CN110264435A - Enhancement Method, device, computer equipment and the storage medium of low dosage MIP image - Google Patents
Enhancement Method, device, computer equipment and the storage medium of low dosage MIP image Download PDFInfo
- Publication number
- CN110264435A CN110264435A CN201910504782.5A CN201910504782A CN110264435A CN 110264435 A CN110264435 A CN 110264435A CN 201910504782 A CN201910504782 A CN 201910504782A CN 110264435 A CN110264435 A CN 110264435A
- Authority
- CN
- China
- Prior art keywords
- image
- mip
- mip image
- low dosage
- module
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 239000000700 radioactive tracer Substances 0.000 claims abstract description 46
- 230000002708 enhancing effect Effects 0.000 claims abstract description 33
- 238000013528 artificial neural network Methods 0.000 claims abstract description 30
- 238000004590 computer program Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 10
- 230000013016 learning Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 239000003814 drug Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 230000002073 mitogenetic effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000007794 visualization technique Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/08—Volume rendering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- 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
- G06T2207/10104—Positron emission tomography [PET]
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Nuclear Medicine (AREA)
Abstract
This application involves Enhancement Method, device, computer equipment and the storage mediums of a kind of low dosage MIP image.Method includes: the first MIP image and corresponding first PET image obtained under full dose of radiation tracer, rebuilds the first MIP image and corresponding first PET image, obtains the second MIP image under the low dosage radioactive tracer of corresponding position;The second MIP image is learnt by the feedback of the first MIP image using confrontation neural network is generated, establishes image enhancement model;The current MIP image under the radioactive tracer of low dosage is obtained, current MIP image input picture enhancing model is subjected to image enhancement, output MIP enhances image.Image can be enhanced using the MIP of the quick outputting high quality of image enhancement model using this method;So as to meet the MIP image of low dosage as the quality requirement with reference to figure.
Description
Technical field
This application involves technical field of image processing, Enhancement Method, dress more particularly to a kind of low dosage MIP image
It sets, computer equipment and storage medium.
Background technique
Direct volume drawing (direct volume rendering, DVR) be most effective volume visualization method it
One, it is widely used in various fields such as medicine, geography, physics.Transmission function (transfer function, TF) is responsible for will
Volume data attribute (such as density value, gradient-norm) is mapped as the optical properties such as color, transparency, and quality has the effect of DVR
Decisive influence.The DVR method special as one kind, maximum intensity projection (maximum intensity projection,
MIP) maximum density values on throw light are projected on screen, transmission function is not necessarily to, has many advantages, such as simple and practical, curing
Etc. fields are widely applied.
In PET-MR equipment, need to generate the PET image of one group of MIP image.The effect of MIP image is: as reference
Figure, auxiliary MR picture positioning.In order to obtain the PET image of high quality, the radioactive tracer of full dosage is necessary,
But the radioactive tracer of full dosage can cause the problem of potential patient health injury.And simply by reducing dosage
MIP image can make the reduction of MIP image quality because of more noises again, be unable to satisfy as the quality requirement with reference to figure.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide Enhancement Method, the device, meter of a kind of low dosage MIP image
Machine equipment and storage medium are calculated, can satisfy the MIP image of low dosage as the quality requirement with reference to figure.
A kind of Enhancement Method of low dosage MIP image, which comprises
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, rebuilds described first
MIP image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;
Second MIP image is learnt by the feedback of the first MIP image using confrontation neural network is generated, is built
Vertical image enhancement model;
The current MIP image under the radioactive tracer of low dosage is obtained, by the input described image enhancing of current MIP image
Model carries out image enhancement, and output MIP enhances image.
The reconstruction first MIP image and corresponding first PET image in one of the embodiments, obtain pair
Answer the second MIP image under the low dosage radioactive tracer of position, comprising:
The reconstruction under various dose is carried out to the first PET image according to default reconstruction parameter, obtains corresponding 2nd PET figure
Picture;And the extraction of pixel value is carried out to second PET image, obtain the second MIP image.
It is described in one of the embodiments, that first is passed through to second MIP image using generation confrontation neural network
The feedback of MIP image is learnt, and image enhancement model is established, comprising:
Generation confrontation neural network is built, the generation confrontation neural network includes generation module and discrimination module;
Using first MIP image as label data, exercised supervision using the generation module to the second MIP image
It practises, obtains sub-picture;
The sub-picture and the first MIP image are differentiated using the discrimination module, obtain differentiating result;
Training is completed after differentiating that result meets termination condition, obtains image enhancement model.
In one of the embodiments, further include:
After building generation confrontation neural network, first MIP image and the second MIP image are carried out one-to-one
Shuffle processing.
Part in one of the embodiments, will then differentiate that result feeds back to the generation module, utilizes the generation module
Continuation supervised learning is carried out to the second MIP image, obtains new sub-picture.
In one of the embodiments, further include:
Output MIP enhancing image after, by MIP enhancing image be stored as auxiliary positioning with reference to figure.
Correspondingly, the present invention also provides a kind of enhancement device of low dosage MIP image, described device include obtain module,
Establish module and enhancing module:
The acquisition module, for obtaining the first MIP image and corresponding first PET under full dose of radiation tracer
Image rebuilds first MIP image and corresponding first PET image, obtains the low dosage radioactive tracer of corresponding position
Under the second MIP image;
It is described to establish module, for passing through the first MIP image to second MIP image using generation confrontation neural network
Feedback learnt, establish image enhancement model;
The enhancing module, the current MIP image under radioactive tracer for obtaining current low dosage, will be current
MIP image inputs described image enhancing model and carries out image enhancement, and output MIP enhances image.
Memory module is gone back in one of the embodiments:
The memory module, for after output MIP enhancing image, MIP enhancing image to be stored as to the ginseng of auxiliary positioning
Examine figure.
Correspondingly, the present invention also provides a kind of computer equipment, including memory and processor, the memory are stored with
Computer program, the processor execute the computer program and constantly perform the steps of
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, rebuilds described first
MIP image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;
Second MIP image is learnt by the feedback of the first MIP image using confrontation neural network is generated, is built
Vertical image enhancement model;
The current MIP image under the radioactive tracer of low dosage is obtained, by the input described image enhancing of current MIP image
Model carries out image enhancement, and output MIP enhances image.
Correspondingly, being stored thereon with computer program, the meter the present invention also provides a kind of computer readable storage medium
Calculation machine program performs the steps of when being executed by processor
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, rebuilds described first
MIP image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;
Second MIP image is learnt by the feedback of the first MIP image using confrontation neural network is generated, is built
Vertical image enhancement model;
The current MIP image under the radioactive tracer of low dosage is obtained, by the input described image enhancing of current MIP image
Model carries out image enhancement, and output MIP enhances image.
Enhancement Method, device, computer equipment and the storage medium of above-mentioned low dosage MIP image, first reconstruction is got
First MIP image and corresponding first PET image obtain the 2nd MIP figure under the low dosage radioactive tracer of corresponding position
Picture;It recycles generation confrontation neural network to learn second MIP image by the feedback of the first MIP image, establishes
Image enhancement model;Image enhancement directly is carried out to current MIP image using image enhancement model, to quickly export high-quality
The MIP of amount enhances image;So as to meet the MIP image of low dosage as the quality requirement with reference to figure.
Detailed description of the invention
Fig. 1 is the applied environment figure of the Enhancement Method of low dosage MIP image in one embodiment;
Fig. 2 is the flow diagram of the Enhancement Method of low dosage MIP image in one embodiment;
Fig. 3 is the flow diagram of step S200 in Fig. 2 embodiment;
Fig. 4 is the flow diagram of the Enhancement Method of low dosage MIP image in another embodiment;
Fig. 5 is the structural block diagram of the enhancement device of low dosage MIP image in one embodiment;
Fig. 6 is the structural block diagram of the enhancement device of low dosage MIP image in another embodiment;
Fig. 7 is the internal structure chart of computer equipment in one embodiment.
In figure: 1, terminal;2, server;100, module is obtained;200, module is established;300, enhance module;400, it stores
Module.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
The Enhancement Method of low dosage MIP image provided by the present application, can be applied in application environment as shown in Figure 1.
Wherein, terminal 1 is communicated with server 2 by network by network.The first MIP image and corresponding is rebuild in terminal 1
First PET image obtains the second MIP image under the low dosage radioactive tracer of corresponding position;According to the first MIP image and
Second MIP image utilizes generation confrontation neural network image enhancement model;Obtain the radioactive tracer of current low dosage
Under current MIP image, by current MIP image input picture enhancing model carry out image enhancement, output MIP enhance image.?
After establishing image enhancement model, image enhancement model can be uploaded in server 2 by terminal 1 by network, to facilitate other
The load of terminal 1 uses.Wherein, terminal 1 can be the various personal computers connecting with PET-MR equipment, laptop, intelligence
Energy mobile phone, tablet computer and portable wearable device, terminal 1 can also be exactly PET-CT equipment, and server 2 can be with independently
The server cluster of server either multiple servers composition realize.
In one embodiment, it as shown in Fig. 2, providing a kind of Enhancement Method of low dosage MIP image, answers in this way
For being illustrated for the terminal in Fig. 1, comprising the following steps:
S100, the first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, rebuilds the
One MIP image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;
The clinical data that equipment measures under the full dose of radiation tracer of acquisition is as training sample.The clinic of each patient
Data include the first MIP image and corresponding first PET image;The second MIP image under low dosage radioactive tracer and right
The second PET image answered can be measured by equipment, but since the MIP image of low dosage can make because of more noises again
MIP image quality reduces, it is also necessary to which algorithm in addition carrys out noise reduction, increases computation burden.Low dosage is for full dosage
, low dosage refers to only if it were not for full dosage, can be understood as low dosage.In one embodiment, for the same trouble
Person can rebuild the first MIP image and corresponding first PET image, obtain under the low dosage radioactive tracer of corresponding position
The second MIP image.First MIP image is the maximum intensity projection (MIP) that the equipment under full dose of radiation tracer obtains
Image.First PET image is generated using the first MIP image as with reference to figure, and picture format can be DICOM (medicine figure
As format) image.
S200, the second MIP image is learnt by the feedback of the first MIP image using generation confrontation neural network,
Establish image enhancement model;
In one embodiment, using confrontation neural network (GAN) is generated, the second MIP image using the first MIP by being schemed
As the information that feedback is come, come continuous feedback learning, the image enhancement model of foundation.Image enhancement model can store in database
In, load image enhances model.In other embodiments, other neural fusions can be used, to this and without
Limitation.
S300, current MIP image under the radioactive tracer of current low dosage is obtained, current MIP image is inputted and is schemed
Image intensifying model carries out image enhancement, and output MIP enhances image.
In different application scenarios, the case where use, can be different.It in one embodiment, can be by image
Enhancing model remains in the publication of PETMR product, after the current MIP image for generating MIP workflow, is loaded directly into trained
Image enhancement model carries out image enhancement to current MIP image, so that quickly the MIP of outputting high quality enhances image.
The Enhancement Method of above-mentioned low dosage MIP image first rebuilds the first MIP image got and corresponding first PET
Image obtains the second MIP image under the low dosage radioactive tracer of corresponding position;It recycles and generates confrontation neural network pair
Second MIP image is learnt by the feedback of the first MIP image, establishes image enhancement model;Directly utilize Image Enhancement Based
Type to carry out image enhancement to current MIP image, so that quickly the MIP of outputting high quality enhances image;It is low so as to meet
The MIP image of dosage is as the quality requirement with reference to figure.
In one embodiment, the first MIP image and corresponding first PET image are rebuild, low dose of corresponding position is obtained
Measure the second MIP image under radioactive tracer, comprising the following steps:
The reconstruction under various dose is carried out to the first PET image according to default reconstruction parameter, obtains corresponding 2nd PET figure
Picture;And the extraction of pixel value is carried out to the second PET image, obtain the second MIP image.
Specifically, dynamic reconstruction parameter is arranged by offline-recon (offline to scout) processing in the first PET image
It is rebuild, obtains corresponding second PET image under low dosage.Then it to a series of pixel value content of second PET images, deposits
Storage is gray level image to get to the second MIP image.In one embodiment, dynamic reconstruction parameter is arranged to be rebuild, it can be with
For the 10 minutes time of cutting is arranged using the second PET image of the patient of data cutting function acquisition low dosage.
The foundation of Image Enhancement Based type is described in detail below.
Specifically, as shown in figure 3, step S200, includes the following steps;
S210, generation confrontation neural network is built, generating confrontation neural network includes generation module and discrimination module;
S220, using the first MIP image as label data, exercised supervision study using generation module to the second MIP image,
Obtain sub-picture;
S230, sub-picture and the first MIP image are differentiated using discrimination module, obtains differentiating result;
S240, training is completed after differentiating that result meets termination condition, obtains image enhancement model.
In one embodiment, the generation confrontation neural network built includes at least generation module and discrimination module;It generates
Module and discrimination module are mutually indepedent, mutual Game Learnings.In this present embodiment, the portion G for generating confrontation neural network is improved
Point, so that the not simple noise signal of input, the second MIP image of two-dimensional low dosage.Using the first MIP image as
Label data, generate model generate sub-picture, discrimination model learning region it is mitogenetic at sub-picture and the first MIP image, generate mould
Type improves oneself according to the differentiation result of discrimination model, generates new sub-picture;If differentiating, result is unsatisfactory for terminating, and basis is sentenced
The feedback of other result generates model and continues to generate new sub-picture, discrimination model continue study distinguish newly-generated sub-picture and
First MIP image;Training is completed after differentiating that result meets termination condition, obtains image enhancement model.
In generating model, the loss function for generating model includes the sub-picture and first for generating confrontation neural network and generating
The two-part weighting of the first normal form of MIP image.In continuous learning process, using the two-part weighting of first normal form, make
Sub-picture and the first MIP image become closer to.
In discrimination model, the loss function of discrimination model can accurately identify that the first MIP image sentences sub-picture
Not;The value of the loss function output of discrimination model is a probability value between 0-1.In this present embodiment, it is arranged one to preset
Threshold value when the probability value of output is higher than preset threshold, differentiates that result is 1;Meet termination condition, obtains image enhancement model.It is defeated
When probability value out is lower than preset threshold, differentiate that result is 0;It is unsatisfactory for termination condition, generation model is fed back to and generates new pair
Image continues learning training.
In one embodiment, epoch number, the decline of Lai Shixian gradient are set.1 epoch is equal to using in training set
Whole sample trainings it is primary, that is to say, that the value of epoch is exactly that the first MIP image and the second MIP image are read several times,
Input generates in model and discrimination model.Second MIP image is input to generation model and obtained by the output for monitoring entire iterative process
To sub-picture;And monitor the output of two loss functions, it is through to meet termination condition.
It is further comprising the steps of in one embodiment:
After building generation confrontation neural network, the first MIP image and the second MIP image are carried out one-to-one
Shuffle processing ensures the instruction of image enhancement model so that the distribution and sequence of the first MIP image and the second MIP image are orderly
Practice.
As shown in figure 4, in one embodiment, it is further comprising the steps of on the basis of in Fig. 2 embodiment:
S400, output MIP enhancing image after, by MIP enhancing image be stored as auxiliary positioning with reference to figure.It can with reference to figure
To be the storage by MIP enhancing image into the Tag to DICOM pixel data, and be written on database and disk.To
Enhanced MIP enhancing image is provided, the format that MIP enhances image is DICOM image, and MIP can be enhanced to image as auxiliary
Help directly using with reference to figure for positioning.
It should be understood that although each step in the flow chart of Fig. 2-4 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 2-4
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
In one embodiment, as shown in figure 5, a kind of enhancement device of low dosage MIP image, device include obtaining module
100, module 200 and enhancing module 300 are established: module 100 is obtained, for obtaining first under full dose of radiation tracer
MIP image and corresponding first PET image rebuild the first MIP image and corresponding first PET image, obtain corresponding position
The second MIP image under low dosage radioactive tracer;Module 200 is established, for fighting neural network to second using generation
MIP image is learnt by the feedback of the first MIP image, establishes image enhancement model;Enhance module 300, works as obtaining
Current MIP image input picture enhancing model is carried out image increasing by the current MIP image under the radioactive tracer of preceding low dosage
By force, output MIP enhances image.
The MIP that the enhancement device of above-mentioned low dosage MIP image is capable of quick outputting high quality enhances image;So as to full
The MIP image of sufficient low dosage is as the quality requirement with reference to figure.
In one embodiment, module 100 is obtained, is also used to carry out not the first PET image according to default reconstruction parameter
With the reconstruction under dosage, corresponding second PET image is obtained;And the extraction of pixel value is carried out to the second PET image, obtain second
MIP image.
In one embodiment, it establishes module 200 to be also used to, builds generation confrontation neural network, generate confrontation nerve net
Network includes generation module and discrimination module;Using the first MIP image as label data, using generation module to the second MIP image
Exercise supervision study, obtains sub-picture;Sub-picture and the first MIP image are differentiated using discrimination module, obtain differentiating knot
Fruit;Training is completed after differentiating that result meets termination condition, obtains image enhancement model.
In one embodiment, it establishes module 200 to be also used to after obtaining differentiating result, if differentiating, result is unsatisfactory for terminating
Condition will then differentiate that result feeds back to generation module, carries out continuation supervised learning to the second MIP image using generation module, obtains
To new sub-picture.
As shown in fig. 6, in one embodiment, on the basis of Fig. 5 embodiment, going back memory module 400: memory module
400 for output MIP enhancing image after, by MIP enhancing image be stored as auxiliary positioning with reference to figure.
The specific restriction of enhancement device about low dosage MIP image may refer to above for low dosage MIP image
Enhancement Method restriction, details are not described herein.Modules in the enhancement device of above-mentioned low dosage MIP image can all or
It is realized by software, hardware and combinations thereof part.Above-mentioned each module can be embedded in the form of hardware or set independently of computer
It in processor in standby, can also be stored in a software form in the memory in computer equipment, in order to processor calling
Execute the corresponding operation of the above modules.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 7.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is used for storage model and the relevant data of model training.The network interface of the computer equipment be used for it is outer
The terminal in portion passes through network connection communication.To realize a kind of low dosage MIP image when the computer program is executed by processor
Enhancement Method.
It will be understood by those skilled in the art that structure shown in Fig. 7, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, the first MIP is rebuild
Image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;It utilizes
It generates confrontation neural network to learn the second MIP image by the feedback of the first MIP image, establishes image enhancement model;
The current MIP image under the radioactive tracer of low dosage is obtained, current MIP image input picture enhancing model is subjected to image
Enhancing, output MIP enhance image.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, the first MIP is rebuild
Image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;It utilizes
It generates confrontation neural network to learn the second MIP image by the feedback of the first MIP image, establishes image enhancement model;
The current MIP image under the radioactive tracer of low dosage is obtained, current MIP image input picture enhancing model is subjected to image
Enhancing, output MIP enhance image.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of Enhancement Method of low dosage MIP image, which is characterized in that the described method includes:
The first MIP image and corresponding first PET image under full dose of radiation tracer are obtained, the first MIP is rebuild
Image and corresponding first PET image obtain the second MIP image under the low dosage radioactive tracer of corresponding position;
Second MIP image is learnt by the feedback of the first MIP image using confrontation neural network is generated, establishes figure
Image intensifying model;
The current MIP image under the radioactive tracer of low dosage is obtained, current MIP image input described image is enhanced into model
Image enhancement is carried out, output MIP enhances image.
2. the method according to claim 1, wherein described rebuild first MIP image and corresponding first
PET image obtains the second MIP image under the low dosage radioactive tracer of corresponding position, comprising:
The reconstruction under various dose is carried out to the first PET image according to default reconstruction parameter, obtains corresponding second PET image;
And the extraction of pixel value is carried out to second PET image, obtain the second MIP image.
3. the method according to claim 1, wherein described fight neural network to described second using generation
MIP image is learnt by the feedback of the first MIP image, establishes image enhancement model, comprising:
Generation confrontation neural network is built, the generation confrontation neural network includes generation module and discrimination module;
Using first MIP image as label data, exercised supervision study using the generation module to the second MIP image,
Obtain sub-picture;
The sub-picture and the first MIP image are differentiated using the discrimination module, obtain differentiating result;
Training is completed after differentiating that result meets termination condition, obtains image enhancement model.
4. according to the method described in claim 3, it is characterized by further comprising:
After building generation confrontation neural network, first MIP image and the second MIP image are carried out one-to-one
Shuffle processing.
5. according to the method described in claim 3, it is characterized by further comprising:
After obtaining differentiating result, if differentiating, result is unsatisfactory for termination condition, will differentiate that result feeds back to the generation module,
Continuation supervised learning is carried out to the second MIP image using the generation module, obtains new sub-picture.
6. according to claim 1 to method described in 5 any one, which is characterized in that further include:
Output MIP enhancing image after, by MIP enhancing image be stored as auxiliary positioning with reference to figure.
7. a kind of enhancement device of low dosage MIP image, which is characterized in that described device include obtain module, establish module with
And enhancing module:
The acquisition module, for obtaining the first MIP image and corresponding first PET figure under full dose of radiation tracer
Picture rebuilds first MIP image and corresponding first PET image, obtains under the low dosage radioactive tracer of corresponding position
The second MIP image;
It is described to establish module, for passing through the anti-of the first MIP image to second MIP image using generation confrontation neural network
The study of feed row, establishes image enhancement model;
The enhancing module, the current MIP image under radioactive tracer for obtaining current low dosage, current MIP is schemed
As input described image enhancing model progress image enhancement, output MIP enhances image.
8. device according to claim 7, which is characterized in that go back memory module:
The memory module, for output MIP enhancing image after, by MIP enhancing image be stored as auxiliary positioning with reference to figure.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 6 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 6 is realized when being executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910504782.5A CN110264435B (en) | 2019-06-12 | 2019-06-12 | Method and device for enhancing low-dose MIP (MIP) image, computer equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910504782.5A CN110264435B (en) | 2019-06-12 | 2019-06-12 | Method and device for enhancing low-dose MIP (MIP) image, computer equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110264435A true CN110264435A (en) | 2019-09-20 |
CN110264435B CN110264435B (en) | 2022-05-17 |
Family
ID=67917725
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910504782.5A Active CN110264435B (en) | 2019-06-12 | 2019-06-12 | Method and device for enhancing low-dose MIP (MIP) image, computer equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110264435B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991636A (en) * | 2019-11-14 | 2020-04-10 | 东软医疗***股份有限公司 | Training method and device of generative confrontation network, image enhancement method and equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150201895A1 (en) * | 2012-08-31 | 2015-07-23 | The University Of Chicago | Supervised machine learning technique for reduction of radiation dose in computed tomography imaging |
CN108537794A (en) * | 2018-04-19 | 2018-09-14 | 上海联影医疗科技有限公司 | Medical image processing method, device and computer readable storage medium |
CN109035356A (en) * | 2018-07-05 | 2018-12-18 | 四川大学 | A kind of system and method based on PET pattern imaging |
CN109493951A (en) * | 2017-11-08 | 2019-03-19 | 上海联影医疗科技有限公司 | For reducing the system and method for dose of radiation |
-
2019
- 2019-06-12 CN CN201910504782.5A patent/CN110264435B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150201895A1 (en) * | 2012-08-31 | 2015-07-23 | The University Of Chicago | Supervised machine learning technique for reduction of radiation dose in computed tomography imaging |
CN109493951A (en) * | 2017-11-08 | 2019-03-19 | 上海联影医疗科技有限公司 | For reducing the system and method for dose of radiation |
CN108537794A (en) * | 2018-04-19 | 2018-09-14 | 上海联影医疗科技有限公司 | Medical image processing method, device and computer readable storage medium |
CN109035356A (en) * | 2018-07-05 | 2018-12-18 | 四川大学 | A kind of system and method based on PET pattern imaging |
Non-Patent Citations (1)
Title |
---|
XIAOLI LI ET AL.: "Low-Dose CT Image Denoising Based on Improved WGAN-gp", 《JOURNAL OF NEW MEDIA》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991636A (en) * | 2019-11-14 | 2020-04-10 | 东软医疗***股份有限公司 | Training method and device of generative confrontation network, image enhancement method and equipment |
CN110991636B (en) * | 2019-11-14 | 2023-07-04 | 东软医疗***股份有限公司 | Training method and device of generated countermeasure network, image enhancement method and device |
Also Published As
Publication number | Publication date |
---|---|
CN110264435B (en) | 2022-05-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20210374474A1 (en) | Method, apparatus, and electronic device for training neural network model | |
US11011275B2 (en) | System and method for diagnosing gastrointestinal neoplasm | |
WO2021164306A1 (en) | Image classification model training method, apparatus, computer device, and storage medium | |
US20190164012A1 (en) | State detection using machine-learning model trained on simulated image data | |
Gao et al. | Bone age assessment based on deep convolution neural network incorporated with segmentation | |
CN109767461A (en) | Medical image registration method, device, computer equipment and storage medium | |
KR102132375B1 (en) | Deep learning model based image diagnosis apparatus and method thereof | |
Ntinou et al. | A transfer learning approach to heatmap regression for action unit intensity estimation | |
JP2022540487A (en) | Image processing for skin analysis, system and method for visualizing skin analysis | |
US12008793B2 (en) | Object behavior analysis method, information display method, and electronic device | |
Cheng et al. | DDU-Net: A dual dense U-structure network for medical image segmentation | |
US11676295B2 (en) | Image contouring using spiking neural networks | |
Li et al. | Hdrnet: Single-image-based hdr reconstruction using channel attention cnn | |
Wang et al. | Light attention embedding for facial expression recognition | |
KR102157867B1 (en) | Apparatus and method for ultra-fast acquiring positron emission tomographic image based on deep learning | |
Hou et al. | Af-net: A medical image segmentation network based on attention mechanism and feature fusion | |
Mao et al. | Pseudo-labeling generative adversarial networks for medical image classification | |
CN110264435A (en) | Enhancement Method, device, computer equipment and the storage medium of low dosage MIP image | |
Bhise et al. | Plant disease detection using machine learning | |
Fu et al. | Total‐body low‐dose CT image denoising using a prior knowledge transfer technique with a contrastive regularization mechanism | |
Yu et al. | Self-supervised multi-task learning for medical image analysis | |
US20200302290A1 (en) | Method for feature data recalibration and apparatus thereof | |
CN116721032A (en) | Image restoration method and device, computer equipment and storage medium | |
Hsieh et al. | Based on DICOM RT structure and multiple loss function deep learning algorithm in organ segmentation of head and neck image | |
Wang et al. | A joint local spatial and global temporal CNN-Transformer for dynamic facial expression recognition |
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 | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 201807 Shanghai City, north of the city of Jiading District Road No. 2258 Applicant after: Shanghai Lianying Medical Technology Co., Ltd Address before: 201807 Shanghai City, north of the city of Jiading District Road No. 2258 Applicant before: SHANGHAI UNITED IMAGING HEALTHCARE Co.,Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |