CN117173057B - Noise reduction method, system, equipment and storage medium for coronary angiography image - Google Patents
Noise reduction method, system, equipment and storage medium for coronary angiography image Download PDFInfo
- Publication number
- CN117173057B CN117173057B CN202311450579.7A CN202311450579A CN117173057B CN 117173057 B CN117173057 B CN 117173057B CN 202311450579 A CN202311450579 A CN 202311450579A CN 117173057 B CN117173057 B CN 117173057B
- Authority
- CN
- China
- Prior art keywords
- cardiac
- coronary angiography
- information
- preset
- noise reduction
- 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
- 238000002586 coronary angiography Methods 0.000 title claims abstract description 126
- 230000009467 reduction Effects 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000000747 cardiac effect Effects 0.000 claims abstract description 155
- 230000009466 transformation Effects 0.000 claims abstract description 23
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 5
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000011946 reduction process Methods 0.000 claims description 4
- 238000013519 translation Methods 0.000 claims description 4
- 230000007246 mechanism Effects 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 10
- 230000010247 heart contraction Effects 0.000 abstract description 6
- 239000011159 matrix material Substances 0.000 description 22
- 230000008569 process Effects 0.000 description 6
- 238000002583 angiography Methods 0.000 description 4
- 239000002872 contrast media Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000005855 radiation Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000010009 beating Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 208000029078 coronary artery disease Diseases 0.000 description 2
- 210000004351 coronary vessel Anatomy 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000005489 elastic deformation Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 230000003143 atherosclerotic effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention discloses a noise reduction method, a noise reduction system, noise reduction equipment and a noise reduction storage medium for a first coronary angiography image, wherein the noise reduction method, the noise reduction system, the noise reduction equipment and the noise reduction storage medium are used for preprocessing the first coronary angiography image to obtain first coronary angiography image sequence information; analyzing one-dimensional cardiac data, generating a cardiac feature map, slicing, performing linear transformation to obtain serialized cardiac coding information, and extracting the relation between every two cardiac coding information to obtain cardiac reference information; information encoding is carried out on the first coronary angiography sequence information to obtain a first feature image set, and frame features are obtained by combining cardiac reference information; adding position codes to the frame features and analyzing to obtain a frame sequence feature atlas and analyzing to obtain a second feature atlas; and decoding the second feature image set to generate a second noise-reduced coronary angiography image. According to the embodiment of the invention, better noise reduction effect can be obtained through interframe noise reduction, the problem of artifacts caused by heart beating is solved, and the coronary angiography is noise reduced by utilizing multi-mode information, so that better noise reduction effect is achieved.
Description
Technical Field
The embodiment of the invention relates to the technical field of image noise reduction, in particular to a method, a system, equipment and a storage medium for noise reduction of a coronary angiography image.
Background
Coronary artery is the coronary artery of the heart of an important organ in a human body, coronary angiography images are a common method for diagnosing coronary atherosclerotic heart disease (coronary heart disease), are a safer and more reliable invasive diagnosis technology, and are regarded as 'gold standard' for diagnosing coronary heart disease.
In the current image denoising algorithm, the image denoising based on deep learning achieves good denoising effect, but the following problems are also encountered when denoising the coronary angiography image:
1) The method is more complex, is greatly influenced by the radiation agent/contrast agent, and in the angiography process, the relation among the radiation dose, the contrast agent dose and the image quality is often balanced due to health and safety consideration, so that the noise reduction of the angiography image is more complex compared with the noise reduction of the visible light image;
2) The heart beating is easy to generate artifacts, and the noise reduction method considering the relation between the upper frame and the lower frame has better noise reduction effect due to more reference information. However, in the coronary angiography process, rapid movement and elastic deformation of the region of interest are caused by heart beating, so that the effect of interframe noise reduction is poor, and the problem of artifact is generated;
3) The traditional single-mode noise reduction method only considers information in the image, but in the coronary angiography process, data of other modes can also provide useful information for noise reduction, for example, a cardiac cycle can provide cardiac information to help interframe noise reduction.
Disclosure of Invention
Therefore, the embodiment of the invention provides a method, a system, equipment and a storage medium for denoising a coronary angiography image, which are used for solving the technical problem that the beating of a heart is easy to generate artifacts when denoising the coronary angiography image in the prior art.
In order to achieve the above object, the embodiment of the present invention provides the following technical solutions:
according to a first aspect of an embodiment of the present invention, there is provided a method for denoising a coronary angiography image, the method comprising:
s1, receiving a group of first coronary angiography images, and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information;
s2, acquiring one-dimensional cardiac data, analyzing the one-dimensional cardiac data, generating a cardiac feature map, slicing the cardiac feature map, performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information into a preset transducer encoder, extracting relation information among the cardiac coding information, and sequentially acquiring cardiac reference information through a preset reshape layer and a linear transformation layer;
s3, carrying out information encoding on the first coronary angiography sequence information through a plurality of groups of dense chain blocks and downsampling convolution to obtain a first feature image set, and obtaining frame features in a hidden space through a preset reshape layer and combining the cardiac reference information;
s4, adding position codes to the frame features, obtaining a frame sequence feature image set containing inter-frame information through a preset transform coder, and obtaining a second feature image set through a preset reshape layer;
s5, decoding the second feature image set to generate a second coronary angiography image after noise reduction.
Further, receiving a set of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information, including:
receiving a group of first coronary angiography images, and performing data enhancement operation on the first coronary angiography images to obtain a preprocessed coronary angiography image and a preprocessed coronary angiography image set;
sequentially carrying out normalization calculation on the preprocessed coronary angiography images in the preprocessed coronary angiography image set to obtain first coronary angiography image sequence information;
wherein the data enhancement operation includes random rotation, random translation, random affine transformation, and random erasure for obtaining a motion relationship between the respective first coronary angiography images.
Further, acquiring one-dimensional cardiac data and analyzing the one-dimensional cardiac data to generate a cardiac feature map, including:
acquiring one-dimensional cardiac data, inputting the one-dimensional cardiac data into a preset one-dimensional encoder, and generating encoded cardiac data;
and inputting the encoded cardiac data into a preset decoder for decoding operation, and splicing the middle features in the encoded cardiac data with the feature images with the same size of the preset decoder through transverse connecting edges to generate cardiac feature images.
Further, the preset one-dimensional encoder and the preset one-dimensional decoder form a one-dimensional U-shaped structure.
Further, slicing the cardiac feature map and performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information to a preset transducer encoder, extracting relation information between the cardiac coding information, and sequentially obtaining cardiac reference information through a preset reshape layer and a linear transformation layer, wherein the method comprises the following steps:
slicing the cardiac feature map, and performing dimension adjustment through a preset full-connection layer to generate serialized cardiac coding information;
position coding is carried out on the serialized cardiac coding information, the position coding is input to a preset transducer coder, and the relation information among the cardiac coding information is extracted;
acquiring cardiac reference information by sequentially passing the relation information through a preset reshape layer and a linear transformation layer;
the cardiac reference information is calculated with a preset one-dimensional encoder and a preset decoder through a self-attention mechanism, and the noise reduction process is guided.
Further, the method includes the steps of performing information encoding on the first coronary angiography sequence information through multiple groups of dense chain blocks and downsampling convolution to obtain a first feature atlas, and obtaining frame features in a hidden space through a preset reshape layer and combining the cardiac reference information by the first feature atlas, wherein the method comprises the following steps:
information encoding is carried out on the first coronary angiography sequence information, and a first characteristic image set is obtained;
performing attention calculation by using the first feature map in the first feature map set and the cardiac reference information, and fusing cardiac information to generate frame features in a hidden space;
the first feature map is output to the preset decoder through the transverse connection edge.
Further, decoding the second feature atlas to generate a denoised second coronary angiography image, including:
upsampling the second feature atlas through a plurality of layers of dense link blocks plus bilinear differences to generate a third feature atlas;
performing attention calculation on the third feature map and the cardiac reference information, fusing cardiac information, and generating a second coronary angiography image after noise reduction;
the output data of the previous layer of convolution block is the input data of the current layer of convolution block, the input data and the encoder same-size feature map are stacked, and the output data of the current layer of convolution block is generated.
According to a second aspect of an embodiment of the present invention, there is provided a noise reduction system for a coronary angiography image, the noise reduction system including:
the preprocessing module is used for receiving a group of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information;
the cardiac reference information acquisition module is used for acquiring one-dimensional cardiac data, analyzing the one-dimensional cardiac data, generating a cardiac feature map, slicing the cardiac feature map, performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information into a preset transducer encoder, extracting relation information among the cardiac coding information, and sequentially acquiring cardiac reference information through a preset reshape layer and a linear transformation layer;
the frame characteristic acquisition module is used for carrying out information encoding on the first coronary angiography sequence information through a plurality of groups of dense link blocks and downsampling convolution to acquire a first characteristic atlas, and acquiring frame characteristics in a hidden space through a preset reshape layer and combining the cardiac reference information;
the second feature atlas acquisition module is used for adding position codes to the frame features, acquiring a frame sequence feature atlas containing inter-frame information through a preset transform encoder, and acquiring the second feature atlas through a preset reshape layer;
and the decoding module is used for decoding the second characteristic image set and generating a second noise-reduced coronary angiography image.
According to a third aspect of embodiments of the present invention, there is provided a noise reduction apparatus for a coronary angiography image, the apparatus comprising: a processor and a memory;
the memory is used for storing one or more program instructions;
the processor is configured to execute one or more program instructions for performing the steps of a method of denoising a coronary angiography image according to any one of the preceding claims.
According to a fourth aspect of embodiments of the present invention, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of a method of denoising a coronary angiography image according to any one of the above.
The embodiment of the invention has the following advantages:
according to the embodiment of the invention, better noise reduction effect can be obtained through interframe noise reduction, the problem of artifacts caused by heart beating is solved, and the coronary angiography is noise reduced by utilizing multi-mode information, so that better noise reduction effect is achieved.
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. It will be apparent to those of ordinary skill in the art that the drawings in the following description are exemplary only and that other implementations can be obtained from the extensions of the drawings provided without inventive effort.
The structures, proportions, sizes, etc. shown in the present specification are shown only for the purposes of illustration and description, and are not intended to limit the scope of the invention, which is defined by the claims, so that any structural modifications, changes in proportions, or adjustments of sizes, which do not affect the efficacy or the achievement of the present invention, should fall within the ambit of the technical disclosure.
FIG. 1 is a schematic diagram of a logic structure of a noise reduction system for a coronary angiography image according to an embodiment of the present invention;
FIG. 2 is a flowchart of a method for denoising a coronary angiography image according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a noise reduction model in a noise reduction method for a coronary angiography image according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a logic structure of an encoder in a method for denoising a coronary angiography image according to an embodiment of the present invention;
fig. 5 is a schematic logic structure diagram of a decoder in a method for denoising a coronary angiography image according to an embodiment of the present invention.
Detailed Description
Other advantages and advantages of the present invention will become apparent to those skilled in the art from the following detailed description, which, by way of illustration, is to be read in connection with certain specific embodiments, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the current image denoising algorithm, the image denoising based on deep learning achieves good denoising effect, but the following problems are also encountered when denoising the coronary angiography image:
1) The method is more complex, is greatly influenced by the radiation agent/contrast agent, and in the angiography process, the relation among the radiation dose, the contrast agent dose and the image quality is often balanced due to health and safety consideration, so that the noise reduction of the angiography image is more complex compared with the noise reduction of the visible light image;
2) The heart beating is easy to generate artifacts, and the noise reduction method considering the relation between the upper frame and the lower frame has better noise reduction effect due to more reference information. However, in the coronary angiography process, rapid movement and elastic deformation of the region of interest are caused by heart beating, so that the effect of interframe noise reduction is poor, and the problem of artifact is generated;
3) The traditional single-mode noise reduction method only considers information in the image, but in the coronary angiography process, data of other modes can also provide useful information for noise reduction, for example, a cardiac cycle can provide cardiac information to help interframe noise reduction.
In order to solve the technical problem that the beating of the heart is easy to generate artifact when the coronary angiography image is noise reduced.
Referring to fig. 1, an embodiment of the present invention discloses a noise reduction system for a coronary angiography image, the system comprising: a preprocessing module 1; a cardiac reference information acquisition module 2; a frame feature acquisition module 3; a second feature atlas acquisition module 4; a decoding module 5.
Corresponding to the above-disclosed noise reduction system for the coronary angiography image, the embodiment of the invention also discloses a noise reduction method for the coronary angiography image. The following describes in detail a method for denoising a coronary angiography image disclosed in an embodiment of the present invention in conjunction with a denoising system for a coronary angiography image described above.
Referring to fig. 3, the invention discloses a method for denoising a coronary angiography image, which comprises the following steps:
s1, receiving a group of first coronary angiography images, and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information.
S2, acquiring one-dimensional cardiac data, analyzing the one-dimensional cardiac data, generating a cardiac feature map, slicing the cardiac feature map, performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information into a preset transducer encoder, extracting relation information among the cardiac coding information, and sequentially acquiring cardiac reference information through a preset reshape layer and a linear transformation layer.
S3, carrying out information encoding on the first coronary angiography sequence information through a plurality of groups of dense chain blocks and downsampling convolution to obtain a first feature image set, and obtaining frame features in a hidden space through a preset reshape layer and combining the cardiac reference information.
S4, adding position codes to the frame features, obtaining a frame sequence feature image set containing inter-frame information through a preset transform coder, and obtaining a second feature image set through a preset reshape layer.
S5, decoding the second feature image set to generate a second coronary angiography image after noise reduction.
Further, receiving a set of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information, including: receiving a group of first coronary angiography images, and performing data enhancement operation on the first coronary angiography images to obtain a preprocessed coronary angiography image and a preprocessed coronary angiography image set; and sequentially carrying out normalization calculation on the preprocessed coronary angiography images in the preprocessed coronary angiography image set to obtain first coronary angiography image sequence information.
Wherein the data enhancement operation includes random rotation, random translation, random affine transformation, and random erasure for obtaining a motion relationship between the respective first coronary angiography images.
Further, acquiring one-dimensional cardiac data and analyzing the one-dimensional cardiac data to generate a cardiac feature map, including: acquiring one-dimensional cardiac data, inputting the one-dimensional cardiac data into a preset one-dimensional encoder, and generating encoded cardiac data; and inputting the encoded cardiac data into a preset decoder for decoding operation, and splicing the middle features in the encoded cardiac data with the feature images with the same size of the preset decoder through transverse connecting edges to generate cardiac feature images.
The cardiac reference information is calculated through a K matrix and a V matrix which are used as attention, and an encoder and a decoder, so that the noise reduction process is guided.
Further, the preset one-dimensional encoder is connected with the preset one-dimensional decoder to form a one-dimensional U-shaped structure.
Further, slicing the cardiac feature map and performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information to a preset transducer encoder, extracting relation information between the cardiac coding information, and sequentially obtaining cardiac reference information through a preset reshape layer and a linear transformation layer, wherein the method comprises the following steps: slicing the cardiac feature map, and performing dimension adjustment through a preset full-connection layer to form serialized cardiac coding information; position coding is carried out on the serialized cardiac coding information, the position coding is input to a preset transducer coder, and the relation information among the cardiac coding information is extracted; and acquiring the cardiac reference information by sequentially passing the relation information through a preset reshape layer and a linear transformation layer.
The cardiac reference information is calculated with a preset one-dimensional encoder and a preset decoder through a self-attention mechanism, and the noise reduction process is guided.
Further, the method includes the steps of performing information encoding on the first coronary angiography sequence information through multiple groups of dense chain blocks and downsampling convolution to obtain a first feature atlas, and obtaining frame features in a hidden space through a preset reshape layer and combining the cardiac reference information by the first feature atlas, wherein the method comprises the following steps: information encoding is carried out on the first coronary angiography sequence information, and a first characteristic image set is obtained; and performing attention calculation by using the first feature map in the first feature map set and the cardiac reference information, and fusing cardiac information to generate frame features in the hidden space.
The intermediate feature map is used as a Q matrix to calculate attention with a K matrix and a V matrix output by the cardiac reference information acquisition module 2 and used for fusing cardiac information, and the attention calculation formula is as follows:
wherein, the Q matrix is an intermediate feature map, namely a first/second/third feature map, the K matrix and the V matrix are the cardiac reference information output by the cardiac reference information acquisition module 2, and the K matrix is the cardiac reference information output by the cardiac reference information acquisition module 2 T The transpose of the matrix K, KT is the column number of the matrix Q and the matrix K.
The first feature map is output to the preset decoder through the transverse connection edge.
Further, decoding the second feature atlas to generate a denoised second coronary angiography image, including: upsampling the second feature atlas through a plurality of layers of dense link blocks plus bilinear differences to generate a third feature atlas; and performing attention calculation on the third feature map and the cardiac reference information, fusing cardiac information, and generating a second coronary angiography image after noise reduction.
The output data of the previous layer of convolution block is the input data of the current layer of convolution block, the input data and the encoder same-size feature map are stacked, and the output data of the current layer of convolution block is generated.
The intermediate feature map is used as a matrix K and a matrix V output by the Q matrix and cardiac reference information acquisition module 2 for performing attention calculation and is used for fusing cardiac information, and the attention calculation formula is as follows:
wherein, the Q matrix is an intermediate feature map, namely a first/second/third feature map, the K matrix and the V matrix are the cardiac reference information output by the cardiac reference information acquisition module 2, and the K matrix is the cardiac reference information output by the cardiac reference information acquisition module 2 T The transpose of the matrix K, KT is the column number of the matrix Q and the matrix K.
In addition, the embodiment of the invention also provides a noise reduction device for the coronary angiography image, which comprises: a processor and a memory; the memory is used for storing one or more program instructions; the processor is configured to execute one or more program instructions for performing the steps of a method of denoising a coronary angiography image according to any one of the preceding claims.
In addition, an embodiment of the present invention further provides a computer readable storage medium, where a computer program is stored, where the computer program is executed by a processor to implement the steps of a method for denoising a coronary angiography image according to any one of the above.
In the embodiment of the invention, the processor may be an integrated circuit chip with signal processing capability. The processor may be a general purpose processor, a digital signal processor (Digital Signal Processor, DSP for short), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC for short), a field programmable gate array (FieldProgrammable Gate Array, FPGA for short), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components.
The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be embodied directly in the execution of a hardware decoding processor, or in the execution of a combination of hardware and software modules in a decoding processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The processor reads the information in the storage medium and, in combination with its hardware, performs the steps of the above method.
The storage medium may be memory, for example, may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory.
The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an electrically Erasable ROM (Electrically EPROM, EEPROM), or a flash Memory.
The volatile memory may be a random access memory (Random Access Memory, RAM for short) which acts as an external cache. By way of example, and not limitation, many forms of RAM are available, such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (Double Data RateSDRAM), enhanced SDRAM (ESDRAM), synchronous DRAM (SLDRAM), and direct memory bus RAM (directracram, DRRAM).
The storage media described in embodiments of the present invention are intended to comprise, without being limited to, these and any other suitable types of memory.
Those skilled in the art will appreciate that in one or more of the examples described above, the functions described in the present invention may be implemented in a combination of hardware and software. When the software is applied, the corresponding functions may be stored in a computer-readable medium or transmitted as one or more instructions or code on the computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
While the invention has been described in detail in the foregoing general description and specific examples, it will be apparent to those skilled in the art that modifications and improvements can be made thereto. Accordingly, such modifications or improvements may be made without departing from the spirit of the invention and are intended to be within the scope of the invention as claimed.
Claims (8)
1. A method of denoising a coronary angiography image, the method comprising:
s1, receiving a group of first coronary angiography images, and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information;
s2, acquiring one-dimensional cardiac data, analyzing the one-dimensional cardiac data, generating a cardiac feature map, slicing the cardiac feature map, performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information into a preset transducer encoder, extracting relation information among the cardiac coding information, and sequentially acquiring cardiac reference information through a preset reshape layer and a linear transformation layer;
s3, carrying out information encoding on the first coronary angiography image sequence information through a plurality of groups of dense chain blocks and downsampling convolution to obtain a first feature image set, and obtaining frame features in a hidden space through a preset reshape layer and combining the cardiac reference information;
s4, adding position codes to the frame features, obtaining a frame sequence feature image set containing inter-frame information through a preset transform coder, and obtaining a second feature image set through a preset reshape layer;
s5, decoding the second feature atlas to generate a second coronary angiography image after noise reduction;
receiving a group of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information, wherein the method comprises the following steps of:
receiving a group of first coronary angiography images, and performing data enhancement operation on the first coronary angiography images to obtain a preprocessed coronary angiography image and a preprocessed coronary angiography image set;
sequentially carrying out normalization calculation on the preprocessed coronary angiography images in the preprocessed coronary angiography image set to obtain first coronary angiography image sequence information;
the data enhancement operation comprises random rotation, random translation, random affine transformation and random erasure, and is used for obtaining the motion relation between the first coronary angiography images;
acquiring one-dimensional cardiac data and analyzing the one-dimensional cardiac data to generate a cardiac feature map, including:
acquiring one-dimensional cardiac data, inputting the one-dimensional cardiac data into a preset one-dimensional encoder, and generating encoded cardiac data;
inputting the encoded cardiac data into a preset one-dimensional decoder for decoding operation, and splicing the middle features in the encoded cardiac data with the same-size feature images of the preset one-dimensional decoder through transverse connecting edges to generate cardiac feature images.
2. The method of claim 1, wherein the predetermined one-dimensional encoder is connected to the predetermined one-dimensional decoder to form a one-dimensional U-shaped structure.
3. The method for denoising a coronary angiography image according to claim 2, wherein slicing and linearly transforming the cardiac feature map to obtain serialized cardiac coding information, inputting the serialized cardiac coding information to a preset transducer encoder, extracting relation information between the cardiac coding information, and sequentially obtaining cardiac reference information through a preset reshape layer and a linear transformation layer, comprises:
slicing the cardiac feature map, and performing dimension adjustment through a preset full-connection layer to generate serialized cardiac coding information;
position coding is carried out on the serialized cardiac coding information, the position coding is input to a preset transducer coder, and the relation information among the cardiac coding information is extracted;
acquiring cardiac reference information by sequentially passing the relation information through a preset reshape layer and a linear transformation layer;
the cardiac reference information is calculated with a preset one-dimensional encoder and a preset one-dimensional decoder through a self-attention mechanism, and the noise reduction process is guided.
4. A method of denoising a coronary angiography image according to claim 3, wherein the step of information encoding the first coronary angiography image sequence information by means of a plurality of sets of dense chain blocks and downsampling convolutions to obtain a first feature atlas, and wherein the step of obtaining frame features in hidden space by means of a preset reshape layer and combining the cardiac reference information comprises the steps of:
information encoding is carried out on the first coronary angiography image sequence information, and a first characteristic image set is obtained;
performing attention calculation by using the first feature map in the first feature map set and the cardiac reference information, and fusing cardiac information to generate frame features in a hidden space;
the first feature map is output to the preset one-dimensional decoder through the transverse connection edge.
5. The method of denoising a coronary angiography image of claim 4, wherein decoding the second feature set of images to generate a denoised second coronary angiography image comprises:
upsampling the second feature atlas through a plurality of layers of dense link blocks plus bilinear differences to generate a third feature atlas;
performing attention calculation on the third feature map and the cardiac reference information, fusing cardiac information, and generating a second coronary angiography image after noise reduction;
the output data of the previous layer of convolution block is the input data of the current layer of convolution block, the input data and the encoder same-size feature map are stacked, and the output data of the current layer of convolution block is generated.
6. A noise reduction system for a coronary angiography image, the noise reduction system comprising:
the preprocessing module is used for receiving a group of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information;
the cardiac reference information acquisition module is used for acquiring one-dimensional cardiac data, analyzing the one-dimensional cardiac data, generating a cardiac feature map, slicing the cardiac feature map, performing linear transformation to obtain serialized cardiac coding information, inputting the serialized cardiac coding information into a preset transducer encoder, extracting relation information among the cardiac coding information, and sequentially acquiring cardiac reference information through a preset reshape layer and a linear transformation layer;
the frame characteristic acquisition module is used for carrying out information encoding on the first coronary angiography image sequence information through a plurality of groups of dense link blocks and downsampling convolution to acquire a first characteristic atlas, and acquiring frame characteristics in a hidden space through a preset reshape layer and combining the cardiac reference information;
the second feature atlas acquisition module is used for adding position codes to the frame features, acquiring a frame sequence feature atlas containing inter-frame information through a preset transform encoder, and acquiring the second feature atlas through a preset reshape layer;
the decoding module is used for decoding the second characteristic image set and generating a second coronary angiography image after noise reduction;
receiving a group of first coronary angiography images and sequentially preprocessing the first coronary angiography images to generate first coronary angiography image sequence information, wherein the method comprises the following steps of:
receiving a group of first coronary angiography images, and performing data enhancement operation on the first coronary angiography images to obtain a preprocessed coronary angiography image and a preprocessed coronary angiography image set;
sequentially carrying out normalization calculation on the preprocessed coronary angiography images in the preprocessed coronary angiography image set to obtain first coronary angiography image sequence information;
the data enhancement operation comprises random rotation, random translation, random affine transformation and random erasure, and is used for obtaining the motion relation between the first coronary angiography images;
acquiring one-dimensional cardiac data and analyzing the one-dimensional cardiac data to generate a cardiac feature map, including:
acquiring one-dimensional cardiac data, inputting the one-dimensional cardiac data into a preset one-dimensional encoder, and generating encoded cardiac data;
inputting the encoded cardiac data into a preset one-dimensional decoder for decoding operation, and splicing the middle features in the encoded cardiac data with the same-size feature images of the preset one-dimensional decoder through transverse connecting edges to generate cardiac feature images.
7. A noise reduction device for a coronary angiography image, the noise reduction device comprising: a processor and a memory;
the memory is used for storing one or more program instructions;
the processor is configured to execute one or more program instructions for performing the steps of a method of denoising a coronary angiographic image according to any one of claims 1 to 5.
8. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, implements the steps of a method of denoising a coronary angiographic image according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311450579.7A CN117173057B (en) | 2023-11-03 | 2023-11-03 | Noise reduction method, system, equipment and storage medium for coronary angiography image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311450579.7A CN117173057B (en) | 2023-11-03 | 2023-11-03 | Noise reduction method, system, equipment and storage medium for coronary angiography image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117173057A CN117173057A (en) | 2023-12-05 |
| CN117173057B true CN117173057B (en) | 2024-02-06 |
Family
ID=88930229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311450579.7A Active CN117173057B (en) | 2023-11-03 | 2023-11-03 | Noise reduction method, system, equipment and storage medium for coronary angiography image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117173057B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110490863A (en) * | 2019-08-22 | 2019-11-22 | 北京红云智胜科技有限公司 | Whether there is or not the systems of total occlusion for detection coronary angiography based on deep learning |
| CN111261277A (en) * | 2018-11-30 | 2020-06-09 | 上海图灵医疗科技有限公司 | Heart disease detection method based on deep learning model |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210334963A1 (en) * | 2017-03-24 | 2021-10-28 | Pie Medical Imaging B.V. | Method and system for assessing vessel obstruction based on machine learning |
| EP3884868A1 (en) * | 2020-03-26 | 2021-09-29 | Pie Medical Imaging BV | Method and system for registering intra-object data with extra-object data |
-
2023
- 2023-11-03 CN CN202311450579.7A patent/CN117173057B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111261277A (en) * | 2018-11-30 | 2020-06-09 | 上海图灵医疗科技有限公司 | Heart disease detection method based on deep learning model |
| CN110490863A (en) * | 2019-08-22 | 2019-11-22 | 北京红云智胜科技有限公司 | Whether there is or not the systems of total occlusion for detection coronary angiography based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 图注意力网络的场景图到图像生成模型;兰红;刘秦邑;;中国图象图形学报(第08期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117173057A (en) | 2023-12-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102389173B1 (en) | Training method, image processing method, device and storage medium for generative adversarial network | |
| US11540798B2 (en) | Dilated convolutional neural network system and method for positron emission tomography (PET) image denoising | |
| CN111179372B (en) | Image attenuation correction method, image attenuation correction device, computer equipment and storage medium | |
| JP7340107B2 (en) | Image reconstruction method, apparatus, device, system and computer-readable storage medium | |
| CN112907439B (en) | Deep learning-based supine position and prone position breast image registration method | |
| CN111784788A (en) | PET (positron emission tomography) rapid imaging method and system based on deep learning | |
| CN104569880B (en) | A kind of magnetic resonance fast imaging method and system | |
| CN112837244B (en) | Low-dose CT image denoising and artifact removing method based on progressive generation confrontation network | |
| CN111951165B (en) | Image processing method, apparatus, computer device, and computer-readable storage medium | |
| CN115439651A (en) | DSA (digital Signal amplification) cerebrovascular segmentation system and method based on space multi-scale attention network | |
| CN111986092B (en) | Dual-network-based image super-resolution reconstruction method and system | |
| CN116993610A (en) | Method and equipment for reconstructing restricted CT image based on denoising diffusion model | |
| CN111325695A (en) | Low-dose image enhancement method and system based on multi-dose grade and storage medium | |
| CN114299185A (en) | Magnetic resonance image generation method, magnetic resonance image generation device, computer equipment and storage medium | |
| CN117173057B (en) | Noise reduction method, system, equipment and storage medium for coronary angiography image | |
| CN112419175A (en) | Weight-sharing dual-region generation countermeasure network and image generation method thereof | |
| CN117292704A (en) | Voice-driven gesture action generation method and device based on diffusion model | |
| CN112102251A (en) | Method and device for segmenting image, electronic equipment and storage medium | |
| Yang et al. | Quasi-supervised learning for super-resolution PET | |
| CN108961161B (en) | Image data processing method, device and computer storage medium | |
| CN112950478B (en) | Face super-resolution method and system based on dual identity attribute constraint | |
| CN113053496B (en) | Deep learning method for low-dose estimation of medical image | |
| KR20220071554A (en) | Medical Image Fusion System | |
| Liu et al. | Medical CT image super-resolution via cyclic feature concentration network | |
| CN114373532B (en) | Multi-mode medical image translation method for generating countermeasure network based on target perception |
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 |