first magnetic resonance image in the first sample training set to obtain spatial probability distribution information corresponding to the first magnetic resonance image, and finally the first sample training set and the corresponding spatial probability distribution information are input into a visual access segmentation model to be trained to obtain the trained visual access segmentation model. It should be noted that, in the embodiment of the present invention, the network weight of the visual pathway segmentation model is initialized through Xavier initialization, and is optimized through adaptive moment estimation (Adam) algorithm, and in addition, when the model is trained, a learning rate attenuation strategy is adopted for training, a larger learning rate is used in an early stage of training to accelerate convergence, and a smaller learning rate is used in a later stage to ensure stability.
Further, in the embodiment of the present invention, after the trained visual pathway segmentation model is obtained, in the test stage, the spatial probability distribution map is registered to the magnetic resonance image to be segmented, so as to obtain the spatial probability distribution map corresponding to the magnetic resonance image to be segmented, and thus the positioning of the visual pathway ROI is completed. And then, cutting the magnetic resonance image to be segmented and the corresponding spatial probability distribution information into sub-blocks, sending the sub-blocks into a visual path segmentation model for prediction, and splicing the obtained prediction result sub-blocks in sequence to obtain a final segmentation result.
On the basis of the foregoing embodiment, the performing image registration on the first sample training set to obtain sample spatial probability distribution information includes:
aligning the first magnetic resonance image of the first sample training set to a preset reference image to obtain a second magnetic resonance image and a corresponding deformation field;
converting the label on the first magnetic resonance image into a reference space according to the deformation field to obtain a label corresponding to the second magnetic resonance image;
and carrying out summation and average processing on the labels of the second magnetic resonance image to obtain sample space probability distribution information.
In an embodiment of the invention, the probability that a voxel at a certain position in the magnetic resonance image belongs to the visual path is reflected by the spatial probability distribution map. In the training stage, firstly, images in a first sample training set are registered, the images in the first sample training set are aligned to a preset reference image, labels corresponding to the images in the first sample training set are converted to a reference space according to an obtained deformation field, and after the image registration is completed, all the labels after the conversion are summed and averaged to obtain a spatial probability distribution map.
On the basis of the embodiment, the visual pathway segmentation model is constructed by a 3D full convolution neural network.
Further, in addition to the above embodiments, the 3D full convolution neural network is composed of three layers of convolution layers with convolution kernel sizes of 7 × 7 × 7 and four layers of convolution layers with convolution kernel sizes of 1 × 1 × 1.
In an embodiment of the present invention, fig. 2 is a schematic structural diagram of a 3D full convolution neural network provided in an embodiment of the present invention, and as shown in fig. 2, the 3D full convolution neural network is composed of three convolutional layers with convolution kernel sizes of 7 × 7 × 7 and four convolutional layers with convolution kernel sizes of 1 × 1 × 1. Preferably, in the embodiment of the present invention, in order to obtain a deeper network to extract more discriminative features, the embodiment of the present invention uses three layers of consecutive convolution kernels of 3 × 3 × 3 size instead of convolution kernels of 7 × 7 × 7 size, both of which have the same size receptive field, and connects a prilu nonlinear activation layer behind each convolution layer, wherein the last convolution layer uses a Softmax loss function, and specifically, the loss function uses cross entropy in the embodiment of the present invention. In addition, in order to utilize different levels of information, the embodiment of the invention connects the feature maps obtained at different stages for final classification.
On the basis of the foregoing embodiment, the training a visual pathway segmentation model according to the first sample training set and the sample spatial probability distribution information to obtain a trained visual pathway segmentation model includes:
cutting the first sample training set and the sample spatial probability distribution information into subblocks, and training a visual access segmentation model to obtain a sample prediction result subblock;
and splicing according to the sequence of the sample prediction result sub-blocks to obtain a sample visual path segmentation image so as to obtain a trained visual path segmentation model.
In the embodiment of the invention, each image in the first sample training set and the corresponding spatial probability distribution information are used as two channels input by the 3D full convolution neural network and input into the 3D full convolution neural network for training and learning. Specifically, in the embodiment of the present invention, each image in the first sample training set and the corresponding spatial probability distribution information are cut into three-dimensional sub-blocks, the three-dimensional sub-blocks with the size of 27 × 27 × 27 are input, the label values with the size of 9 × 9 × 9 at the center of the sub-blocks are output, and then the output sub-blocks are sequentially spliced, so as to obtain the sample visual pathway segmentation image. In addition, in order to overcome the problem of class imbalance, the embodiment of the invention selects the three-dimensional sub-blocks containing the visual channels, and adds the same number of three-dimensional sub-blocks containing only the background for training.
On the basis of the foregoing embodiment, the training a visual pathway segmentation model according to the first sample training set and the sample spatial probability distribution information to obtain a trained visual pathway segmentation model further includes:
and training a visual access segmentation model according to the first sample training set and the sample spatial probability distribution information based on an early-stopping strategy to obtain the trained visual access segmentation model.
In the embodiment of the invention, in order to prevent overfitting, an Early Stopping (Early Stopping) strategy is adopted to train the 3D full convolution neural network, the performance of the 3D full convolution neural network model on the verification set is calculated, and when the performance of the model on the verification set does not decrease any more, the training is stopped, so that the problem of overfitting caused by continuous training can be avoided. Specifically, in the embodiment of the present invention, namely when the loss function of the verification set does not decrease after 10 epochs, the training is terminated, so as to obtain a trained visual pathway segmentation model.
Fig. 3 is a schematic structural diagram of an optic nerve segmentation apparatus in a magnetic resonance image according to an embodiment of the present invention, and as shown in fig. 3, an embodiment of the present invention provides an optic nerve segmentation apparatus in a magnetic resonance image, including an image registration module 301 and a visual pathway acquisition module 302, where the image registration module 301 is configured to perform image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, where the spatial probability distribution information includes shape information and position information of a visual pathway; the visual path obtaining module 302 is configured to predict the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model, and obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
In the embodiment Of the present invention, before segmenting the visual pathway on the target magnetic resonance image, the image registration module 301 performs image registration on the target magnetic resonance image to obtain a spatial probability distribution map about the visual pathway in the target magnetic resonance image, and obtains shape information and position information Of the visual pathway, that is, spatial probability distribution information, according to the spatial probability distribution map, so as to position a Region Of Interest (ROI) Of the visual pathway. Then, the visual pathway acquisition module 302 performs predictive identification on the visual pathway on the target magnetic resonance image according to the shape information and the position information of the visual pathway in the spatial probability distribution information, and adds a label to the identified visual pathway region, thereby segmenting the visual pathway image on the target magnetic resonance image.
According to the optic nerve segmentation device in the magnetic resonance image, provided by the embodiment of the invention, the visual passage in the magnetic resonance image is segmented by acquiring the spatial probability distribution information of the visual passage in the magnetic resonance image and according to the shape information and the position information of the spatial probability distribution information, so that the problem of fuzzy boundary of the visual passage is effectively overcome, and the accurate segmentation of the visual passage is realized.
The apparatus provided in the embodiment of the present invention is used for executing the above method embodiments, and for details of the process and the details, reference is made to the above embodiments, which are not described herein again.
Fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 4, the electronic device may include: a Processor (Processor)401, a communication Interface (communication Interface)402, a Memory (Memory)403 and a communication bus 404, wherein the Processor 401, the communication Interface 402 and the Memory 403 complete communication with each other through the communication bus 404. Processor 401 may call logic instructions in memory 403 to perform the following method: carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage; predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
In addition, the logic instructions in the memory 403 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
An embodiment of the present invention discloses a computer program product, which includes a computer program stored on a non-transitory computer readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer can execute the methods provided by the above method embodiments, for example, the method includes: carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage; predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
An embodiment of the present invention provides a non-transitory computer-readable storage medium storing server instructions, where the server instructions cause a computer to execute the method for optic nerve segmentation in magnetic resonance images provided in the foregoing embodiment, for example, the method includes: carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage; predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for optic nerve segmentation in a magnetic resonance image is characterized by comprising the following steps:
carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage;
predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to obtain a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
2. The method of optic nerve segmentation in magnetic resonance images as set forth in claim 1, wherein the trained visual pathway segmentation model is trained by the following steps:
adding a label to a visual passage on a sample magnetic resonance image to obtain a first sample training set;
carrying out image registration on the first sample training set to obtain sample space probability distribution information;
and training a visual access segmentation model according to the first sample training set and the sample spatial probability distribution information to obtain the trained visual access segmentation model.
3. The method of optic nerve segmentation in magnetic resonance images as set forth in claim 2, wherein the image registration of the first training set of samples to obtain the spatial probability distribution information of the samples includes:
aligning the first magnetic resonance image of the first sample training set to a preset reference image to obtain a second magnetic resonance image and a corresponding deformation field;
converting the label on the first magnetic resonance image into a reference space according to the deformation field to obtain a label corresponding to the second magnetic resonance image;
and carrying out summation and average processing on the labels of the second magnetic resonance image to obtain sample space probability distribution information.
4. The method of claim 2, wherein the visual pathway segmentation model is constructed by a 3D full convolution neural network.
5. An optic nerve segmentation method in magnetic resonance images as claimed in claim 4, characterized in that the 3D full convolution neural network is composed of three layers of convolution layers with convolution kernel size of 7 x 7 and four layers of convolution layers with convolution kernel size of 1 x 1.
6. The method of claim 2, wherein the training a visual pathway segmentation model according to the first training set of samples and the spatial probability distribution information of the samples to obtain a trained visual pathway segmentation model comprises:
cutting the first sample training set and the sample spatial probability distribution information into a plurality of subblocks, and training a visual access segmentation model to obtain a sample prediction result subblock;
and splicing according to the sequence of the sample prediction result sub-blocks to obtain a sample visual path segmentation image so as to obtain a trained visual path segmentation model.
7. The method of claim 2, wherein the training a visual pathway segmentation model according to the first training set of samples and the spatial probability distribution information of the samples to obtain a trained visual pathway segmentation model, further comprises:
and training a visual access segmentation model according to the first sample training set and the sample spatial probability distribution information based on an early-stopping strategy to obtain the trained visual access segmentation model.
8. An apparatus for optic nerve segmentation in magnetic resonance images, comprising:
the image registration module is used for carrying out image registration on a target magnetic resonance image to obtain spatial probability distribution information of the target magnetic resonance image, wherein the spatial probability distribution information comprises shape information and position information of a visual passage;
the visual path acquisition module is used for predicting the target magnetic resonance image and the spatial probability distribution information based on a trained visual path segmentation model to acquire a visual path segmentation image of the target magnetic resonance image; the trained visual passage segmentation model is obtained by training a sample magnetic resonance image and sample space probability distribution information.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method according to any of claims 1 to 7 are implemented when the processor executes the program.
10. A non-transitory computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910509876.1A CN110211166B (en) | 2019-06-13 | 2019-06-13 | Optic nerve dividing method and device in magnetic resonance image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910509876.1A CN110211166B (en) | 2019-06-13 | 2019-06-13 | Optic nerve dividing method and device in magnetic resonance image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110211166A CN110211166A (en) | 2019-09-06 |
| CN110211166B true CN110211166B (en) | 2021-10-12 |
Family
ID=67792370
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910509876.1A Active CN110211166B (en) | 2019-06-13 | 2019-06-13 | Optic nerve dividing method and device in magnetic resonance image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110211166B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738635A (en) * | 2019-09-11 | 2020-01-31 | 深圳先进技术研究院 | feature tracking method and device |
| CN111862022B (en) * | 2020-07-13 | 2022-08-16 | 广州柏视医疗科技有限公司 | Automatic delineation method for organs at risk in whole body multi-part radiotherapy |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090163800A1 (en) * | 2007-12-20 | 2009-06-25 | Siemens Corporate Research, Inc. | Tools and methods for visualization and motion compensation during electrophysiology procedures |
| CN107563996A (en) * | 2017-08-23 | 2018-01-09 | 东北大学 | A kind of new discus nervi optici dividing method and system |
| CN107886508A (en) * | 2017-11-23 | 2018-04-06 | 上海联影医疗科技有限公司 | Difference subtracts image method and medical image processing method and system |
| CN108492302A (en) * | 2018-03-26 | 2018-09-04 | 北京市商汤科技开发有限公司 | Nervous layer dividing method and device, electronic equipment, storage medium, program |
| CN109785334A (en) * | 2018-12-17 | 2019-05-21 | 深圳先进技术研究院 | Cardiac magnetic resonance images dividing method, device, terminal device and storage medium |
-
2019
- 2019-06-13 CN CN201910509876.1A patent/CN110211166B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090163800A1 (en) * | 2007-12-20 | 2009-06-25 | Siemens Corporate Research, Inc. | Tools and methods for visualization and motion compensation during electrophysiology procedures |
| CN107563996A (en) * | 2017-08-23 | 2018-01-09 | 东北大学 | A kind of new discus nervi optici dividing method and system |
| CN107886508A (en) * | 2017-11-23 | 2018-04-06 | 上海联影医疗科技有限公司 | Difference subtracts image method and medical image processing method and system |
| CN108492302A (en) * | 2018-03-26 | 2018-09-04 | 北京市商汤科技开发有限公司 | Nervous layer dividing method and device, electronic equipment, storage medium, program |
| CN109785334A (en) * | 2018-12-17 | 2019-05-21 | 深圳先进技术研究院 | Cardiac magnetic resonance images dividing method, device, terminal device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110211166A (en) | 2019-09-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110232383B (en) | Focus image recognition method and focus image recognition system based on deep learning model | |
| US9968257B1 (en) | Volumetric quantification of cardiovascular structures from medical imaging | |
| CN105574859B (en) | A kind of liver neoplasm dividing method and device based on CT images | |
| Zhang et al. | Intelligent scanning: Automated standard plane selection and biometric measurement of early gestational sac in routine ultrasound examination | |
| Wen et al. | A novel statistical cerebrovascular segmentation algorithm with particle swarm optimization | |
| US20230104173A1 (en) | Method and system for determining blood vessel information in an image | |
| CN108550156B (en) | Method and apparatus for automatic delineation of structural shapes for image-guided treatment planning | |
| CN111640120B (en) | Pancreas CT automatic segmentation method based on significance dense connection expansion convolution network | |
| CN102324109B (en) | Method for three-dimensionally segmenting insubstantial pulmonary nodule based on fuzzy membership model | |
| CN110298844B (en) | X-ray radiography image blood vessel segmentation and identification method and device | |
| CN111105424A (en) | Lymph node automatic delineation method and device | |
| US20100027865A1 (en) | Method and System for Brain Tumor Segmentation in 3D Magnetic Resonance Images | |
| RU2654199C1 (en) | Segmentation of human tissues in computer image | |
| CN109102511A (en) | A kind of cerebrovascular dividing method, system and electronic equipment | |
| CN110110727B (en) | Image segmentation method based on conditional random field and Bayes post-processing | |
| CN113096169B (en) | Non-rigid multimode medical image registration model establishing method and application thereof | |
| CN110211166B (en) | Optic nerve dividing method and device in magnetic resonance image | |
| US20230005140A1 (en) | Automated detection of tumors based on image processing | |
| CN116309571B (en) | Three-dimensional cerebrovascular segmentation method and device based on semi-supervised learning | |
| CN111932495B (en) | Medical image detection method, device and storage medium | |
| CN115222713A (en) | Method and device for calculating coronary artery calcium score and storage medium | |
| CN113763371A (en) | Pathological image cell nucleus segmentation method and device | |
| CN112419340A (en) | Generation method, application method and device of cerebrospinal fluid segmentation model | |
| CN113393445B (en) | Breast cancer image determination method and system | |
| CN112862785B (en) | CTA image data identification method, device and storage medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |