CN109816650A - A kind of target area recognition methods and its system based on two-dimentional DSA image - Google Patents
A kind of target area recognition methods and its system based on two-dimentional DSA image Download PDFInfo
- Publication number
- CN109816650A CN109816650A CN201910066634.XA CN201910066634A CN109816650A CN 109816650 A CN109816650 A CN 109816650A CN 201910066634 A CN201910066634 A CN 201910066634A CN 109816650 A CN109816650 A CN 109816650A
- Authority
- CN
- China
- Prior art keywords
- image
- target area
- processed
- dsa
- dimentional
- 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
Abstract
The invention discloses a kind of target area recognition methods based on two-dimentional DSA image, comprising: determines the image to be processed in two dimension DSA image sequence;The potential target region in the image to be processed is identified according to the default feature of target area, obtains potential target area image;Connected domain analysis is carried out to the potential target area image, using the corresponding region of the connected domain as final target area.The present invention can more accurately analyze image, directly obtain the position of every frame objective area in image in two-dimentional DSA image sequence.
Description
Technical field
The invention belongs to field of image processing more particularly to a kind of target area recognition methods based on two-dimentional DSA image
And its system.
Background technique
Increasingly mature with image processing techniques, application of the image processing techniques in medical image is also increasingly wider
General, good positive effect can be played to field of medical imaging by handling medical image by image processing techniques.With blood vessel at
As for, mainly there are Magnetic Resonance Angiography (MRA), CT angiography (CTA) and number for the detection method of blood vessel at present
Word subtractive angiography (DSA), relative to Magnetic Resonance Angiography and CT angiography, digital subtraction angiography is for blood vessel
Imaging effect be more clear.
The basic principle of DSA (Digital Subtraction Angiography) imaging is respectively to injection contrast medium
By inspection position and be not injected into contrast medium by inspection position carry out angiography, to injection contrast medium by inspection position blood vessel
Radiography and the angiography by inspection position for being not injected into contrast medium carry out computer disposal.Computer is by two different blood vessels
The digital information of contrastographic picture is subtracted each other, and removal bone, muscle and other soft tissues, leave behind simple blood vessel image subtracts shadow figure
Picture is shown by display.By taking encephalic DSA blood vessel imaging as an example, DSA photography is carried out to intracranial vessel, is needed from a certain
Branch enters cranium blood vessel and squeezes into contrast medium (can be contrast agent), probably passes through 5~8 seconds, and contrast agent can reach blood from cranium blood vessel is entered
Pipe end.In contrast agent from during entering cranium blood vessel arrival blood vessel end, 20~30 frame of digital subtractive angiography figures are shot
Picture shows 20~30 frame of digital subtraction angiography images of shooting by display.
DSA image only shows intracranial vessel, if staff wants to directly obtain a certain area in DSA image
The position in domain, it is clear that existing technology can not directly show some region of position that staff wants.Work
Personnel need to observe the image in the DSA image sequence of shooting, but there are larger subjectivity and errors for such mode
Property.
Summary of the invention
In order to solve the above technical problems, the main purpose of the present invention is to provide a kind of targets based on two-dimentional DSA image
Area recognizing method and its system can not be directly displayed out with the image in solution in the prior art DSA image sequence and be worked
The position of target area needed for personnel, needs related personnel carefully to observe image, so as to cause there are larger subjectivity,
Error technical problem.
The technical scheme is that be accomplished by the following way:
A kind of target area recognition methods based on two-dimentional DSA image, comprising:
Determine the image to be processed in two dimension DSA image sequence;
The potential target region in the image to be processed is identified according to the default feature of target area, obtains potential target
Area image;
To the potential target area image carry out connected domain analysis, using the corresponding region of the connected domain as finally
Target area.
Preferably, the default feature according to target area identify the potential target region in the image to be processed it
Before, further includes:
The default feature of target area in two-dimentional DSA image sequence is marked;
The two-dimentional DSA image sequence of default signature with target area is input to neural network model, according to
The two-dimentional DSA image sequence training neural network model of default signature with target area, the neural network
Model includes convolutional neural networks model;
It is identified in the image to be processed by the convolutional neural networks model according to the default feature of target area
Potential target region.
Preferably, described to be processed according to the identification of the default feature of target area by the convolutional neural networks model
Before potential target region in image, further includes: the image to be processed is pre-processed, pretreatment image is obtained, institute
State pretreatment specifically:
Image normalization processing is carried out to the image to be processed, described image normalized further comprises in coordinate
The heart, x-shearing normalization, scaling normalization or rotational normalization;
It is identified in the pretreatment image by the convolutional neural networks model according to the default feature of target area
Potential target region.
Preferably, belong to the probability value of target area in the potential target area image including potential target region.
Preferably, connected domain analysis is carried out to the potential target area image to specifically include:
Binary conversion treatment is carried out to the potential target area image according to the probability value, generates binary image;
Connected domain analysis is carried out to the binary image.
Preferably, binary conversion treatment is carried out to the potential target administrative division map according to the probability value, generates binary picture
The step of picture, specifically:
The second preset threshold is set, the size of the probability value Yu second preset threshold is compared, according to comparing result
Generate binary image;The comparing result includes that the probability value is less than greater than second preset threshold and the probability value
Second preset threshold.
Preferably, the image to be processed chosen in two dimension DSA image sequence, specifically:
Every frame image in the two dimension DSA image sequence is successively detected, when gray value in the image detected for the first time
When quantity less than the pixel of the first preset threshold is greater than present count magnitude, using the frame image as the of the image to be processed
One frame image;
When in the image detected for the last time gray value less than the pixel of the first preset threshold quantity be less than present count
When magnitude, using the frame image as the last frame image of the image to be processed;
Using the first frame image to the last frame image as the image to be processed.
A kind of target area identifying system based on two-dimentional DSA image, comprising:
Image chooses module, for determining the image to be processed in two dimension DSA image sequence;
Identification module identifies the potential target area in the image to be processed for the default feature according to target area
Domain obtains potential target area image;
Module is reprocessed, is used for the potential target area image connected domain analysis, by the corresponding area of the connected domain
Domain is as final target area.
Preferably, described image chooses module, is specifically used for:
Every frame image in the two dimension DSA image sequence is successively detected, when gray value in the image detected for the first time
When quantity less than the pixel of the first preset threshold is greater than present count magnitude, using the frame image as the of the image to be processed
One frame image;
When in the image detected for the last time gray value less than the pixel of the first preset threshold quantity be less than present count
When magnitude, using the frame image as the last frame image of the image to be processed;
By image to be processed described in conduct of the first frame image to the last frame image.
Preferably, further include preprocessing module, for pre-processing to the image to be processed, obtain pretreatment figure
Picture;The pretreatment includes that image normalization is handled, and described image normalized includes that coordinate centralization, x-shearing return
One changes, scales normalization or rotational normalization.
It preferably, further include neural metwork training module, for the default spy to target area in two-dimentional DSA image sequence
Sign is marked;The two-dimentional DSA image sequence of default signature with target area is input to neural network model, root
According to the two-dimentional DSA image sequence training neural network model of the default signature with target area, the nerve net
Network model includes convolutional neural networks model;It is identified by the convolutional neural networks model according to the default feature of target area
Potential target region in the pretreatment image.
Preferably, the identification module, specifically for passing through the convolutional neural networks model according to the pre- of target area
If feature identifies the potential target region in the pretreatment image, to obtain potential target area image.
Compared with the prior art, the present invention is based on the target area recognition methods of two-dimentional DSA image and its system at least to have
Have the advantage that or the utility model has the advantages that
The present invention is by pre-processing, identifying and locating again to several frame images chosen in two-dimentional DSA image sequence
Reason, can more accurately analyze image, final to determine connected domain, that is, target area position in image.By this hair
Bright treated two dimension DSA image sequence can be handled image to be processed according to the convolutional neural networks that training is completed,
Immediately arrive at processing result, i.e., in two dimension DSA image sequence in every frame image connected domain position.So as to so that staff
More easily, more directly image is observed.Meanwhile this method is also solved and is manually directly analyzed medical image
When, there are larger subjectivities, error technical problem.
Detailed description of the invention
In order to illustrate more clearly of this specification embodiment or technical solution in the prior art, below will to embodiment or
Attached drawing needed to be used in the description of the prior art is briefly described, it should be apparent that, the accompanying drawings in the following description is only
The some embodiments recorded in this specification, for those of ordinary skill in the art, in not making the creative labor property
Under the premise of, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of flow diagram of the recognition methods of image provided in an embodiment of the present invention;
Fig. 2 is a kind of structural schematic diagram of the identifying system of image provided in an embodiment of the present invention.
Specific embodiment
To keep the purposes, technical schemes and advantages of the application clearer, below in conjunction with the application specific embodiment and
Technical scheme is clearly and completely described in corresponding attached drawing.Obviously, described embodiment is only the application
A part of the embodiment, instead of all the embodiments.Based on this specification embodiment, those of ordinary skill in the art are not making
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present application.
In conjunction with attached drawing, invention is further explained.
As embodiment one, as shown in Figure 1, the flow diagram of the recognition methods for image of the present invention, this method are main
The following steps are included:
Step S100: determining the image to be processed in two dimension DSA image sequence, includes multiframe in two-dimentional DSA image sequence
Image therefrom selects image to be processed.The image to be processed is the image of blood vessel occur in two dimension DSA image sequence.
Step S200: the image to be processed is pre-processed, pretreatment image is obtained.The step is to go out to selection
The image to be processed of existing blood vessel is pre-processed, and pretreatment image is obtained.
Step S300: the potential target region in the pretreatment image is identified according to the default feature of target area, is obtained
To potential target area image.
Step S400: connected domain analysis is carried out to the potential target area image, by the corresponding region of the connected domain
As final target area, complete to the objective area in image to be processed chosen in the two dimension DSA image sequence
Identification.
Step in embodiment one realizes the target area for being included to image to be processed in two-dimentional DSA image sequence
Identification.The present invention by two-dimentional DSA image sequence determine several frame images pre-process, identify and reprocess,
More accurately image can be analyzed, it is final to determine connected domain, that is, target area position in image.Through the invention
Treated, and two dimension DSA image sequence can be handled image to be processed according to the convolutional neural networks that training is completed, directly
Connect and obtain processing result, i.e., in two dimension DSA image sequence in every frame image connected domain position.The position of connected domain is target
It the position in region can be from image directly so as to so that staff is easier, more directly observes image
To the region with preset target area with same characteristic features.Meanwhile this method also solve it is artificial directly to medical image into
When row analysis, there are larger subjectivities, error technical problem.
The present invention also provides embodiment two, the embodiment as preferred embodiment, be to the method in embodiment one into
Row further optimizes, in which:
Step S100: determine two dimension DSA image sequence in image to be processed the step of can there are many method, the following are
Two kinds of implementation methods that the present invention enumerates, certain step can also be other methods that technical solution of the present invention may be implemented.
Two different methods are described in detail separately below.
Method one:
Half of image sequence being located among two dimension DSA image sequence is chosen, as image to be processed.Entire two dimension DSA
Include several frame images in image sequence, is the image of not blood vessel in former frame images of two-dimentional DSA image sequence, works as blood vessel
Can generally occur at a quarter of two-dimentional DSA image sequence, disappear at 3/4ths, therefore from two-dimentional DSA image sequence
The frame at frame to 3/4ths at a quarter of column is as image to be processed.For example, two dimension DSA image sequence one shares 24
Frame image, can be from the image of the 7th frame image to the 18th frame be chosen as image to be processed.
It is determined according to the actual situation due to the image totalframes of two-dimentional DSA image sequence, the two dimension of different parts
The image totalframes of DSA image sequence is different.When the totalframes amount of image in two-dimentional DSA image sequence is larger and in two dimension
When only having the image of only a few frame blood vessel occur in DSA image sequence, the image conduct for the frame of blood vessel occur can be directly chosen
Image to be processed.For example, two dimension DSA image sequence one shares 32 frame images, the image of only the 28th frame to 30 frames has blood vessel to go out
It is existing, then the image of the 28th frame image to the 30th frame can be directly chosen as image to be processed.
This kind of method is not needed by carrying out other image procossings to the image in two-dimentional DSA image sequence, can be quick
Ground directly can be used as image to be processed to which image in determining two dimension DSA image sequence.
Method two:
Every frame image in the two dimension DSA image sequence is successively detected, when gray value is pre- less than first in described image
If the quantity of the pixel of threshold value is greater than present count magnitude, using the frame image as the first frame image of image to be processed.For example,
For the gray level of every frame image generally between 0-4095, image size is 512*512, the first preset threshold in DSA image sequence
1200 are set as, present count magnitude is set as 1000.It is detected since the first frame image of DSA image sequence, when detecting certain
When the quantity of pixel of the gray value less than 1200 is greater than 1000 in one frame image, then it is assumed that there is blood vessel appearance in the frame image, it will
First frame image of the frame image as image to be processed.
When the quantity that pixel is less than preset threshold in described image is less than present count magnitude, using the frame image as wait locate
Manage the last frame image of image.Since the tail vein of brain is close to skull, can from two-dimentional DSA image sequence last
Frame image detects forward, a frame image of blood vessel can be detected as the last frame figure of image to be processed last frame
Picture.Can also be since first frame image, every frame image after sequence detection first frame image, when detecting pixel in image
When quantity less than preset threshold is less than present count magnitude, using the frame image as the last frame image of image to be processed.
For example, whether have the method for blood vessel to be similarly in detection image ought detect that gray value is less than in a certain frame image
When the quantity of 1200 pixel is less than 1000, then it is assumed that have blood vessel appearance in the frame image, using the frame image as figure to be processed
The last frame image of picture.
Using the first frame image to the last frame image as the image to be processed, regardless of first frame image with
The quantity that pixel is less than preset threshold in intermediate frame image between last frame image is above or below present count magnitude,
As image to be processed.
This method corresponds to the quantity of the gray value of preset threshold by pixel in every frame image to two-dimentional DSA image sequence
It is compared with present count magnitude, determines whether each frame image belongs to image to be processed.Due in two-dimentional DSA image sequence
The similarity-rough set of the image of consecutive frame is high, so needing to detect every frame image, this method can be determined more accurately
Whether the image in two-dimentional DSA image sequence belongs to image to be processed, to improve the figure determined in two dimension DSA image sequence
It seem the accuracy of no image to be processed.
Further, it is also possible to method one and method two are combined to determine the figure to be processed in two-dimentional DSA image sequence
Picture.For example, can first pass through method one chooses the image for having blood vessel to occur in two-dimentional DSA image sequence, in order to more accurately examine
Whether there is blood vessel to occur in the image chosen in survey method one or further choose the display clearer image of blood vessel, so as to true
Determine the target area in image, then the image chosen by method one is detected by method two.It is combined by method one
This mode of method two determines the image to be processed in two dimension DSA image sequence, on the basis of fast selecting image to be processed
On, and increase the accuracy for determining image to be processed.
Image to be processed in two dimension DSA image sequence can also be determined by the method not enumerated in this specification, it is same
Sample belongs to the scope of protection of the present invention.
Step S200: pretreated step is carried out to every frame image in the image to be processed determined in step S100.
It specifically can be the step of image normalization processing is carried out to every frame image in image to be processed, image normalization
Processing can be one or more of coordinate centralization, x-shearing normalization, scaling normalization and rotational normalization,
It is to meet image as defined in pre-treatment step by image procossing to be processed.For example, image size as defined in pre-treatment step is
200mm*200mm, pel spacing are the image of 1mm, if it is 512mm*512mm that image to be processed, which is size, pel spacing is
The image of 0.5mm.By image to be processed is normalized in the step, first image to be processed can be zoomed in and out
Normalized, is 256mm*256mm by image down to be processed to size, and pel spacing is the image of 1mm.
Then it is being cut out, is being 256mm*256mm by size, pel spacing is that the image of 1mm is each up and down
28 pixels are cut out, obtained image is that size is 200mm*200mm, and pel spacing is the image of 1mm.It is obtained in the step
Image be pretreatment image, equally include the image of quantity identical as number of image frames to be processed in pretreatment image.May be used also
To carry out one or more of coordinate centralization, x-shearing normalization, rotational normalization etc. place to image to be processed
Reason, no longer enumerates, purpose is for that image is normalized, and can preferably determine target area herein
Domain.
By a series of transformation, (finding one group of parameter using the not bending moment of image can disappear for image normalization processing
The influence that image is converted except other transforming function transformation functions), original image to be processed is converted into corresponding sole criterion form (should
Canonical form image has invariant feature to translation, rotation, scaling equiaffine transformation).
Step S300: according to potential in pretreatment image obtained in the default feature identification step S200 of target area
Target area, obtains potential target area image, includes that potential target region belongs to target in the potential target area image
The probability value in region.The step specifically may is that
The label that default feature is first carried out to the target area of two-dimentional DSA image sequence, will be default with target area
The two-dimentional DSA image sequence of signature is input to neural network model, according to the default signature with target area
The two-dimentional DSA image sequence training neural network model, target area are staff before manually according to two-dimentional DSA image
The region that sequence judges.Two-dimentional DSA image sequence in the technical program is also needed to the image with same characteristic features
It is handled, identifies and preset the region that feature has same characteristic features with target area.Wherein, the neural network model includes
Convolutional neural networks model can also be that the nerve with neural network identical function in the technical program may be implemented in other certainly
Network model.Using pretreatment image as the input of neural network model, potential target region is obtained by neural network model
Image, the feature of the potential target area image obtained by the neural network model and the image for training the neural network to use
In target area default feature it is identical.Label has region or spy in the image that training neural network uses
Sign, the neural network that training is completed export same specific region or spy after inputting information, according to the rule that training is completed
The information of sign.The output to pretreatment image processing can be directly obtained as a result, i.e. potential target by the neural network model
Area image, includes the probability value that potential target region belongs to target area in the potential target area image, range 0-1 it
Between.The step is to be respectively processed in pretreatment image to each frame image for including.
Step S400: it is raw that binary conversion treatment is carried out to the potential target area image according to the probability value in step S300
At binary image, connected domain analysis is carried out to the binary image, using the corresponding region of the connected domain as finally
Target area, to complete the step of the identification to the objective area in image to be processed chosen in the two dimension DSA image sequence
Suddenly, it specifically may is that
Binary conversion treatment is carried out to potential target area image, the second preset threshold is preset, compares the probability value
With the size of second preset threshold, binary image is generated according to comparing result.The comparing result includes the probability
Value is greater than second preset threshold and the probability value is less than second preset threshold, and probability value is greater than described second
The probability value of preset threshold is set as 1 in binary image, white or black in corresponding binary image.Probability value is less than
The probability value of second preset threshold is set as 0 in binary image, black or white in corresponding binary image, thus
Generate binary image.
For example, the second preset threshold is 0.8, then by probability value in potential target area image greater than 0.8 in binaryzation
It is set as 1 in image, is otherwise 0.
Connected domain analysis is carried out to obtained binary image, marks all connected domains, it can be by 1 in binary image
Corresponding probability value carries out connected domain analysis, and the corresponding region of connected domain that 1 corresponding probability value is generated is as finally
Target area.The same step is also to reprocess respectively to every frame image in potential target area image.
Method provided by the invention realizes the identification to target area, and this method can pass through corresponding system or dress
It sets to complete.
As shown in Fig. 2, the present invention also provides a kind of target area identifying system based on two-dimentional DSA image, the system
It specifically includes that
Image chooses module 1, for choosing the image to be processed in two-dimentional DSA image sequence.Specifically, which can be with
For choosing half of image sequence being located among two dimension DSA image sequence, as image to be processed.
The module can be also used for successively detecting every frame image in the two dimension DSA image sequence, when pixel in image
When quantity less than the first preset threshold is greater than present count magnitude, using the frame image as the first frame image of image to be processed.
When the quantity that pixel is less than preset threshold in image is less than present count magnitude, using the frame image as the last of image to be processed
One frame image.Using first frame image to last frame figure as image to be processed.
Preprocessing module 2 obtains pretreatment image for pre-processing to image.Specifically, which can be used for
Image normalization processing is carried out to described image.
Neural metwork training module 3, the two-dimentional DSA image sequence for that will have goal-selling region signature input
To neural network model, according to the two-dimentional DSA image sequence training neural network with goal-selling region signature
Model, the neural network model include convolutional neural networks model.
Identification module 4 identifies the potential target area in the pretreatment image for the default feature according to target area
Domain obtains potential target area image;It include that potential target region belongs to target area in the potential target area image
Probability value.
Module 5 is reprocessed, for carrying out binary conversion treatment generation to the potential target administrative division map according to the probability value
Binary image carries out connected domain analysis to the binary image, using the corresponding region of the connected domain as final mesh
Region is marked, to complete the identification to the objective area in image to be processed chosen in the two dimension DSA image sequence.
The step in the above method may be implemented in the system, reaches same technical effect.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device,
For electronic equipment, nonvolatile computer storage media embodiment, since it is substantially similar to the method embodiment, so description
It is fairly simple, the relevent part can refer to the partial explaination of embodiments of method.
Device that this specification embodiment provides, electronic equipment, nonvolatile computer storage media with method are corresponding
, therefore, device, electronic equipment, nonvolatile computer storage media also have the Advantageous effect similar with corresponding method
Fruit, since the advantageous effects of method being described in detail above, which is not described herein again corresponding intrument,
The advantageous effects of electronic equipment, nonvolatile computer storage media.
In the 1990s, the improvement of a technology can be distinguished clearly be on hardware improvement (for example,
Improvement to circuit structures such as diode, transistor, switches) or software on improvement (improvement for method flow).So
And with the development of technology, the improvement of current many method flows can be considered as directly improving for hardware circuit.
Designer nearly all obtains corresponding hardware circuit by the way that improved method flow to be programmed into hardware circuit.Cause
This, it cannot be said that the improvement of a method flow cannot be realized with hardware entities module.For example, programmable logic device
(Programmable Logic Device, PLD) (such as field programmable gate array (Field Programmable
GateArray, FPGA)) it is exactly such a integrated circuit, logic function determines device programming by user.By designing
Personnel, which voluntarily program, to come a digital display circuit " integrated " on a piece of PLD, designed without asking chip maker and
Make dedicated IC chip.Moreover, nowadays, substitution manually makes IC chip, and this programming also changes mostly
It is realized with " logic compiler (logic compiler) " software, software compiler phase used when it writes with program development
It is similar, and the source code before compiling also write by handy specific programming language, this is referred to as hardware description language
(Hardware Description Language, HDL), and HDL is also not only a kind of, but there are many kind, such as ABEL
(Advanced Boolean Expression Language)、AHDL(Altera Hardware Description
Language)、Confluence、CUPL(Cornell UniversityProgramming Language)、HDCal、JHDL
(Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby
Hardware Description Language) etc., VHDL (Very-High-Speed is most generally used at present
Integrated Circuit Hardware Description Language) and Verilog.Those skilled in the art also answer
This understands, it is only necessary to method flow slightly programming in logic and is programmed into integrated circuit with above-mentioned several hardware description languages,
The hardware circuit for realizing the logical method process can be readily available.
Controller can be implemented in any suitable manner, for example, controller can take such as microprocessor or processing
The computer for the computer readable program code (such as software or firmware) that device and storage can be executed by (micro-) processor can
Read medium, logic gate, switch, specific integrated circuit (Application Specific Integrated Circuit,
ASIC), the form of programmable logic controller (PLC) and insertion microcontroller, the example of controller includes but is not limited to following microcontroller
Device: ARC 625D, AtmelAT91SAM, Microchip PIC18F26K20 and Silicone Labs C8051F320 are deposited
Memory controller is also implemented as a part of the control logic of memory.It is also known in the art that in addition to
Pure computer readable program code mode is realized other than controller, can be made completely by the way that method and step is carried out programming in logic
Controller is obtained to come in fact in the form of logic gate, switch, specific integrated circuit, programmable logic controller (PLC) and insertion microcontroller etc.
Existing identical function.Therefore this controller is considered a kind of hardware component, and to including for realizing various in it
The device of function can also be considered as the structure in hardware component.Or even, it can will be regarded for realizing the device of various functions
For either the software module of implementation method can be the structure in hardware component again.
System, device, module or the unit that above-described embodiment illustrates can specifically realize by computer chip or entity,
Or it is realized by the product with certain function.It is a kind of typically to realize that equipment is computer.Specifically, computer for example may be used
Think personal computer, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media play
It is any in device, navigation equipment, electronic mail equipment, game console, tablet computer, wearable device or these equipment
The combination of equipment.
For convenience of description, it is divided into various units when description apparatus above with function to describe respectively.Certainly, implementing this
The function of each unit can be realized in the same or multiple software and or hardware when specification one or more embodiment.
It should be understood by those skilled in the art that, this specification embodiment can provide as method, system or computer program
Product.Therefore, this specification embodiment can be used complete hardware embodiment, complete software embodiment or combine software and hardware
The form of the embodiment of aspect.Moreover, it wherein includes that computer is available that this specification embodiment, which can be used in one or more,
It is real in the computer-usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) of program code
The form for the computer program product applied.
This specification is referring to the method, equipment (system) and computer program product according to this specification embodiment
Flowchart and/or the block diagram describes.It should be understood that can be realized by computer program instructions every in flowchart and/or the block diagram
The combination of process and/or box in one process and/or box and flowchart and/or the block diagram.It can provide these computers
Processor of the program instruction to general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices
To generate a machine, so that generating use by the instruction that computer or the processor of other programmable data processing devices execute
In the dress for realizing the function of specifying in one or more flows of the flowchart and/or one or more blocks of the block diagram
It sets.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described want
There is also other identical elements in the process, method of element, commodity or equipment.
This specification can describe in the general context of computer-executable instructions executed by a computer, such as journey
Sequence module.Generally, program module include routines performing specific tasks or implementing specific abstract data types, programs, objects,
Component, data structure etc..Specification can also be practiced in a distributed computing environment, in these distributed computing environments,
By executing task by the connected remote processing devices of communication network.In a distributed computing environment, program module can
To be located in the local and remote computer storage media including storage equipment.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.
More than, it is merely preferred embodiments of the present invention, but the protection scope invented is not limited thereto, it is any ripe
Know those skilled in the art in the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of, should all contain
Lid is within protection scope of the present invention.Therefore, the scope of protection of the invention shall be subject to the scope of protection specified in the patent claim.
Claims (12)
1. a kind of target area recognition methods based on two-dimentional DSA image characterized by comprising
Determine the image to be processed in two dimension DSA image sequence;
The potential target region in the image to be processed is identified according to the default feature of target area, obtains potential target region
Image;
Connected domain analysis is carried out to the potential target area image, using the corresponding region of the connected domain as final target
Region.
2. target area recognition methods according to claim 1, which is characterized in that in the default feature according to target area
Before identifying the potential target region in the image to be processed, further includes:
The default feature of target area in two-dimentional DSA image sequence is marked;
The two-dimentional DSA image sequence of default signature with target area is input to neural network model, according to having
The two-dimentional DSA image sequence training neural network model of the default signature of target area, the neural network model
Including convolutional neural networks model;
It is identified by the convolutional neural networks model according to the default feature of target area potential in the image to be processed
Target area.
3. target area recognition methods according to claim 2, which is characterized in that passing through the convolutional neural networks mould
Type is identified according to the default feature of target area before the potential target region in the image to be processed, further includes: to described
Image to be processed is pre-processed, and pretreatment image, the pretreatment are obtained specifically:
Image normalization processing is carried out to the image to be processed, described image normalized further comprises coordinate center
Change, x-shearing normalization, scaling normalizes or rotational normalization;
It is identified by the convolutional neural networks model according to the default feature of target area potential in the pretreatment image
Target area.
4. target area recognition methods according to claim 3, which is characterized in that wrapped in the potential target area image
Include the probability value that potential target region belongs to target area.
5. recognition methods according to claim 4, which is characterized in that carry out connected domain to the potential target area image
Analysis specifically includes:
Binary conversion treatment is carried out to the potential target area image according to the probability value, generates binary image;
Connected domain analysis is carried out to the binary image.
6. target area recognition methods according to claim 5, which is characterized in that according to the probability value to described potential
The step of target area figure carries out binary conversion treatment, generates binary image, specifically:
The second preset threshold is set, the size of the probability value Yu second preset threshold is compared, is generated according to comparing result
Binary image;The comparing result includes that the probability value is greater than second preset threshold and the probability value less than described
Second preset threshold.
7. target area recognition methods according to claim 6, which is characterized in that the selection two dimension DSA image sequence
In image to be processed, specifically:
Every frame image in the two dimension DSA image sequence is successively detected, when gray value is less than in the image detected for the first time
When the quantity of the pixel of first preset threshold is greater than present count magnitude, using the frame image as the first frame of the image to be processed
Image;
When in the image detected for the last time gray value less than the pixel of the first preset threshold quantity be less than present count magnitude
When, using the frame image as the last frame image of the image to be processed;
Using the first frame image to the last frame image as the image to be processed.
8. a kind of target area identifying system based on two-dimentional DSA image characterized by comprising
Image chooses module, for determining the image to be processed in two dimension DSA image sequence;
Identification module identifies the potential target region in the image to be processed for the default feature according to target area, obtains
To potential target area image;
Module is reprocessed, for the potential target area image connected domain analysis, the corresponding region of the connected domain to be made
For final target area.
9. target area identifying system according to claim 8, which is characterized in that described image chooses module, specific to use
In:
Every frame image in the two dimension DSA image sequence is successively detected, when gray value is less than in the image detected for the first time
When the quantity of the pixel of first preset threshold is greater than present count magnitude, using the frame image as the first frame of the image to be processed
Image;
When in the image detected for the last time gray value less than the pixel of the first preset threshold quantity be less than present count magnitude
When, using the frame image as the last frame image of the image to be processed;
By image to be processed described in conduct of the first frame image to the last frame image.
10. target area identifying system according to claim 9, which is characterized in that further include preprocessing module, for pair
The image to be processed is pre-processed, and pretreatment image is obtained;The pretreatment includes that image normalization is handled, described image
Normalized includes coordinate centralization, x-shearing normalization, scaling normalization or rotational normalization.
11. target area identifying system according to claim 10, which is characterized in that further include neural metwork training mould
Block is marked for the default feature to target area in two-dimentional DSA image sequence;The default feature of target area will be had
The two-dimentional DSA image sequence of label is input to neural network model, according to the two dimension of the default signature with target area
The DSA image sequence training neural network model, the neural network model includes convolutional neural networks model;By described
Convolutional neural networks model identifies the potential target region in the pretreatment image according to the default feature of target area.
12. target area identifying system according to claim 11, which is characterized in that the identification module is specifically used for
The potential target in the pretreatment image is identified according to the default feature of target area by the convolutional neural networks model
Region, to obtain potential target area image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910066634.XA CN109816650B (en) | 2019-01-24 | 2019-01-24 | Target area identification method and system based on two-dimensional DSA image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910066634.XA CN109816650B (en) | 2019-01-24 | 2019-01-24 | Target area identification method and system based on two-dimensional DSA image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109816650A true CN109816650A (en) | 2019-05-28 |
CN109816650B CN109816650B (en) | 2022-11-25 |
Family
ID=66604936
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910066634.XA Active CN109816650B (en) | 2019-01-24 | 2019-01-24 | Target area identification method and system based on two-dimensional DSA image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109816650B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110503642A (en) * | 2019-08-23 | 2019-11-26 | 强联智创(北京)科技有限公司 | A kind of localization method and system based on DSA image |
CN110517244A (en) * | 2019-08-23 | 2019-11-29 | 首都医科大学宣武医院 | A kind of localization method and system based on DSA image |
CN110634129A (en) * | 2019-08-23 | 2019-12-31 | 首都医科大学宣武医院 | Positioning method and system based on DSA image |
CN111062852A (en) * | 2019-12-16 | 2020-04-24 | 百度国际科技(深圳)有限公司 | Map rendering method and device, electronic equipment and storage medium |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101127117A (en) * | 2007-09-11 | 2008-02-20 | 华中科技大学 | Method for segmenting blood vessel data using serial DSA image |
CN103003880A (en) * | 2010-07-26 | 2013-03-27 | 皇家飞利浦电子股份有限公司 | Determining representative images for a video |
CN103135755A (en) * | 2011-12-02 | 2013-06-05 | 深圳泰山在线科技有限公司 | Interaction system and interactive method |
CN103385692A (en) * | 2012-05-07 | 2013-11-13 | 佳能株式会社 | Image processing apparatus and image processing method |
CN105787481A (en) * | 2016-04-05 | 2016-07-20 | 湖南人文科技学院 | Target detection algorithm based on targeted potential areas analysis and application thereof |
CN106940816A (en) * | 2017-03-22 | 2017-07-11 | 杭州健培科技有限公司 | Connect the CT image Lung neoplasm detecting systems of convolutional neural networks entirely based on 3D |
CN107239727A (en) * | 2016-12-07 | 2017-10-10 | 北京深鉴智能科技有限公司 | Gesture identification method and system |
CN108052957A (en) * | 2017-11-07 | 2018-05-18 | 聊城大学 | A kind of spacecraft target method for quickly identifying |
WO2018095082A1 (en) * | 2016-11-28 | 2018-05-31 | 江苏东大金智信息系统有限公司 | Rapid detection method for moving target in video monitoring |
WO2018137612A1 (en) * | 2017-01-25 | 2018-08-02 | 武汉极目智能技术有限公司 | Road sign recognition method and system |
CN108596891A (en) * | 2018-04-23 | 2018-09-28 | 中国计量大学 | A kind of method of counting towards multiple types mixing silk cocoon |
CN108899075A (en) * | 2018-06-28 | 2018-11-27 | 众安信息技术服务有限公司 | A kind of DSA image detecting method, device and equipment based on deep learning |
CN109034266A (en) * | 2018-08-16 | 2018-12-18 | 新智数字科技有限公司 | A kind of target image detection method, apparatus and system |
CN109044555A (en) * | 2018-08-30 | 2018-12-21 | 姜润浩 | Intervention method establishes after dog autologous blood clots embolic cerebral infarction thrombolysis intracranial hemorrhage transformation model and its application |
CN109218833A (en) * | 2017-06-29 | 2019-01-15 | 深圳泰山体育科技股份有限公司 | The method and system that control television image is shown |
CN109255784A (en) * | 2018-09-13 | 2019-01-22 | 北京市商汤科技开发有限公司 | Image processing method and device, electronic equipment and storage medium |
-
2019
- 2019-01-24 CN CN201910066634.XA patent/CN109816650B/en active Active
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101127117A (en) * | 2007-09-11 | 2008-02-20 | 华中科技大学 | Method for segmenting blood vessel data using serial DSA image |
CN103003880A (en) * | 2010-07-26 | 2013-03-27 | 皇家飞利浦电子股份有限公司 | Determining representative images for a video |
CN103135755A (en) * | 2011-12-02 | 2013-06-05 | 深圳泰山在线科技有限公司 | Interaction system and interactive method |
CN103385692A (en) * | 2012-05-07 | 2013-11-13 | 佳能株式会社 | Image processing apparatus and image processing method |
CN105787481A (en) * | 2016-04-05 | 2016-07-20 | 湖南人文科技学院 | Target detection algorithm based on targeted potential areas analysis and application thereof |
WO2018095082A1 (en) * | 2016-11-28 | 2018-05-31 | 江苏东大金智信息系统有限公司 | Rapid detection method for moving target in video monitoring |
CN107239727A (en) * | 2016-12-07 | 2017-10-10 | 北京深鉴智能科技有限公司 | Gesture identification method and system |
WO2018137612A1 (en) * | 2017-01-25 | 2018-08-02 | 武汉极目智能技术有限公司 | Road sign recognition method and system |
CN106940816A (en) * | 2017-03-22 | 2017-07-11 | 杭州健培科技有限公司 | Connect the CT image Lung neoplasm detecting systems of convolutional neural networks entirely based on 3D |
CN109218833A (en) * | 2017-06-29 | 2019-01-15 | 深圳泰山体育科技股份有限公司 | The method and system that control television image is shown |
CN108052957A (en) * | 2017-11-07 | 2018-05-18 | 聊城大学 | A kind of spacecraft target method for quickly identifying |
CN108596891A (en) * | 2018-04-23 | 2018-09-28 | 中国计量大学 | A kind of method of counting towards multiple types mixing silk cocoon |
CN108899075A (en) * | 2018-06-28 | 2018-11-27 | 众安信息技术服务有限公司 | A kind of DSA image detecting method, device and equipment based on deep learning |
CN109034266A (en) * | 2018-08-16 | 2018-12-18 | 新智数字科技有限公司 | A kind of target image detection method, apparatus and system |
CN109044555A (en) * | 2018-08-30 | 2018-12-21 | 姜润浩 | Intervention method establishes after dog autologous blood clots embolic cerebral infarction thrombolysis intracranial hemorrhage transformation model and its application |
CN109255784A (en) * | 2018-09-13 | 2019-01-22 | 北京市商汤科技开发有限公司 | Image processing method and device, electronic equipment and storage medium |
Non-Patent Citations (4)
Title |
---|
NISREEN SULAYMAN 等: "Semi-automatic detection and segmentation algorithm of saccular aneurysms in 2D cerebral DSA images", 《THE EGYPTIAN JOURNAL OF RADIOLOGY AND NUCLEAR MEDICINE》 * |
徐珊 等: "基于连通域分析的仪表数字区域定位方法研究", 《科研信息化技术与应用》 * |
方智文 等: "深度图像的目标潜在区域提取算法", 《信号处理》 * |
韩光辉 等: "肺部CT图像病变区域检测方法", 《自动化学报》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110503642A (en) * | 2019-08-23 | 2019-11-26 | 强联智创(北京)科技有限公司 | A kind of localization method and system based on DSA image |
CN110517244A (en) * | 2019-08-23 | 2019-11-29 | 首都医科大学宣武医院 | A kind of localization method and system based on DSA image |
CN110634129A (en) * | 2019-08-23 | 2019-12-31 | 首都医科大学宣武医院 | Positioning method and system based on DSA image |
CN110503642B (en) * | 2019-08-23 | 2021-10-29 | 强联智创(北京)科技有限公司 | Positioning method and system based on DSA image |
CN111062852A (en) * | 2019-12-16 | 2020-04-24 | 百度国际科技(深圳)有限公司 | Map rendering method and device, electronic equipment and storage medium |
CN111062852B (en) * | 2019-12-16 | 2023-10-17 | 阿波罗智联(北京)科技有限公司 | Map rendering method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN109816650B (en) | 2022-11-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109816650A (en) | A kind of target area recognition methods and its system based on two-dimentional DSA image | |
US10929708B2 (en) | Deep learning network for salient region identification in images | |
CN109101919B (en) | Method and apparatus for generating information | |
CN110517780A (en) | A kind of aneurysm rupture methods of risk assessment and system | |
CN109583364A (en) | Image-recognizing method and equipment | |
RU2513905C2 (en) | Image classification based on image segmentation | |
CN110517242A (en) | A kind of aneurysmal analysis method and device | |
US20230015384A1 (en) | Method, Apparatus, Storage Medium and Processor for Medical Image Auto-segmentation | |
CN117333529B (en) | Template matching-based vascular ultrasonic intima automatic measurement method and system | |
Wang et al. | Segmentation of ultrasound brachial plexus based on U-Net | |
CN116030247B (en) | Medical image sample generation method and device, storage medium and electronic equipment | |
CN110503642A (en) | A kind of localization method and system based on DSA image | |
CN109377504A (en) | A kind of entocranial artery blood-vessel image dividing method and system | |
CN114612484B (en) | Retina OCT image segmentation method based on unsupervised learning | |
CN109472803A (en) | A kind of entocranial artery blood vessel segmentation method and system | |
CN113379770B (en) | Construction method of nasopharyngeal carcinoma MR image segmentation network, image segmentation method and device | |
CN112734726B (en) | Angiography typing method, angiography typing device and angiography typing equipment | |
CN114972241A (en) | Fetus ultrasonic section quantitative quality control assessment method and device based on residual error neural network | |
CN114037830A (en) | Training method for enhanced image generation model, image processing method and device | |
CN110517244A (en) | A kind of localization method and system based on DSA image | |
CN109671065A (en) | A kind of method and system for taking bolt to judge based on head CT images | |
Schumacher et al. | Weakly supervised pancreas segmentation based on class activation maps | |
CN117036830B (en) | Tumor classification model training method and device, storage medium and electronic equipment | |
Patil et al. | Early detection of hemiplegia by analyzing the gait characteristics and walking patterns using convolutional neural networks | |
US20230298326A1 (en) | Image augmentation method, electronic device and readable storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |