CN109816650B - Target area identification method and system based on two-dimensional DSA image - Google Patents
Target area identification method and system based on two-dimensional DSA image Download PDFInfo
- Publication number
- CN109816650B CN109816650B CN201910066634.XA CN201910066634A CN109816650B CN 109816650 B CN109816650 B CN 109816650B CN 201910066634 A CN201910066634 A CN 201910066634A CN 109816650 B CN109816650 B CN 109816650B
- Authority
- CN
- China
- Prior art keywords
- image
- preset
- processed
- target area
- frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses a target region identification method based on a two-dimensional DSA image, which comprises the following steps: determining an image to be processed in a two-dimensional DSA image sequence; identifying a potential target area in the image to be processed according to preset characteristics of the target area to obtain a potential target area image; and analyzing the connected domain of the potential target area image, and taking the area corresponding to the connected domain as a final target area. The invention can more accurately analyze the images and directly obtain the position of the target area in each frame of image in the two-dimensional DSA image sequence.
Description
Technical Field
The invention belongs to the field of image processing, and particularly relates to a target region identification method and a target region identification system based on a two-dimensional DSA image.
Background
With the increasing maturity of image processing technology, the application of image processing technology in medical images is also more and more extensive, and the medical image processing technology can play a good positive role in the field of medical images. Taking blood vessel imaging as an example, the current methods for detecting blood vessels mainly include Magnetic Resonance Angiography (MRA), CT angiography (CTA), and Digital Subtraction Angiography (DSA), which has a clearer imaging effect on blood vessels compared with magnetic resonance angiography and CT angiography.
The basic principle of DSA (Digital microscopy) imaging is to perform Angiography on a contrast agent-injected examined region and an examined region without contrast agent injection, respectively, and to perform computer processing on the Angiography of the contrast agent-injected examined region and the Angiography of the examined region without contrast agent injection. The computer subtracts the digital information of two different angiographic images, removes bones, muscles and other soft tissues, only leaves the subtraction image of the pure blood vessel image, and displays the subtraction image through the display. Taking intracranial DSA vessel imaging as an example, DSA radiography of an intracranial vessel requires injecting a contrast agent (which may be a contrast agent) from one intracranial injection vessel, and the contrast agent can reach the end of the vessel from the intracranial injection vessel after about 5-8 seconds. During the process that the contrast agent reaches the tail end of the blood vessel from the intracranial blood vessel, 20-30 frames of digital subtraction angiography images are shot, and the shot 20-30 frames of digital subtraction angiography images are displayed through a display.
The DSA image only displays the intracranial blood vessel, and if a worker wants to directly obtain the position of a certain region in the DSA image, the prior art obviously cannot directly display the position of the certain region wanted by the worker. Workers need to observe images in a captured DSA image sequence, but the method has large subjectivity and error.
Disclosure of Invention
In order to solve the above technical problems, a primary object of the present invention is to provide a method and a system for identifying a target region based on a two-dimensional DSA image, so as to solve the technical problems that in the prior art, images in a DSA image sequence cannot directly display the position of a target region required by a worker, and the worker needs to carefully observe the images, which results in large subjectivity and error.
The technical scheme of the invention is realized by the following modes:
a target region identification method based on two-dimensional DSA images comprises the following steps:
determining an image to be processed in a two-dimensional DSA image sequence;
identifying a potential target area in the image to be processed according to preset characteristics of the target area to obtain a potential target area image;
and analyzing the connected domain of the potential target area image, and taking the area corresponding to the connected domain as a final target area.
Preferably, before the potential target region in the image to be processed is identified according to the preset features of the target region, the method further includes:
marking preset characteristics of a target region in a two-dimensional DSA image sequence;
inputting a two-dimensional DSA image sequence with a preset feature marker of a target region into a neural network model, and training the neural network model according to the two-dimensional DSA image sequence with the preset feature marker of the target region, wherein the neural network model comprises a convolutional neural network model;
and identifying potential target areas in the image to be processed according to preset characteristics of the target areas through the convolutional neural network model.
Preferably, before the potential target region in the image to be processed is identified according to the preset features of the target region through the convolutional neural network model, the method further includes: preprocessing the image to be processed to obtain a preprocessed image, wherein the preprocessing specifically comprises the following steps:
carrying out image normalization processing on the image to be processed, wherein the image normalization processing further comprises coordinate centralization, x-sharpening normalization, scaling normalization or rotation normalization;
and identifying potential target areas in the preprocessed image according to preset characteristics of the target areas through the convolutional neural network model.
Preferably, the potential target region image includes a probability value that the potential target region belongs to the target region.
Preferably, the performing connected component analysis on the potential target area image specifically includes:
carrying out binarization processing on the potential target area image according to the probability value to generate a binarized image;
and carrying out connected domain analysis on the binary image.
Preferably, the step of performing binarization processing on the potential target region map according to the probability value to generate a binarized image specifically includes:
setting a second preset threshold, comparing the probability value with the second preset threshold, and generating a binary image according to a comparison result; the comparison result comprises that the probability value is greater than the second preset threshold value and the probability value is smaller than the second preset threshold value.
Preferably, the selecting an image to be processed in the two-dimensional DSA image sequence specifically includes:
sequentially detecting each frame of image in the two-dimensional DSA image sequence, and when the number of pixels with the gray scale value smaller than a first preset threshold value in the first detected image is larger than a preset number value, taking the frame of image as a first frame of image of the image to be processed;
when the number of the pixels with the gray values smaller than the first preset threshold value in the last detected image is smaller than a preset number value, taking the frame image as the last frame image of the image to be processed;
and taking the first frame image to the last frame image as the images to be processed.
A target region identification system based on two-dimensional DSA images, comprising:
the image selection module is used for determining an image to be processed in a two-dimensional DSA image sequence;
the identification module is used for identifying a potential target area in the image to be processed according to preset characteristics of the target area to obtain a potential target area image;
and the reprocessing module is used for analyzing the potential target area image connected domain and taking the area corresponding to the connected domain as a final target area.
Preferably, the image selecting module is specifically configured to:
sequentially detecting each frame of image in the two-dimensional DSA image sequence, and when the number of pixels with the gray scale value smaller than a first preset threshold value in the first detected image is larger than a preset number value, taking the frame of image as a first frame of image of the image to be processed;
when the number of the pixels with the gray values smaller than the first preset threshold value in the last detected image is smaller than a preset number value, taking the frame image as the last frame image of the image to be processed;
and taking the first frame image to the last frame image as the image to be processed.
Preferably, the image processing device further comprises a preprocessing module, configured to preprocess the image to be processed to obtain a preprocessed image; the preprocessing comprises image normalization processing, and the image normalization processing comprises coordinate centralization, x-sharpening normalization, scaling normalization or rotation normalization.
Preferably, the device further comprises a neural network training module, which is used for marking preset features of a target region in the two-dimensional DSA image sequence; inputting a two-dimensional DSA image sequence with a preset feature marker of a target region into a neural network model, and training the neural network model according to the two-dimensional DSA image sequence with the preset feature marker of the target region, wherein the neural network model comprises a convolutional neural network model; and identifying potential target areas in the preprocessed image according to preset characteristics of the target areas through the convolutional neural network model.
Preferably, the identification module is specifically configured to identify a potential target area in the preprocessed image according to preset features of the target area through the convolutional neural network model, so as to obtain a potential target area image.
Compared with the prior art, the target region identification method and the system thereof based on the two-dimensional DSA image have the advantages or beneficial effects that:
the invention can more accurately analyze the images by preprocessing, identifying and reprocessing a plurality of frames of images selected from the two-dimensional DSA image sequence, and finally determine the position of a connected domain, namely a target region in the images. The two-dimensional DSA image sequence processed by the method can process the image to be processed according to the trained convolutional neural network, and directly obtain a processing result, namely the position of a connected domain in each frame of image in the two-dimensional DSA image sequence. So that the worker can more easily and directly observe the image. Meanwhile, the method also solves the technical problems of high subjectivity and high error when the medical image is directly analyzed manually.
Drawings
In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort.
Fig. 1 is a schematic flowchart of an image recognition method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an image recognition system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be described in detail and completely with reference to the following specific embodiments of the present application and the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any inventive step based on the embodiments of the present disclosure, shall fall within the scope of protection of the present application.
The invention is further explained with reference to the drawings.
As a first embodiment, as shown in fig. 1, a schematic flow chart of the image recognition method of the present invention is shown, and the method mainly includes the following steps:
step S100: determining an image to be processed in a two-dimensional DSA image sequence, wherein the two-dimensional DSA image sequence comprises a plurality of frames of images, and selecting the image to be processed from the images. The image to be processed is an image in which a blood vessel appears in a two-dimensional DSA image sequence.
Step S200: and preprocessing the image to be processed to obtain a preprocessed image. The step is to preprocess the selected to-be-processed image with the blood vessel to obtain a preprocessed image.
Step S300: and identifying a potential target area in the preprocessed image according to preset characteristics of the target area to obtain a potential target area image.
Step S400: and performing connected domain analysis on the potential target region image, and taking a region corresponding to the connected domain as a final target region, thereby completing the identification of the target region in the to-be-processed image selected from the two-dimensional DSA image sequence.
The steps in the first embodiment realize the identification of the target region contained in the image to be processed in the two-dimensional DSA image sequence. The invention can more accurately analyze the images by preprocessing, identifying and reprocessing a plurality of frames of images determined in the two-dimensional DSA image sequence, and finally determine the position of a connected domain, namely a target region in the images. The two-dimensional DSA image sequence processed by the method can process the image to be processed according to the trained convolutional neural network, and directly obtain a processing result, namely the position of a connected domain in each frame of image in the two-dimensional DSA image sequence. The position of the connected domain is the position of the target region, so that workers can observe the image more easily and directly, and the region with the same characteristics as the preset target region can be directly obtained from the image. Meanwhile, the method also solves the technical problems of high subjectivity and high error when the medical image is directly analyzed manually.
The present invention further provides a second embodiment, which is a preferred embodiment, and further optimizes the method in the first embodiment, wherein:
step S100: there are many methods for determining the to-be-processed image in the two-dimensional DSA image sequence, and the following are two exemplary implementation methods of the present invention, but the step may also be other methods that can implement the technical solution of the present invention. Two different methods are described in detail below.
The method comprises the following steps:
and selecting a half image sequence positioned in the middle of the two-dimensional DSA image sequence as an image to be processed. The whole two-dimensional DSA image sequence comprises a plurality of frames of images, the first frames of images of the two-dimensional DSA image sequence are images without blood vessels, when the blood vessels generally appear at one quarter of the two-dimensional DSA image sequence, the blood vessels disappear at three quarters of the two-dimensional DSA image sequence, and therefore the frames from one quarter of the two-dimensional DSA image sequence to three quarters of the two-dimensional DSA image sequence are used as images to be processed. For example, a total of 24 images of the two-dimensional DSA image sequence may be selected from the seventh image to the eighteenth image as the image to be processed.
The total image frame number of the two-dimensional DSA image sequence is determined according to actual conditions, and the total image frame number of the two-dimensional DSA image sequence at different positions is different. When the total frame number of the images in the two-dimensional DSA image sequence is large and only a few frames of images appear in the two-dimensional DSA image sequence, the images of the frames with the blood vessels can be directly selected as the images to be processed. For example, a total of 32 frames of images in the two-dimensional DSA image sequence, and only the images from the 28 th frame to the 30 th frame have blood vessels, the images from the 28 th frame to the 30 th frame can be directly selected as the images to be processed.
The method can quickly and directly determine which images in the two-dimensional DSA image sequence can be used as the images to be processed without carrying out other image processing on the images in the two-dimensional DSA image sequence.
The second method comprises the following steps:
and sequentially detecting each frame of image in the two-dimensional DSA image sequence, and when the number of pixels of which the gray value is smaller than a first preset threshold value in the image is larger than a preset number value, taking the frame of image as a first frame of image of the image to be processed. For example, the gray scale level of each frame of image in the DSA image sequence is generally between 0-4095, the image size is 512 × 512, the first preset threshold value is set to 1200, and the preset number value is set to 1000. And starting detection from the first frame image of the DSA image sequence, and when the number of pixels with the gray value smaller than 1200 in a certain frame image is detected to be larger than 1000, determining that blood vessels appear in the frame image, and taking the frame image as the first frame image of the image to be processed.
And when the number of the pixels in the image smaller than the preset threshold value is smaller than the preset number value, taking the frame image as the last frame image of the image to be processed. Because the terminal blood vessel of the brain is close to the skull, the detection can be carried out from the last frame of image in the two-dimensional DSA image sequence, and the image with the blood vessel detected in the last frame is taken as the last frame of image of the image to be processed. Each frame of image after the first frame of image is detected in sequence from the first frame of image, and when the number of the detected pixels in the image smaller than the preset threshold value is smaller than the preset number value, the frame of image is taken as the last frame of image to be processed.
For example, the method for detecting whether there is a blood vessel in an image is also to regard that there is a blood vessel in a frame of image when it is detected that the number of pixels with gray scale values less than 1200 in the frame of image is less than 1000, and take the frame of image as the last frame of image of the image to be processed.
And taking the first frame image to the last frame image as the images to be processed, and taking the images to be processed regardless of whether the number of pixels in an intermediate frame image between the first frame image and the last frame image, which are smaller than a preset threshold value, is larger than or smaller than a preset number value.
The method comprises the step of comparing the number of gray values of pixels corresponding to a preset threshold value in each frame of image of a two-dimensional DSA image sequence with a preset number value to determine whether each frame of image belongs to an image to be processed. Because the similarity of the images of the adjacent frames in the two-dimensional DSA image sequence is higher, each frame of image needs to be detected, and the method can more accurately determine whether the images in the two-dimensional DSA image sequence belong to the images to be processed, so that the accuracy of determining whether the images in the two-dimensional DSA image sequence belong to the images to be processed is improved.
In addition, the method one and the method two can be combined to determine an image to be processed in the two-dimensional DSA image sequence. For example, an image with a blood vessel appearing in a two-dimensional DSA image sequence may be selected first by the first method, in order to detect whether a blood vessel appears in the image selected by the first method or further select an image showing a clearer blood vessel to determine a target region in the image, and then the image selected by the first method may be detected by the second method. The method I is combined with the method II to determine the images to be processed in the two-dimensional DSA image sequence, and the accuracy of determining the images to be processed is improved on the basis of quickly selecting the images to be processed.
The images to be processed in the two-dimensional DSA image sequence can also be determined by methods not exemplified in the present specification, which also fall within the scope of the present invention.
Step S200: and a step of preprocessing each frame of image in the images to be processed determined in the step S100.
Specifically, the image normalization processing may be one or more of coordinate centering, x-sharpening normalization, scaling normalization, and rotation normalization, and the image to be processed is processed into an image that meets the requirements of the preprocessing step. For example, the preprocessing step specifies an image size of 200mm with a pixel pitch of 1mm, and if the image to be processed is an image size of 512mm with a pixel pitch of 0.5 mm. By performing normalization processing on the image to be processed in the step, scaling normalization processing can be performed on the image to be processed first, and the image to be processed is reduced to an image with the size of 256mm × 256mm and the pixel pitch of 1 mm.
Then, the image is cut, and 28 pixels are cut respectively from the upper, lower, left and right sides of the image with the size of 256mm × 256mm and the pixel pitch of 1mm, so that the obtained image is the image with the size of 200mm × 200mm and the pixel pitch of 1 mm. The image obtained in the step is a preprocessed image, and the preprocessed image also comprises images with the same number of frames as the images to be processed. One or more of coordinate centering, x-sharpening normalization, rotation normalization and the like can be performed on the image to be processed, which is not illustrated here, and the purpose is to perform normalization processing on the image, so that the target area can be better determined.
The image normalization process transforms the original image to be processed into a corresponding unique standard form (which has invariant properties to affine transformations such as translation, rotation, scaling, etc.) through a series of transformations (i.e., finding a set of parameters using the invariant moment of the image so that it can eliminate the effect of other transformation functions on the image transformation).
Step S300: and identifying a potential target area in the preprocessed image obtained in the step S200 according to preset features of the target area to obtain a potential target area image, wherein the potential target area image includes probability values of the potential target area belonging to the target area. The step may specifically be:
the method comprises the steps of firstly marking a preset feature of a target region of a two-dimensional DSA image sequence, inputting the two-dimensional DSA image sequence with the preset feature mark of the target region into a neural network model, training the neural network model according to the two-dimensional DSA image sequence with the preset feature mark of the target region, wherein the target region is a region which is manually judged by a worker according to the two-dimensional DSA image sequence. In the technical scheme, the two-dimensional DSA image sequence also needs to process images with the same characteristics, and identify a region with the same characteristics as preset characteristics of a target region. The neural network model includes a convolutional neural network model, and may be other neural network models capable of realizing the same function as the neural network in the present technical solution. And taking the preprocessed image as the input of a neural network model, obtaining a potential target area image through the neural network model, wherein the characteristics of the potential target area image obtained through the neural network model are the same as the preset characteristics of a target area in an image used for training the neural network. And after the trained neural network inputs information, outputting the same information of the specific region or the characteristic according to the trained rule. The output result of the preprocessing image processing, namely the potential target area image, can be directly obtained through the neural network model, and the potential target area image comprises the probability value of the potential target area belonging to the target area, and the range of the probability value is 0-1. This step is to process each frame image included in the preprocessed image separately.
Step S400: performing binarization processing on the potential target region image according to the probability value in step S300 to generate a binarized image, performing connected domain analysis on the binarized image, and taking a region corresponding to the connected domain as a final target region, thereby completing the step of identifying a target region in the to-be-processed image selected from the two-dimensional DSA image sequence, which may specifically be:
and carrying out binarization processing on the potential target area image, presetting a second preset threshold value, comparing the probability value with the second preset threshold value, and generating a binarization image according to the comparison result. And the comparison result comprises that the probability value is greater than the second preset threshold value and the probability value is less than the second preset threshold value, and the probability value with the probability value greater than the second preset threshold value is set as 1 in the binary image, and corresponds to white or black in the binary image. And setting the probability value with the probability value smaller than the second preset threshold value as 0 in the binary image, wherein the probability value corresponds to black or white in the binary image, so that the binary image is generated.
For example, if the second preset threshold is 0.8, the probability value in the potential target region image is greater than 0.8, and is set to 1 in the binarized image, otherwise, the probability value is 0.
And analyzing the connected domain of the obtained binary image, marking all the connected domains, analyzing the connected domain of the probability value corresponding to 1 in the binary image, and taking the region corresponding to the connected domain generated by the probability value corresponding to 1 as a final target region. This step is also to reprocess each frame of the potential target area image separately.
The method provided by the invention realizes the identification of the target area, and can be completed through a corresponding system or device.
As shown in fig. 2, the present invention further provides a target region identification system based on two-dimensional DSA images, which mainly comprises:
and the image selection module 1 is used for selecting an image to be processed in the two-dimensional DSA image sequence. In particular, the module may be configured to select a half of the image sequence located in the middle of the two-dimensional DSA image sequence as the image to be processed.
The module can also be used for sequentially detecting each frame of image in the two-dimensional DSA image sequence, and when the number of pixels in the image smaller than the first preset threshold value is larger than a preset number value, the frame of image is used as a first frame of image of the image to be processed. And when the number of the pixels in the image smaller than the preset threshold value is smaller than the preset number value, taking the frame image as the last frame image of the image to be processed. And taking the first frame image to the last frame image as the images to be processed.
And the preprocessing module 2 is used for preprocessing the image to obtain a preprocessed image. In particular, the module may be configured to perform an image normalization process on the image.
And the neural network training module 3 is used for inputting the two-dimensional DSA image sequence with the preset target region characteristic mark into a neural network model, and training the neural network model according to the two-dimensional DSA image sequence with the preset target region characteristic mark, wherein the neural network model comprises a convolutional neural network model.
The identification module 4 is configured to identify a potential target area in the preprocessed image according to preset features of the target area, so as to obtain a potential target area image; the potential target region image includes probability values of potential target regions belonging to the target region.
And the reprocessing module 5 is configured to perform binarization processing on the potential target region map according to the probability value to generate a binarized image, perform connected domain analysis on the binarized image, and use a region corresponding to the connected domain as a final target region, thereby completing identification of a target region in the to-be-processed image selected from the two-dimensional DSA image sequence.
The system can realize the steps in the method and achieve the same technical effect.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the embodiments of the apparatus, the electronic device, and the nonvolatile computer storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and the relevant points can be referred to the partial description of the embodiments of the method.
The apparatus, the electronic device, the nonvolatile computer storage medium and the method provided in the embodiments of the present description correspond to each other, and therefore, the apparatus, the electronic device, and the nonvolatile computer storage medium also have similar advantageous technical effects to the corresponding method.
In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing an Integrated Circuit chip, such programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as ABEL (Advanced Boolean Expression Language), AHDL (alternate Hardware Description Language), traffic, CUPL (computer unified programming Language), HDCal, jhddl (Java Hardware Description Language), lava, lola, HDL, PALASM, rhyd (Hardware Description Language), and vhigh-Language (Hardware Description Language), which is currently used in most popular applications. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.
The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: the ARC 625D, atmelAT SAM, microchip PIC18F26K20, and silicon Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in purely computer readable program code means, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be regarded as a hardware component and the means for performing the various functions included therein may also be regarded as structures within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the various elements may be implemented in the same one or more software and/or hardware implementations in implementing one or more embodiments of the present description.
As will be appreciated by one skilled in the art, the present specification embodiments may be provided as a method, system, or computer program product. Accordingly, embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are also included in the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.
Claims (6)
1. A target region identification method based on two-dimensional DSA images is characterized by comprising the following steps:
determining an image to be processed in a two-dimensional DSA image sequence;
firstly, a frame in one quarter to three quarters of the two-dimensional DSA image sequence is selected,
then, sequentially detecting each frame of image in the selected two-dimensional DSA image sequence, and when the number of pixels with the gray scale value smaller than a first preset threshold value in the first detected image is larger than a preset number value, taking the frame of image as a first frame of image of the image to be processed;
when the number of the pixels with the gray values smaller than the first preset threshold value in the last detected image is smaller than a preset number value, taking the frame image as the last frame image of the image to be processed;
taking the first frame image to the last frame image as the image to be processed;
the first preset threshold is 1200, and the preset quantity value is 1000;
identifying a potential target area in the image to be processed according to preset characteristics of the target area to obtain a potential target area image; wherein the potential target area image comprises probability values of potential target areas belonging to the target areas; the image to be processed is a preprocessed image, the preprocessing is to carry out image normalization processing on the image to be processed, and the image normalization processing comprises coordinate centering, x-sharpening normalization, scaling normalization or rotation normalization;
performing connected domain analysis on the potential target area image, and taking an area corresponding to the connected domain as a final target area;
wherein performing connected component analysis on the potential target region image comprises: and comparing the probability value with a preset second preset threshold value, generating a binary image according to the comparison result, analyzing the connected domain of the probability value larger than the second preset threshold value, and taking the region corresponding to the connected domain generated by the probability value larger than the second preset threshold value as a target region.
2. The method for identifying the target area according to claim 1, before identifying the potential target area in the image to be processed according to the preset features of the target area, further comprising:
marking preset characteristics of a target region in a two-dimensional DSA image sequence;
inputting a two-dimensional DSA image sequence with a preset feature marker of a target region into a neural network model, and training the neural network model according to the two-dimensional DSA image sequence with the preset feature marker of the target region, wherein the neural network model comprises a convolutional neural network model;
and identifying potential target areas in the image to be processed according to preset characteristics of the target areas through the convolutional neural network model.
3. The identification method according to claim 2, wherein the performing connected component analysis on the potential target area image specifically comprises:
carrying out binarization processing on the potential target area image according to the probability value to generate a binarized image;
and carrying out connected domain analysis on the binary image.
4. The target region identification method according to claim 3, wherein the step of performing binarization processing on the potential target region map according to the probability value to generate a binarized image specifically comprises:
setting a second preset threshold, comparing the probability value with the second preset threshold, and generating a binary image according to a comparison result; the comparison result comprises that the probability value is greater than the second preset threshold value and the probability value is smaller than the second preset threshold value.
5. A target region identification system based on two-dimensional DSA images, comprising:
the image selection module is used for determining an image to be processed in a two-dimensional DSA image sequence;
firstly, a frame at one quarter to a frame at three quarters of the two-dimensional DSA image sequence are selected,
then, sequentially detecting each frame of image in the selected two-dimensional DSA image sequence, and when the number of pixels with the gray scale value smaller than a first preset threshold value in the first detected image is larger than a preset number value, taking the frame of image as a first frame of image of the image to be processed;
when the number of the pixels with the gray values smaller than the first preset threshold value in the last detected image is smaller than a preset number value, taking the frame image as the last frame image of the image to be processed;
taking the first frame image to the last frame image as the image to be processed;
the first preset threshold value is 1200, and the preset quantity value is 1000;
the identification module is used for identifying a potential target area in the image to be processed according to preset characteristics of the target area to obtain a potential target area image; wherein the potential target area image comprises probability values of potential target areas belonging to the target areas;
the preprocessing module is used for preprocessing the image to be processed to obtain a preprocessed image; the preprocessing comprises image normalization processing, and the image normalization processing comprises coordinate centralization, x-sharpening normalization, scaling normalization or rotation normalization;
the reprocessing module is used for analyzing the potential target area image connected domain and taking the area corresponding to the connected domain as a final target area;
wherein performing connected component analysis on the potential target region image comprises: and comparing the probability value with a preset second preset threshold value, generating a binary image according to the comparison result, analyzing the connected domain of the probability value larger than the second preset threshold value, and taking the region corresponding to the connected domain generated by the probability value larger than the second preset threshold value as a target region.
6. The target region identification system of claim 5, further comprising a neural network training module for labeling pre-set features of the target region in a two-dimensional DSA image sequence; inputting a two-dimensional DSA image sequence with a preset feature marker of a target region into a neural network model, and training the neural network model according to the two-dimensional DSA image sequence with the preset feature marker of the target region, wherein the neural network model comprises a convolutional neural network model; and identifying potential target areas in the preprocessed image according to preset characteristics of the target areas through the convolutional neural network model.
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 CN109816650A (en) | 2019-05-28 |
| CN109816650B true 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) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517244B (en) * | 2019-08-23 | 2023-04-28 | 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 |
| CN110634129B (en) * | 2019-08-23 | 2022-08-23 | 首都医科大学宣武医院 | Positioning method and system based on DSA image |
| 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 |
Citations (10)
| 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 |
| 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 |
| 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 |
| 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 |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN108052957B (en) * | 2017-11-07 | 2021-09-14 | 聊城大学 | Spacecraft target rapid identification method |
| 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 |
| 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 |
| CN109255784B (en) * | 2018-09-13 | 2021-06-25 | 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 (10)
| 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 |
| 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 |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109816650A (en) | 2019-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109816650B (en) | Target area identification method and system based on two-dimensional DSA image | |
| Ren et al. | Tenet: Triple excitation network for video salient object detection | |
| CN107274442B (en) | Image identification method and device | |
| CN109034183B (en) | Target detection method, device and equipment | |
| CN112016475B (en) | Human body detection and identification method and device | |
| CN111127428A (en) | Method and system for extracting target region based on brain image data | |
| CN112347512A (en) | Image processing method, device, equipment and storage medium | |
| CN112911281A (en) | Video quality evaluation method and device | |
| CN112784857A (en) | Model training and image processing method and device | |
| CN116342888B (en) | Method and device for training segmentation model based on sparse labeling | |
| CN111368902A (en) | Data labeling method and device | |
| CN109472803B (en) | Intracranial artery blood vessel segmentation method and system | |
| CN113205508B (en) | Segmentation method, device and equipment based on image data | |
| CN113130078B (en) | Method, device and equipment for predicting intracranial aneurysm occlusion | |
| CN114926437A (en) | Image quality evaluation method and device | |
| CN111368813B (en) | Motion blurred image identification method and device, electronic equipment and payment equipment | |
| CN113988162A (en) | Model training and image recognition method and device, storage medium and electronic equipment | |
| CN110517243B (en) | Positioning method and system based on DSA image | |
| CN110517244B (en) | Positioning method and system based on DSA image | |
| CN108388859B (en) | Object detection method, network training method, device and computer storage medium | |
| CN110599527A (en) | MRA image data registration method and device | |
| CN116363390B (en) | Infrared dim target detection method and device, storage medium and electronic equipment | |
| CN110634129B (en) | Positioning method and system based on DSA image | |
| US20230298326A1 (en) | Image augmentation method, electronic device and readable storage medium | |
| US20210295014A1 (en) | Collation device, collation method, and computer program product |
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 |