CN114004859A - Method and system for segmenting echocardiography left atrium map based on multi-view fusion network - Google Patents
Method and system for segmenting echocardiography left atrium map based on multi-view fusion network Download PDFInfo
- Publication number
- CN114004859A CN114004859A CN202111424991.2A CN202111424991A CN114004859A CN 114004859 A CN114004859 A CN 114004859A CN 202111424991 A CN202111424991 A CN 202111424991A CN 114004859 A CN114004859 A CN 114004859A
- Authority
- CN
- China
- Prior art keywords
- view
- chamber
- features
- feature
- apical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000004927 fusion Effects 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 30
- 210000005246 left atrium Anatomy 0.000 title claims abstract description 26
- 238000002592 echocardiography Methods 0.000 title claims abstract description 15
- 230000002776 aggregation Effects 0.000 claims abstract description 24
- 238000004220 aggregation Methods 0.000 claims abstract description 24
- 230000011218 segmentation Effects 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000003709 image segmentation Methods 0.000 claims abstract description 3
- 230000001746 atrial effect Effects 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000007499 fusion processing Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims 1
- 238000013507 mapping Methods 0.000 abstract 1
- 230000000747 cardiac effect Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 230000004217 heart function Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
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
- G06T7/174—Segmentation; Edge detection involving the use of two or more images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10132—Ultrasound image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30048—Heart; Cardiac
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention belongs to the technical field of medical image segmentation, and particularly relates to a method and a system for segmenting an echocardiography left atrium image based on a multi-view fusion network. Acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of an echocardiogram, and respectively preprocessing the medical images; respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for encoding; the feature-based aggregation reallocation module is used for performing spatial reallocation on the single feature obtained by the encoder to obtain the feature after spatial mapping of the three views; and performing decoding operation based on the single view characteristics subjected to characteristic aggregation reallocation processing to obtain the decoding characteristics of the three views respectively, and further performing multi-view characteristic aggregation reallocation to obtain an accurate segmentation result.
Description
Technical Field
The invention belongs to the technical field of medical image segmentation, and particularly relates to a method and a system for segmenting an echocardiography left atrium image based on a multi-view fusion network.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
In clinical applications, echocardiography is an important tool for physicians to judge heart conditions. In clinical treatment, the characteristics of the left atrial motion state in an echocardiogram and the like are the primary basis for doctors to diagnose heart diseases. By segmenting the left atrium, it has a significant role in the assessment of the overall cardiac function. The echocardiogram comprises complete left atrium information in three views of a two-chamber apical chamber, a three-chamber apical chamber and a four-chamber apical chamber, but because the positions detected by the ultrasonic probes are different, the shape of the left atrium in different chambers is different, and similar features exist in the structure of the left atrium among different views. Meanwhile, the echocardiogram also contains a lot of noise, and the inventor finds that the traditional segmentation algorithm cannot accurately segment the left atrium.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a method and a system for segmenting an echocardiography left atrium map based on a multi-view fusion network, which can fuse the characteristics of the left atrium in a plurality of views, automatically and simultaneously segment the left atrium in a cardiac apical two-chamber, a cardiac apical three-chamber and a cardiac apical four-chamber, and improve the segmentation efficiency.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a method for segmenting the left atrium of an echocardiography based on a multi-view fusion network, which comprises the following steps:
acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram, and respectively preprocessing the medical images;
respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for encoding;
calculating the single view characteristic after the three views are fused by carrying out multi-view characteristic aggregation reallocation on each characteristic obtained by the encoder;
and performing decoding operation based on the single view characteristics after the fusion processing to respectively obtain the decoding characteristics of the three views, and further performing multi-view characteristic aggregation redistribution to obtain an accurate segmentation result.
Further, the process of calculating the single view feature after the three views are fused through multi-view feature aggregation reallocation is as follows:
firstly, carrying out feature connection on three different views, then respectively carrying out Non-local operation on the connection features and the single view features to respectively obtain weights of the single view features, multiplying the weights obtained by calculation with the single view features and adding the weights and the original single features to obtain the redistributed single view features.
Further, the encoder uses a ResNet based network.
Further, the encoder cascades for a plurality of convolutional layers, downsamples, and active layers.
Further, the preprocessing includes a random flipping operation, a random clipping operation, and a resizing operation.
Further, the decoder is aware of the pyramid structure.
A second aspect of the present invention provides a system for segmenting a left atrial map of an echocardiogram based on a multi-view fusion network, which comprises:
the image preprocessing module is used for acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram and respectively preprocessing the medical images;
the feature coding module is used for respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for coding;
a multi-view feature fusion module, configured to perform feature aggregation and redistribution on the single view features obtained by the encoder, so that three view features are spatially aligned with each other;
and the characteristic decoding module is used for performing decoding operation on the basis of the aligned single view characteristics and a decoder corresponding to the encoder to respectively obtain the decoding characteristics of the three views, and further performing multi-view characteristic aggregation reallocation to obtain an accurate segmentation result.
A third aspect of the present invention provides a computer readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for segmenting a left atrial map of echocardiography based on a multi-view fusion network as described above.
A fourth aspect of the present invention provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method for segmenting a left atrial map of echocardiography based on a multi-view fusion network as described above.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides an echocardiogram left atrium segmentation method and system based on a depth multi-view fusion network, which encode a cardiac apex two-chamber medical image, a cardiac apex three-chamber medical image and a cardiac apex four-chamber medical image to obtain each characteristic, calculate single view characteristics after three views are fused by multi-view characteristic aggregation redistribution, finally perform decoding operation based on the single view characteristics after the fusion processing to respectively obtain the decoding characteristics of the three views, and further perform another multi-view characteristic aggregation redistribution module to realize accurate automatic segmentation of the left atrium in the cardiac apex two-chamber, the cardiac apex three-chamber and the cardiac apex four-chamber.
Drawings
FIG. 1 is a schematic diagram of experimental data set formation according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a method for segmenting the left atrium of an echocardiogram based on deep learning according to an embodiment of the present invention;
fig. 3 is a schematic diagram of the deep multi-view convergence network according to the first embodiment of the present invention.
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Example one
Referring to fig. 3, the present embodiment provides a method for segmenting an echocardiographic left atrium map based on a multi-view fusion network, which specifically includes the following steps:
step 1: acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram, and respectively preprocessing the medical images;
step 2: respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for encoding;
in a specific implementation, the process of encoding in the encoder is:
the encoder adopts ResNet as a basic network as the encoder and cascades a plurality of convolution layers, down sampling and active layers.
In a specific implementation, the pre-processing includes a random flip operation, a random crop operation, and a resize operation.
The data preprocessing mainly comprises the following three steps:
(1) and (4) random overturning: flipping the input image and the corresponding label with a probability of 0.5;
(2) random cutting: cropping the given image to random size and aspect ratio;
(3) adjusting the size: the input image is resized to a given size.
The basic network layer adopts a LedNet asymmetric encoder-decoder structure, so that network parameters are greatly reduced, and the operation speed is improved. The encoder adopts ResNet as a basic network, and an attention pyramid structure is used in a decoder, so that the complexity of the network is further reduced. A multi-view feature aggregation redistribution module is added between the encoder and decoder that operates using a feature concatenation and spatial attention mechanism such that the left atrial region between the three view features remains the same spatial distribution, as shown in fig. 2.
And step 3: calculating the characteristics of the single view after the three views are fused by the characteristics of the different views obtained by the encoder through characteristic aggregation redistribution;
and 4, step 4: decoding operation is carried out on the basis of the fused features and a decoder corresponding to the encoder, and decoding features of the three views are obtained respectively; the decoding characteristics of the three views are redistributed by multi-view characteristic aggregation to obtain an accurate segmentation result.
In the specific implementation, the process of multi-view feature aggregation reallocation is as follows:
firstly, performing feature connection on three different views, and then respectively performing Non-local operation on the connection features and the single view features to respectively obtain the weight of the single view features. And multiplying the calculated weight by the single view characteristic and adding the multiplied weight to the original single characteristic to obtain the redistributed single view characteristic.
In a specific implementation, the process of multi-view feature aggregation reallocation may be integrated into one software module, for example, implemented by a multi-view feature aggregation reallocation module.
The method of the present embodiment is described in detail below.
Firstly, corresponding equipment is utilized to acquire an echocardiogram image, and the echocardiogram image of each experimental object is acquired under the support of hospital data. After the images are collected, the collected images are processed to be made into an experimental data set.
The process of constructing the experimental data set is shown in fig. 1 and comprises: three parts of data acquisition, data annotation and data enhancement pretreatment;
the data acquisition comprises the acquisition of three different echocardiography images of a patient, and by taking the patient as a unit, the DICOM images of a second chamber of the apex, a third chamber of the apex and a fourth chamber of the apex are selected from the echocardiography and are converted into PNG format.
In the data labeling, the position of the left atrium is manually labeled on the PNG image by adopting LabelMe, the outline of the left atrium is drawn point by point during labeling, and an original drawing (img.png) and a label drawing (label.png) are read from a generated JSON file to serve as an experimental data set.
The experimental data enhancement pretreatment method specifically comprises the following steps:
(1) and (4) random overturning: flipping the input image and the corresponding label with a probability of 0.5;
(2) random cutting: cropping the given image to random size and aspect ratio;
(3) adjusting the size: the input image is resized to a given size.
And inputting the experimental data set into a depth multi-view fusion network to obtain an echocardiogram left atrium segmentation result.
The structure diagram of the depth multi-view fusion network is shown in fig. 3, and the main steps are that medical images of an echocardiogram center point two chamber, a heart point three chamber and a heart point four chamber are respectively input into an encoder for encoding, then single view features obtained after three views are fused are calculated through multi-view feature aggregation reallocation, then decoding operation and another multi-view feature aggregation reallocation module are carried out on the single view features after fusion processing, and accurate segmentation results are obtained.
Example two
The embodiment provides an echocardiography left atrial map segmentation system based on a multi-view fusion network, which comprises:
the image preprocessing module is used for acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram and respectively preprocessing the medical images;
the feature coding module is used for respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for coding;
a multi-view feature fusion module, configured to perform feature aggregation and redistribution on the single view features obtained by the encoder, so that three view features are spatially aligned with each other;
the characteristic decoding module is used for carrying out decoding operation based on the aligned single view characteristics and a decoder corresponding to the encoder to respectively obtain the decoding characteristics of the three views; and the decoding characteristics of the three views are subjected to multi-view characteristic aggregation reallocation to obtain an accurate segmentation result.
It should be noted that, each module in the present embodiment corresponds to each step in the first embodiment one to one, and the specific implementation process is the same, which will not be described again here.
EXAMPLE III
The present embodiment provides a computer readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for segmenting a left atrial map of echocardiography based on a multi-view fusion network as described above.
Example four
The present embodiment provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method for segmenting the echocardiogram left atrium based on the multi-view fusion network as described above when executing the program.
It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general purpose processors, digital signal processors DSP, application specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and so on. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may include both read-only memory and random access memory, and may provide instructions and data to the processor, and a portion of the memory may also include non-volatile random access memory. For example, the memory may also store device type information.
In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (10)
1. A method for segmenting an echocardiography left atrium map based on a multi-view fusion network is characterized by comprising the following steps:
acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram, and respectively preprocessing the medical images;
respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for encoding;
calculating the single view characteristic after the three views are fused by carrying out multi-view characteristic aggregation reallocation on each characteristic obtained by the encoder;
and performing decoding operation based on the single view characteristics after the fusion processing to respectively obtain the decoding characteristics of the three views, and further performing multi-view characteristic aggregation redistribution to obtain an accurate segmentation result.
2. The method for segmenting the echocardiogram left atrium based on the multi-view fusion network as claimed in claim 1, wherein the process of calculating the single view characteristics after the three-view fusion through the multi-view characteristic aggregation reallocation is as follows:
firstly, carrying out feature connection on three different views, then respectively carrying out Non-local operation on the connection features and the single view features to respectively obtain weights of the single view features, multiplying the weights obtained by calculation with the single view features and adding the weights and the original single features to obtain the redistributed single view features.
3. The method of segmenting the echocardiographic left atrium based on a multi-view fusion network of claim 1, wherein the encoder uses a ResNet based network.
4. The method of multi-view fusion network-based echocardiographic left atrial map segmentation as in claim 3, wherein the encoder is cascaded into a plurality of convolutional layers, downsampling, and active layers.
5. The method of multi-view fusion network-based echocardiographic left atrial map segmentation as set forth in claim 1, wherein the preprocessing comprises a stochastic flipping operation, a stochastic cropping operation, and a resizing operation.
6. The method of multi-view fusion network-based echocardiographic left atrial map segmentation as claimed in claim 1, wherein the decoder is focused on a pyramid structure.
7. An echocardiographic left atrium image segmentation system based on a multi-view fusion network, which is characterized by comprising:
the image preprocessing module is used for acquiring medical images of a center point two chamber, a center point three chamber and a center point four chamber of the echocardiogram and respectively preprocessing the medical images;
the feature coding module is used for respectively inputting the preprocessed apical two-chamber medical image, apical three-chamber medical image and apical four-chamber medical image into corresponding encoders for coding;
the multi-view feature fusion module is used for respectively endowing each feature obtained by the encoder with a preset weight value, and averaging the three weight values to obtain a fused feature;
and the feature decoding module is used for performing decoding operation based on the single view features after the fusion processing to respectively obtain the decoding features of the three views, and further performing multi-view feature aggregation redistribution to obtain an accurate segmentation result.
8. The system of claim 7, wherein the computing of the single view feature after three-view fusion by multi-view feature fusion reassignment in the multi-view feature fusion module is performed by:
firstly, carrying out feature connection on three different views, then respectively carrying out Non-local operation on the connection features and the single view features to respectively obtain weights of the single view features, multiplying the weights obtained by calculation with the single view features and adding the weights and the original single features to obtain the redistributed single view features.
9. A 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 for segmentation of the echocardiographic left atrium map based on a multi-view fusion network according to any one of claims 1-6.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps in the method for echocardiography left atrial map segmentation based on a multi-view fusion network according to any one of claims 1-6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111424991.2A CN114004859A (en) | 2021-11-26 | 2021-11-26 | Method and system for segmenting echocardiography left atrium map based on multi-view fusion network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111424991.2A CN114004859A (en) | 2021-11-26 | 2021-11-26 | Method and system for segmenting echocardiography left atrium map based on multi-view fusion network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114004859A true CN114004859A (en) | 2022-02-01 |
Family
ID=79930512
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111424991.2A Pending CN114004859A (en) | 2021-11-26 | 2021-11-26 | Method and system for segmenting echocardiography left atrium map based on multi-view fusion network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114004859A (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020113886A1 (en) * | 2018-12-07 | 2020-06-11 | 中国科学院自动化研究所 | Behavior feature extraction method, system and apparatus based on time-space/frequency domain hybrid learning |
CN111739000A (en) * | 2020-06-16 | 2020-10-02 | 山东大学 | System and device for improving left ventricle segmentation accuracy of multiple cardiac views |
CN112541508A (en) * | 2020-12-21 | 2021-03-23 | 山东师范大学 | Fruit segmentation and recognition method and system and fruit picking robot |
CN112949388A (en) * | 2021-01-27 | 2021-06-11 | 上海商汤智能科技有限公司 | Image processing method and device, electronic equipment and storage medium |
CN113076972A (en) * | 2021-03-04 | 2021-07-06 | 山东师范大学 | Two-stage Logo image detection method and system based on deep learning |
CN113361606A (en) * | 2021-06-07 | 2021-09-07 | 齐鲁工业大学 | Deep map attention confrontation variational automatic encoder training method and system |
US20210326656A1 (en) * | 2020-04-15 | 2021-10-21 | Adobe Inc. | Panoptic segmentation |
-
2021
- 2021-11-26 CN CN202111424991.2A patent/CN114004859A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020113886A1 (en) * | 2018-12-07 | 2020-06-11 | 中国科学院自动化研究所 | Behavior feature extraction method, system and apparatus based on time-space/frequency domain hybrid learning |
US20210326656A1 (en) * | 2020-04-15 | 2021-10-21 | Adobe Inc. | Panoptic segmentation |
CN111739000A (en) * | 2020-06-16 | 2020-10-02 | 山东大学 | System and device for improving left ventricle segmentation accuracy of multiple cardiac views |
CN112541508A (en) * | 2020-12-21 | 2021-03-23 | 山东师范大学 | Fruit segmentation and recognition method and system and fruit picking robot |
CN112949388A (en) * | 2021-01-27 | 2021-06-11 | 上海商汤智能科技有限公司 | Image processing method and device, electronic equipment and storage medium |
CN113076972A (en) * | 2021-03-04 | 2021-07-06 | 山东师范大学 | Two-stage Logo image detection method and system based on deep learning |
CN113361606A (en) * | 2021-06-07 | 2021-09-07 | 齐鲁工业大学 | Deep map attention confrontation variational automatic encoder training method and system |
Non-Patent Citations (2)
Title |
---|
VILSON 等: ""Artificial intelligence applied to support medical decisions for the automatic analysis of echocardiogram images:A systematic review"", 《ARTIFICIAL INTELLIGENCE IN MEDICINE》, 31 October 2021 (2021-10-31) * |
宋雨慧: ""基于深度集成网络的心脏结构分割方法研究"", 《中国优秀硕士学位论文全文数据库》, 31 May 2021 (2021-05-31) * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2020119679A1 (en) | Three-dimensional left atrium segmentation method and apparatus, terminal device, and storage medium | |
US20210049397A1 (en) | Semantic segmentation method and apparatus for three-dimensional image, terminal, and storage medium | |
CN110009669B (en) | 3D/2D medical image registration method based on deep reinforcement learning | |
EP3591616A1 (en) | Automated determination of a canonical pose of a 3d dental structure and superimposition of 3d dental structures using deep learning | |
CN111292314B (en) | Coronary artery segmentation method, device, image processing system and storage medium | |
CN111368849B (en) | Image processing method, image processing device, electronic equipment and storage medium | |
CN111161269B (en) | Image segmentation method, computer device, and readable storage medium | |
CN111598862B (en) | Breast molybdenum target image segmentation method, device, terminal and storage medium | |
CN111144486B (en) | Heart nuclear magnetic resonance image key point detection method based on convolutional neural network | |
CN110570394B (en) | Medical image segmentation method, device, equipment and storage medium | |
CN113554665A (en) | Blood vessel segmentation method and device | |
CN111028248A (en) | Method and device for separating static and dynamic pulses based on CT (computed tomography) image | |
WO2023142781A1 (en) | Image three-dimensional reconstruction method and apparatus, electronic device, and storage medium | |
WO2021184799A1 (en) | Medical image processing method and apparatus, and device and storage medium | |
CN113344950A (en) | CBCT image tooth segmentation method combining deep learning with point cloud semantics | |
WO2021168745A1 (en) | Method and apparatus for training magnetic resonance imaging model | |
CN114549540A (en) | Method for automatically fusing oral scanning tooth data and CBCT (Cone Beam computed tomography) data and application thereof | |
CN114004859A (en) | Method and system for segmenting echocardiography left atrium map based on multi-view fusion network | |
CN113706684A (en) | Three-dimensional blood vessel image reconstruction method, system, medical device and storage medium | |
WO2024051018A1 (en) | Pet parameter image enhancement method and apparatus, device, and storage medium | |
WO2022247004A1 (en) | Image prediction model training method and apparatus, image prediction model application method and apparatus, device, and storage medium | |
CN113379770B (en) | Construction method of nasopharyngeal carcinoma MR image segmentation network, image segmentation method and device | |
CN115471654A (en) | SPECT image left ventricle automatic segmentation method and system | |
WO2022120694A1 (en) | Low-dose image denoising network training method and low-dose image denoising method | |
CN111080588A (en) | Multi-scale neural network-based rapid fetal MR image brain extraction method |
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 |