CN109801277B - Image processing method and device and storage medium - Google Patents

Abstract

The embodiment of the present invention discloses an image processing method, a device, and a storage medium. The image processing method includes: when an initial three-dimensional image of the coronary artery is obtained, based on the shape of the aorta, performing aortic identification in the initial three-dimensional image of the coronary artery, and obtaining Aortic position information; based on the aortic position information, determine the three-dimensional image of the aortic connected area from the initial coronary three-dimensional image; Obtain the boundary position information of the aorta and the left ventricle; perform coronary regrowth based on the boundary position information of the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery.

Description

Image processing method and device, and storage medium

technical field

The present invention relates to image processing technology in the computer field, and in particular, to an image processing method and device, and a storage medium.

Background technique

Coronary artery is a branch of blood vessels supplying blood to the heart. Coronary artery lesions will have a functional impact on the corresponding heart area, which is the cause of coronary heart disease, angina pectoris, myocardial infarction and other heart diseases. Therefore, by performing CTA ( Computed Tomography angiography, computer tomography angiography), allows doctors to know exactly the location and degree of stenosis and occlusion of the patient's heart vessels, which can provide great help in the diagnosis of heart disease, such as MSCT (Mufti-slice Spiral Computed Tomography, Multislice Spiral Computed Tomography). However, because the biomedical image itself has many unavoidable defects, in order to improve the readability of the biomedical image, it is necessary to perform computer processing on the biomedical image to obtain the processed biomedical image. diagnosis of disease. Therefore, CTA images of coronary arteries should also be processed by computer before they can become the basis for diagnosing cardiac diseases.

At present, the computer processing of CTA images of coronary arteries mainly extracts coronary arteries from CTA images of coronary arteries. In the prior art, the extraction of coronary arteries is usually achieved by clustering CTA images of coronary arteries or manually selecting seed points of coronary artery growth for regional growth. However, human intervention is required in the process of clustering and regional growth. , the intelligence of coronary artery extraction is low; in addition, the Frangi algorithm can also be used to extract coronary arteries for CTA images of coronary arteries.

SUMMARY OF THE INVENTION

In order to solve the above technical problems, the embodiments of the present invention are expected to provide an image processing method and device, and a storage medium, which can improve the intelligence of coronary artery acquisition and the effect of coronary artery acquisition.

The technical scheme of the present invention is realized as follows:

In a first aspect, an embodiment of the present invention provides an image processing method, the method comprising:

When the initial three-dimensional image of the coronary artery is obtained, based on the shape of the aorta, the aorta is identified in the initial three-dimensional image of the coronary artery to obtain the location information of the aorta;

based on the aortic position information, determining a three-dimensional image of aortic connectivity from the initial three-dimensional image of the coronary artery;

Based on the location information of the aorta and the physiological structure of the aorta, performing cardiac sinus identification in the three-dimensional image of the aortic connectivity domain to obtain the location information of the boundary between the aorta and the left ventricle;

Coronary artery regrowth is performed based on the position information of the boundary between the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery.

In the above solution, when the initial three-dimensional image of the coronary artery is acquired, the method further includes:

When at least two two-dimensional coronary angiography images are acquired, respectively acquire corresponding image grayscale correction information from the at least two coronary angiography two-dimensional images, where the image grayscale correction information includes the at least two The initial grayscale value and format information corresponding to the two-dimensional coronary angiography image;

obtaining, according to the initial grayscale value and the format information, density values corresponding to the at least two two-dimensional coronary angiography images;

The initial three-dimensional image of the coronary artery is obtained by superimposing the at least two two-dimensional coronary angiography images according to the density value and the preset image hierarchy sequence.

In the above solution, based on the shape of the aorta, the aorta is identified in the initial three-dimensional image of the coronary artery to obtain the position information of the aorta, including:

determining a radius threshold according to the shape of the aorta;

Based on the preset image hierarchy order, binarize the hierarchical coronary 3D image in the initial coronary 3D image according to the preset gray value, and according to the preset positioning algorithm and the radius threshold, in the The aorta is identified in the result image corresponding to the binarization process on the three-dimensional coronary artery image of the hierarchy until the position information of the aorta is obtained.

In the above solution, the determining of the three-dimensional image of the aortic connected domain from the initial three-dimensional coronary artery image based on the aortic position information includes:

obtaining an initial aortic gray value according to the aortic position information;

determining the aortic gray value according to the preset gray value and the initial aortic gray value;

performing binarization processing on the initial three-dimensional coronary artery image according to the gray value of the aorta to obtain a binarized initial three-dimensional coronary artery image;

According to the aortic position information, the aortic connected area is determined from the binarized initial three-dimensional image of the coronary artery, and the three-dimensional image of the aortic connected area is obtained.

In the above solution, based on the location information of the aorta and the physiological structure of the aorta, the identification of the cardiac sinus is performed in the three-dimensional image of the aortic connected domain to obtain the location information of the boundary between the aorta and the left ventricle, including:

According to the physiological structure of the aorta, determine a location determination condition, where the location determination condition is used to identify the cardiac sinus;

In the three-dimensional image of the connected domain of the aorta, a search plane is determined from the position information of the aorta according to a preset search step, until the cross-sectional area of the aorta corresponding to the search plane satisfies the position determination condition, the corresponding The search plane is used as the boundary position information of the aorta and the left ventricle.

In the above solution, performing coronary regrowth based on the position information of the boundary between the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery, including:

Regrowing the coronary artery based on the position information of the boundary between the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery growth;

using a preset algorithm to identify the aortic region on the three-dimensional image of the coronary artery growth;

The aortic region is deleted from the three-dimensional image of the coronary artery growth to obtain the three-dimensional image of the coronary artery.

In the above solution, performing coronary artery regrowth based on the position information of the boundary between the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery growth, including:

According to the position information of the boundary between the aorta and the left ventricle, and the position information of the aorta, determine the gray value of the coronary artery growth;

According to the position information of the boundary between the aorta and the left ventricle, delete the three-dimensional image of the connected area of the aorta from the initial three-dimensional image of the coronary artery to obtain the deleted three-dimensional image of the initial coronary artery;

Coronary artery growth is performed based on the gray value of the coronary artery growth and the deleted initial three-dimensional image of the coronary artery to obtain the three-dimensional image of the coronary artery growth.

In the above solution, determining the gray value of coronary artery growth according to the position information of the boundary between the aorta and the left ventricle and the position information of the aorta includes:

According to the boundary position information of the aorta and the left ventricle, obtain the coronary arteries connected domain;

When the volume corresponding to the coronary connected region is greater than the preset volume, on the basis of the aortic gray value corresponding to the aortic position information, the preset de-adhesion step is cyclically increased until the coronary connected region is reached The corresponding volume is not greater than the preset volume, and the cyclically increased grayscale value is used as the coronary artery growth grayscale value.

In the above solution, performing coronary artery growth based on the grayscale value of the coronary artery growth and the deleted initial three-dimensional image of the coronary artery to obtain the three-dimensional image of the coronary artery growth, including:

According to the coronary artery growth gray value and the preset growth step size, perform coronary artery generation on the deleted initial coronary three-dimensional image until the coronary artery growth gray value is reduced to a preset minimum growth gray value , complete the coronary artery growth, and obtain the three-dimensional image of the coronary artery growth.

In a second aspect, an embodiment of the present invention provides an image processing apparatus, the apparatus includes: a processor, a memory, and a communication bus, the memory communicates with the processor through the communication bus, and the memory stores data A program executable by the processor, when the program is executed, the image processing method as described above is executed by the processor.

In a third aspect, an embodiment of the present invention provides a computer-readable storage medium on which a program is stored, and when the program is executed by a processor, the image processing method as described above is implemented.

Embodiments of the present invention provide an image processing method and device, and a storage medium. First, when the initial coronary 3D image is obtained, based on the shape of the aorta, the aorta is identified in the initial coronary 3D image to obtain aortic position information; based on the aortic position information, the aorta is determined from the initial coronary 3D image. Connected domain 3D image; based on aortic position information and aortic physiological structure, identify the cardiac sinus in the aortic connected domain 3D image to obtain the boundary position information of the aorta and left ventricle; based on the aorta and left ventricular boundary position information Arteries re-grow, and three-dimensional images of coronary arteries are obtained. With the above technical implementation scheme, the image processing device accurately identifies the aorta and coronary arteries in the initial three-dimensional image of the coronary arteries according to the morphological structures of the aorta and the physiological structure of the aorta, and based on the identified aorta and the The left ventricular boundary position information is used to automatically grow the coronary artery, thereby obtaining a three-dimensional image of the coronary artery, realizing the automatic and accurate extraction of the coronary artery, improving the intelligence of coronary artery acquisition, and improving the effect of coronary artery acquisition.

Description of drawings

FIG. 1 is a flowchart for realizing an image processing method according to an embodiment of the present invention;

FIG. 2 is an exemplary three-dimensional image of aortic connected domain provided by an embodiment of the present invention;

3 is a schematic diagram of an exemplary search for position information of the boundary between the aorta and the left ventricle according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of an exemplary determination of a grayscale value of coronary artery growth provided by an embodiment of the present invention;

FIG. 5 is an exemplary aorta and coronary arteries connected domain image provided by an embodiment of the present invention;

6 is a plan view of an exemplary deleted initial three-dimensional coronary artery image provided by an embodiment of the present invention;

FIG. 7 is an exemplary three-dimensional image of coronary artery growth provided by an embodiment of the present invention;

FIG. 8 is an exemplary three-dimensional image of a coronary artery according to an embodiment of the present invention;

FIG. 9 is a schematic structural diagram 1 of a voice signal processing apparatus according to an embodiment of the present invention;

FIG. 10 is a second schematic structural diagram of a voice signal processing apparatus according to an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

Example 1

An embodiment of the present invention provides an image processing method. FIG. 1 is an implementation flowchart of an image processing method provided by an embodiment of the present invention. As shown in FIG. 1 , the image processing method includes:

S101. When the initial three-dimensional image of the coronary artery is obtained, based on the shape of the aorta, perform aortic identification in the initial three-dimensional image of the coronary artery to obtain aortic position information.

In the embodiment of the present invention, the angiography apparatus performs angiography on the coronary artery after injection of the contrast agent, and inputs at least two coronary angiography images obtained by the angiography to the image processing device. At this time, the image processing device obtains at least two coronary angiography images. Two coronary angiography images; when the image processing device superimposes the obtained at least two coronary angiography images, an initial three-dimensional image of the coronary arteries is obtained. The shape of the aorta can be determined based on the prior knowledge: the cross section is circular, and the image processing device determines the position information of the aorta in the initial three-dimensional image of the coronary artery based on the shape of the cross section of the aorta being circular.

Here, the imaging may be CT imaging, X-ray imaging, or other imaging techniques, which are not specifically limited in this embodiment of the present invention. Specifically, when the angiography is CT angiography, the coronary angiography image corresponds to a coronary CT angiography image.

It should be noted that the initial three-dimensional image of the coronary artery includes images corresponding to the region where the coronary arteries are located, such as a stereoscopic image of the human chest (the region including the heart); generally, the initial three-dimensional image of the coronary artery is a stereoscopic image, such as , a three-dimensional matrix can be used to represent the initial coronary three-dimensional image; the aortic position information represents the position information of the aorta in the initial coronary three-dimensional image.

S102 , based on the position information of the aorta, from the initial three-dimensional image of the coronary artery, determine a three-dimensional image of the aorta connected area.

In this embodiment of the present invention, after the image processing apparatus acquires the position information of the aorta, since the aorta is connected to the coronary arteries, after the image processing apparatus obtains the initial three-dimensional coronary artery image, the Based on the location information of the aorta, the connected region where the aorta is located is determined, thereby obtaining a three-dimensional image of the connected region of the aorta including the coronary arteries.

That is to say, the three-dimensional image of the connected area of the aorta represents the stereoscopic image formed by the connected area where the aorta is located.

Specifically, after obtaining the location information of the aorta, the image processing device searches for the connected area where the aorta is located according to the location information of the aorta, that is, obtains a three-dimensional image of the connected area of the aorta.

It can be understood that, since the aorta is connected with the coronary arteries, the coronary arteries are included in the three-dimensional image of the aorta connected region.

S103 , based on the aortic position information and the aortic physiological structure, perform cardiac sinus identification in the three-dimensional image of the aortic connected area to obtain the boundary position information of the aorta and the left ventricle.

In the embodiment of the present invention, after obtaining the aortic position information and the three-dimensional image of the aortic connected area, the image processing device performs the main path from the distal end to the proximal end along the aortic position information in the three-dimensional image of the aortic connected area. The search for the location information of the boundary between the artery and the left ventricle. In this process, the aorta has a physiological structure that is thick and then thin. When the location of the physiological structure is determined, the boundary location information of the aorta and the left ventricle is determined. .

It should be noted that the location represented by the first thick and then thin physiological structure is the cardiac sinus, and the cardiac sinus is the bulge position on the aorta that communicates with the coronary arteries.

It can be understood that, because the three-dimensional image of the aortic connected area represents the connected area where the aorta is located in the initial coronary image, and the aorta is usually connected to the coronary artery, the left ventricle, etc.; After the location information of the ventricular boundary is obtained, the connected domain where the left ventricle connected to the aorta is located can be removed according to the boundary location information of the aorta and the left ventricle, which further provides convenience for the extraction of the coronary arteries.

S104 , performing coronary regrowth based on the boundary position information of the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery.

In the embodiment of the present invention, because the three-dimensional image of the aortic connectivity domain includes not only the connection between the aorta and the coronary artery, but also the connection between the aorta and other parts (for example, the left ventricle), and the connection between the coronary artery and other parts (for example, the atrial appendage), therefore, after the image processing device obtains the boundary position information of the aorta and the left ventricle, it performs coronary regrowth processing on the three-dimensional image of the aorta connected area according to the boundary position information of the aorta and the left ventricle, The above-mentioned unnecessary connected areas (the connection between the aorta and other parts, and the connection between the coronary arteries and other parts) are removed, and the extraction of the coronary arteries is completed, and the three-dimensional images of the coronary arteries are obtained.

Here, the three-dimensional image of coronary artery is processed by computer technology and can be used as a basis for medical diagnosis of heart disease, for example, a three-dimensional mesh model of coronary artery.

It can be understood that the image processing device accurately locates the aorta and coronary arteries in the obtained initial three-dimensional images of coronary arteries through the morphological structure of the aorta, realizes efficient and complete segmentation and reconstruction processing, and completes the CT scan. (Computed Tomography) Automatic and accurate extraction of coronary arteries from two-dimensional images of coronary angiography.

Further, in the embodiment of the present invention, the image processing method further includes that when at least two two-dimensional coronary angiography images are acquired, the image processing apparatus superimposes at least two coronary angiography two-dimensional images based on a preset image hierarchy sequence, Obtain an initial three-dimensional image of the coronary arteries. Specifically include S105-S107, including:

S105. When at least two two-dimensional coronary angiography images are acquired, obtain corresponding image grayscale correction information from the at least two coronary angiography two-dimensional images, respectively, where the image grayscale correction information includes at least two coronary angiography two-dimensional images. The initial gray value and format information corresponding to the dimensional image.

In the embodiment of the present invention, when performing a tomography scan on a coronary artery, at least two two-dimensional coronary angiography images can be captured by injecting a contrast agent intravenously, and the at least two two-dimensional coronary angiography images can be sent to the image processing device During the introduction, the image processing device acquires at least two two-dimensional coronary angiography images. The at least two two-dimensional coronary angiography images obtained by the image processing device include image grayscale correction information, and the image grayscale correction information is used to perform image conversion on the coronary angiography two-dimensional images: converting the coronary angiography two-dimensional images from The first space is converted into the second space, so that the image processing device performs extraction processing of coronary arteries in the second space; here, the first space represents at least two two-dimensional coronary angiography images obtained by the angiography device during coronary angiography The image space corresponding to the image, and the second space represents the image space corresponding to the at least two two-dimensional coronary angiography images when the image processing apparatus performs coronary extraction on the at least two two-dimensional coronary angiography images.

Here, the image grayscale correction information includes initial grayscale values and format information corresponding to at least two two-dimensional coronary angiography images.

It should be noted that the at least two two-dimensional coronary angiography images obtained by the image processing device are images in the first space, and the image format corresponding to each coronary angiography two-dimensional image at this time is the first image format, such as , DICOM images, so that the format information corresponding to the at least two two-dimensional coronary angiography images represents information corresponding to the first image format, such as header file information of the DICOM images.

In addition, the initial grayscale values corresponding to the at least two two-dimensional coronary angiography images represent grayscale values corresponding to the at least two two-dimensional coronary angiography images in the first space.

That is to say, the image processing apparatus reads the information corresponding to the first image format and the corresponding grayscale values in the first space from the obtained at least two two-dimensional coronary angiography images, that is, the image grayscale correction information is obtained.

S106: Obtain density values corresponding to at least two two-dimensional coronary angiography images according to the initial grayscale value and format information.

In the embodiment of the present invention, after obtaining the image grayscale correction information, the image processing apparatus can perform conversion processing from the first space to the second space on at least two two-dimensional coronary angiography images, specifically: the image processing apparatus According to the initial gray value and format information, the density values corresponding to at least two two-dimensional coronary angiography images are obtained.

It should be noted that the density value represents the information in the second image format corresponding to the at least two two-dimensional coronary angiography images in the second space, such as CT value.

Exemplarily, the image processing device obtains a set of CTA images (at least two two-dimensional coronary angiography images) of the human chest including the heart region, and the set of CTA images are DICOM images; Obtain "Rescale Slope" and "Rescale Intercept" (format information) from the file information, and determine the gray value (initial gray value) of each image in at least two two-dimensional coronary angiography images; Formula (1) determines the CT value (density value) corresponding to at least two two-dimensional coronary angiography images in the second space, where formula (1) is:

HU=Rescale Slope×I+Rescale Intercept (1)

Among them, HU represents the CT value corresponding to each pixel in at least two two-dimensional coronary angiography images, and I represents the initial gray value.

S107 , superimposing at least two two-dimensional coronary angiography images according to the density value and the preset image hierarchy order to obtain an initial three-dimensional coronary image.

In the embodiment of the present invention, after the image processing device obtains the density values corresponding to at least two two-dimensional coronary angiography images, the CT value corresponding to each part in the region where the coronary arteries are located is determined, and the computed tomography scan is added. The device sequentially shoots the parts where the coronary arteries are located, so that there is a fixed image level between the at least two obtained two-dimensional coronary angiography images. Therefore, the image processing device, based on the density value and the preset image level sequence, At least two coronary angiography two-dimensional images are superimposed to obtain an initial coronary three-dimensional image for extracting coronary arteries.

Here, each pixel in the initial three-dimensional image of the coronary artery corresponds to its own CT value, and subsequent coronary artery extraction is performed based on the CT value.

It should be noted that the pixels representing the same part exist in at least one 2D coronary angiography image correspondingly, but in the second space, the corresponding density values of the pixels in the same part are the same.

In this embodiment of the present invention, when the initial three-dimensional coronary artery image is a stereoscopic image, for example, a three-dimensional matrix can be used to represent the initial three-dimensional coronary artery image, correspondingly, the two-dimensional coronary angiography image is a two-dimensional matrix, so that the image The processing device superimposes a plurality of two-dimensional coronary angiography images represented by a two-dimensional matrix into an initial coronary three-dimensional image in the form of a three-dimensional matrix according to a preset image hierarchy sequence.

Here, when the image processing device converts at least two coronary angiography images into initial three-dimensional images of coronary arteries according to the density value and the preset image hierarchy order, the image processing device may construct a three-dimensional image according to the density value and the preset image hierarchy order. model, and input at least two coronary angiography images into the three-dimensional image model, so as to obtain an initial three-dimensional coronary artery image; it is also possible that the image processing device analyzes the at least two coronary angiography images according to the density value and the preset image hierarchy order. The planar structures in the image are stretched and combined to obtain an initial three-dimensional image of the coronary artery; this is not specifically limited in the embodiment of the present invention.

Exemplarily, a CT device (angiography device) performs a tomographic scan on a patient's chest (region containing the heart) after injection of a contrast agent, and obtains a set of coronary CTA images (at least two coronary angiography two-dimensional images), wherein, Each picture in the group of coronary CTA images is a two-dimensional matrix, and the image processing device arranges the group of coronary CTA images in a preset image hierarchy order to form a three-dimensional matrix, that is, an initial three-dimensional coronary artery image.

It can be understood that the image processing apparatus obtains an initial three-dimensional image of coronary arteries by processing and superimposing multiple two-dimensional coronary angiography images, which provides original information for subsequent coronary artery extraction.

Further, in the embodiment of the present invention, the image processing device in S101 performs aortic identification in the initial three-dimensional image of the coronary artery based on the shape of the aorta, and obtains the position information of the aorta, which specifically includes: the image processing device determines the shape of the aorta according to the shape of the aorta. Radius threshold; and based on the preset image hierarchy order, according to the preset gray value, the hierarchical coronary 2D image in the initial coronary 3D image is binarized, and according to the preset positioning algorithm and radius threshold, in the hierarchy coronary The aorta is identified in the resulting image corresponding to the binarization of the two-dimensional arterial image until the location information of the aorta is obtained.

It should be noted that in the initial three-dimensional image of the coronary artery, there are two structures with circular cross-sections, one of which is the aorta and the other is the inferior artery, and the radius of the inferior artery is significantly smaller than that of the aorta. , so that the image processing device determines the minimum threshold of the radius, ie, the radius threshold, according to the radius of the aorta, for example, 40 pixels; based on the radius threshold, the interference of the inferior artery is excluded.

Specifically, the image processing device performs the positioning of the aorta based on the preset image hierarchy sequence: starting from the first layer image, the first layer image is binarized by using the preset gray value, and the binarized image is obtained. and using the preset positioning algorithm for locating the circular structure and the set radius threshold to determine the cross-section of the aorta, that is, the aortic position information is obtained. If the location information of the aorta is not determined in the first layer image, the image processing device continues to locate the aorta in the second layer image, and the specific positioning process is the same as the positioning process of performing aortic location in the first layer image. Similarly, if the location information of the aorta has not been determined in the second image, the image processing device will continue to locate the aorta in the third layer image, and thus traverse the initial three-dimensional image of the coronary artery until the location of the aorta is determined; , the aortic position information represents the position information of the aorta on the hierarchical coronary two-dimensional image in the initial coronary three-dimensional image; for example, the aortic position information is formula (2):

position _Aorta = (x ₀ , y ₀ , i) (2)

Among them, position _Aorta represents the position information of the aorta, and (x ₀ , y ₀ , i) represents the position of the center of the aorta (x ₀ , y ₀ ) on the image of the i-th layer.

It should be noted that the preset gray value is a gray value preset by the image processing device, and the preset gray value is obtained based on a combination of a large number of theories and experiments, for example: 226. Preset localization algorithms are used to locate circular structures, for example, Hough transform: Hough transform can determine the possibility that a geometric shape in an image is a standard shape (eg: circle, ellipse, line). In addition, the first slice image, the second slice image, the third slice image .

Further, in the embodiment of the present invention, the image processing device in S102 determines the three-dimensional image of the aortic connected domain from the initial coronary three-dimensional image based on the aortic position information, which specifically includes: the image processing device obtains the initial three-dimensional image according to the aortic position information. The gray value of the aorta; and determining the gray value of the aorta according to the preset gray value and the gray value of the initial aorta; The valued initial coronary three-dimensional image; and according to the aortic position information, the aortic connected area is determined from the binarized initial coronary three-dimensional image, and the aortic connected area three-dimensional image is obtained.

In the embodiment of the present invention, after the image processing device obtains the aortic position information, since the aortic position information represents the position of the aorta in the two-dimensional image of the coronary artery at a certain level, according to the aortic position information, it is possible to obtain The initial grayscale value of the aorta corresponding to the two-dimensional image of the coronary artery at this level. That is to say, the initial aortic gray value represents the gray value of the aorta in a two-dimensional image of the coronary artery at a certain level. In addition, the image processing device determines a three-dimensional image of the aortic connected region from the initial three-dimensional image of the coronary artery according to the aortic position information and the initial aortic gray value. That is to say, after obtaining the aortic position information and the initial aortic gray value, the image processing device can find the location of the aorta at the aortic position information determined in the initial coronary three-dimensional image according to the initial aortic gray value. The connected domain of the aorta is obtained to obtain a three-dimensional image of the connected domain of the aorta.

Specifically, after obtaining the initial aortic gray value, the image processing device performs binarization processing on the initial three-dimensional coronary artery image by using the aortic gray value. Therefore, when determining the three-dimensional image of the aortic connected area, the aortic gray value needs to be obtained according to the initial aortic gray value, and the aortic gray value represents the gray value of the aorta in the initial coronal image. At this time, the image processing device The initial aortic gray value is updated by using the preset gray value, so as to obtain the updated gray value, which is the aortic gray value.

Exemplarily, the image processing device uses formula (3) to realize determining the aortic gray value according to the preset gray value and the initial aortic gray value, and formula (3) is:

Among them, gray _new represents the gray value of the aorta, gray _Aorta represents the initial gray value of the aorta, and gray _exp represents the preset gray value.

After the image processing device obtains the grayscale value of the aorta, it uses the grayscale value of the aorta to perform binarization processing on the initial coronary artery, and uses the known position information of the aorta on the obtained three-dimensional image of the initial coronary artery after the binarization. Find the connected domain of the aorta. For example, the image processing device uses the region growing method to traverse all the pixels with a value of 1 connected with the position information of the aorta on the initial three-dimensional image of the coronary artery after binarization, and mark these connected pixels with a value of 1. Thus, all marked pixels with connected value of 1 are determined as the connected domain where the aorta is located.

It should be noted that although the aortic position information is the position of the aorta on the hierarchical coronary 2D image, since the hierarchical coronary 2D image is the image information in the initial coronary 3D image, the aortic position information also corresponds to Location information of the aorta in the initial coronary 3D image.

FIG. 2 is an exemplary three-dimensional image of the aortic connected area provided by an embodiment of the present invention. As shown in FIG. 2 , the circular structure part is the aorta 10 . In addition, the three-dimensional image of the aortic connected area further includes Coronary artery, left ventricular cavity, left atrium cavity, atrial appendage, and pulmonary vein connected to left atrium, etc.; in addition, depending on the timing of CTA images, the 3D image of aortic connectivity may also include the right ventricular cavity, right atrium cavity and the pulmonary artery to which the right ventricle connects.

It can be understood that, by determining the connected area where the aorta is located in the initial three-dimensional image of the coronary artery, the three-dimensional image of the connected area of the aorta is obtained. Therefore, when the image processing apparatus extracts the coronary arteries from the three-dimensional image of the aortic connected area, the extraction range of the coronary arteries is reduced, and the practicability of the coronary arteries extraction is improved to a certain extent.

Further, in the embodiment of the present invention, the image processing device in S103 performs cardiac sinus identification in the three-dimensional image of the aortic connectivity domain based on the aortic position information and the aortic physiological structure, and obtains the boundary position information of the aorta and the left ventricle, specifically: The method includes: the image processing device determines the position determination condition according to the physiological structure of the aorta, and the position determination condition is used to identify the cardiac sinus; and in the three-dimensional image of the aortic connected domain, the search plane is determined according to the preset search step length from the position information of the aorta , until the cross-sectional area of the aorta corresponding to the search plane satisfies the position determination condition, the corresponding search plane is used as the boundary position information of the aorta and the left ventricle.

Here, the image processing device determines the search plane according to the preset search step size at the location of the aorta, which specifically includes: taking the location information of the aorta as the first location, and constructing a first search plane corresponding to the first location; Presetting a search step size to determine a second position; constructing a second search plane corresponding to the second position; determining a third position according to the first position and the second position; and constructing a third search plane corresponding to the third position. Finally, a cross-sectional image of the aorta is obtained according to the first search plane, the second search plane, and the third search plane. Here, the first search plane, the second search plane and the third search plane are the above search planes.

That is to say, first, the image processing device uses the aortic position information as the first position to construct a first search plane; when the aortic position information is shown in formula (2), the first search plane is shown in formula (4):

(yy ₀ )+(zi)=0 (4)

Among them, the three-dimensional image of the aortic connected area is a three-dimensional image. When represented by three-dimensional coordinates (x, y, z), z represents the direction from the distal end to the proximal end, and (x, y) represents the cross-section captured by the CT equipment. noodle.

In the first search plane, the image processing device removes an area that is not connected to the first position position _Aorta1 as: position _Aorta , and obtains an image IM _Aorta1 that only includes a cross section of the aorta.

Secondly, the image processing device determines the second position of the aorta, position _Aorta2 , according to the preset search step size step (for example, 10 pixels) as: position _Aorta +(0,0,step), so as to obtain the second search plane, as shown in formula (5) ) as shown:

(yy ₀ )+[z-(i+step)]=0 (5)

In the second search plane, the image processing device removes the area that is not connected to the second position position _Aorta2 , and obtains an image IM _Aorta2 that only contains the cross section of the aorta.

如式(6)所示：Then, the image processing device predicts and obtains a third position according to the first position and the second position, and specifically, determines the direction vector of the aorta according to the first position and the second position

As shown in formula (6):

Therefore, the image processing apparatus determines that the third position position _Aorta3 is

The third search plane is further obtained, as shown in formula (7):

为position_Aorta3对应的y值，

为position_Aorta3对应的z值。in,

is the y value corresponding to position _Aorta3 ,

is the z value corresponding to position _Aorta3 .

In the third search plane, the image processing device removes the region that is not connected to the third position position _Aorta3 , and obtains an image IM _Aorta3 that only contains the cross section of the aorta.

Finally, by repeating the above search process, a series of images {IM _Aorta1 , IM _Aorta2 , IM _{Aorta3 ,} . When the image processing device determines that the cross-sectional area of the series of images containing only the aortic cross-section first becomes larger and then becomes smaller, the search for the aortic cross-section is ended, and the position information corresponding to IM _Aortaj is used as the boundary between the aorta and the left ventricle. information.

In the embodiment of the present invention, the position determination condition for the image processing device to determine the position information of the boundary between the aorta and the left ventricle is: the area S (IM _Aortaj+1 ) of the most recent aortic cross-section is larger than the area of the previous aortic cross-section first. S(IM _Aortaj ) indicates that the search position reaches the upper part of the cardiac sinus; after that, the area S(IM _Aortak+1 ) of the latest aortic cross-section is smaller than the area S(IM _Aortak ) of the previous aortic cross-section, indicating that the search The position reaches the lower part of the sinus.

Preferably, the above position determination conditions are: S(IM _Aortaj+1 )>α×S(IM _Aortaj ), and S(IM _Aortak+1 )<β×S(IM _Aortak ), where α is greater than 1, for example, 1.01; β is less than 1, eg, 0.99. In this way, the influence of image noise is reduced, and the judgment accuracy is improved.

In addition, in the embodiment of the present invention, when the obtained aortic cross-section satisfies the above judgment conditions, a two-step search is performed, and the corresponding position at this time is used as the boundary position information of the aorta and the left ventricle.

FIG. 3 is a schematic diagram of an exemplary search for position information of the boundary between the aorta and the left ventricle according to an embodiment of the present invention. As shown in FIG. 3 , 1 to i represent the first search plane to the i th search plane, wherein the position corresponding to the i th search plane is the boundary position information of the aorta and the left ventricle.

It can be understood that since the aorta is in communication with the left ventricle, by determining the position information of the boundary between the aorta and the left ventricle, conditions are provided for removing the left ventricle from the aorta-connected image.

Further, in the embodiment of the present invention, the image processing apparatus in S104 performs coronary regrowth based on the position information of the aorta and left ventricle boundary to obtain a three-dimensional image of the coronary artery, which specifically includes S104a-S104c, wherein:

S104a, performing coronary artery regrowth based on the boundary position information of the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery growth.

In the embodiment of the present invention, after obtaining the boundary position information of the aorta and the left ventricle, the image processing apparatus performs coronary regrowth based on the boundary position information of the aorta and the left ventricle, and obtains a three-dimensional image of the coronary artery growth.

Here, the three-dimensional image representation of the coronary artery growth includes three-dimensional images of the aorta and coronary arteries.

Further, in the embodiment of the present invention, the image processing device in S104a performs coronary artery regrowth based on the position information of the aorta and left ventricle boundary to obtain a three-dimensional image of coronary artery growth, specifically including S104a1-S104a3, wherein:

S104a1: Determine the gray value of coronary artery growth according to the position information of the boundary between the aorta and the left ventricle and the position information of the aorta.

In the embodiment of the present invention, after determining the position information of the boundary between the aorta and the left ventricle, the image processing device divides the three-dimensional image of the connected area of the aorta into two parts: the connected area where the coronary arteries are located and the connected area where the ventricle is located; the image processing device determines In the connected domain where the coronary arteries are located, whether there is adhesion between the coronary arteries and the atrial appendages, when it is determined that the coronary arteries and the atrial appendages are adhesions, the gray value of the aorta corresponding to the location information of the aorta is increased according to the preset de-adhesion step, until the coronary arteries There was no adhesion between the artery and the atrial appendage, and the gray value corresponding to this time was taken as the gray value of the coronary artery growth.

In addition, when the image processing device determines that the coronary arteries are located in the connected domain, and the coronary arteries and the atrial appendages are not adhered, the aortic grayscale value is used as the coronary artery growth grayscale value.

Further, the image processing device determines the grayscale value of the coronary artery growth according to the position information of the aorta and the left ventricle, and the position information of the aorta, which specifically includes: the image processing device obtains the coronary artery according to the position information of the aorta and the left ventricle. connected domain; and when the volume corresponding to the coronary connected domain is greater than the preset volume, on the basis of the aortic gray value corresponding to the aortic position information, the preset de-adhesion step is cyclically increased until the corresponding volume of the coronary connected domain is The volume is not larger than the preset volume, and the gray value after cyclic increase is used as the gray value of coronary artery growth. Here, the coronary connected domain refers to the connected domain where the coronary arteries are located.

It should be noted that the aortic gray value corresponding to the aortic position information refers to the image processing device determining the aortic gray value according to the aortic position information, specifically, the image processing device determining the initial aortic gray value according to the aortic position information, And the aortic gray value is determined according to the initial aortic gray value and the preset gray value.

That is to say, once the coronary arteries and the atrial appendages are adhered, the volume corresponding to the connected region where the coronary arteries are located is greater than the preset volume. Therefore, the image processing device determines whether the coronary arteries and the atrial appendages are adhered according to the preset preset number of connected pixels. , for example, set the preset number of connected pixels to 1,300,000, then, when the number of pixels in the connected region of the aorta and coronary arteries is greater than 1,300,000, it is determined that the coronary arteries and atrial appendages are adhered; when the number of pixels in the connected region of the aorta and coronary artery is different When the number is greater than 1,300,000, it is determined that there is no adhesion between the coronary arteries and the atrial appendages.

At the same time, the image processing device continuously increases the preset de-adhesion step size on the basis of the gray value of the aorta, and judges whether there is still adhesion between the coronary artery and the atrial appendage on the increased gray value; On the basis of the second gray value, increase the preset de-adhesion step size, and judge whether the coronary artery and the atrial appendage still have adhesion on the gray value after the increase again; repeat this process until the coronary artery and the atrial appendage are no longer adhered, and this The corresponding increased gray value is the coronary artery growth gray value.

Exemplarily, FIG. 4 is a schematic diagram of an exemplary determination of the gray value of coronary artery growth provided by an embodiment of the present invention. As shown in FIG. 4 , the preset de-adhesion step size step _gray1 is 20. The upper left picture: the gray value of the aorta is 373.5, and the image processing device determines that the coronary artery and the atrial appendage are adhered at this time, so the gray value is increased by 20, and the upper right picture in Fig. 4 is obtained; in the upper right corner of Fig. 4 In the figure: the corresponding gray value is 393.5, and the image processing device determines that the coronary artery and the atrial appendage are adhered at this time, so the gray value is increased by 20, and the lower left image in Fig. 4 is obtained; the lower left image in Fig. 4 Middle: the corresponding gray value is 413.5, and the image processing device determines that the coronary artery and the atrial appendage are adhered at this time, so the gray value is increased by 20, and the lower right image in Fig. 4 is obtained; in the lower right image in Fig. 4 : The corresponding gray value is 433.5, and the image processing device determines that there is no adhesion between the coronary artery and the atrial appendage at this time, so that the gray value of the coronary artery growth is determined to be 433.5.

In the embodiment of the present invention, after obtaining the boundary position information of the aorta and the left ventricle, the image processing device removes the connected domain where the left ventricle is located in the three-dimensional image of the connected domain of the aorta to obtain the connected domain of the aorta and the coronary artery, and Coronary arteries and atrial appendages were de-adhesed in the connected areas of the aorta and coronary arteries to obtain images of the connected areas of the aorta and coronary arteries for growing coronary arteries. FIG. 5 is an exemplary connected domain image of the aorta and the coronary arteries provided by an embodiment of the present invention. As shown in FIG. 5 , the aorta 10 and the coronary arteries 20 are connected.

S104a2 , according to the position information of the boundary between the aorta and the left ventricle, delete the three-dimensional image of the connected area of the aorta from the initial three-dimensional image of the coronary artery to obtain the deleted three-dimensional image of the initial coronary artery.

In the embodiment of the present invention, after the image processing device obtains the image of the connected area of the aorta and the coronary artery, because the gray value corresponding to the connected area image of the aorta and the coronary artery, that is, the gray value of the coronary artery growth, is relatively high, the main The information of the coronary arteries included in the arterial and coronary connected region images is relatively small; therefore, the image processing device needs to remove the aortic connected region 3D image from the initial coronary 3D image according to the boundary position information of the aorta and the left ventricle Then, the regrowth of the coronary artery is performed; here, the three-dimensional image of the aortic connected area is removed from the initial three-dimensional image of the coronary artery, and the deleted initial three-dimensional image of the coronary artery is obtained.

Specifically, because the image processing device divides the three-dimensional image of the aorta connected area into two parts according to the position information of the aorta and the left ventricle: the connected area where the coronary arteries are located and the connected area where the ventricle is located, therefore, the image processing device uses the initial three-dimensional image of the coronary arteries. The connected domain where the coronary artery is located and the connected domain where the ventricle is located are deleted respectively in the 3D image; wherein, after the image processing device expands the connected domain where the ventricle is located according to a preset degree, the connected domain where the ventricle is located is deleted from the initial three-dimensional image of the coronary artery; and the image processing device The device directly deletes the connected region where the coronary artery is located from the initial three-dimensional image of the coronary artery.

Here, the deletion process represents setting the CT value information in the corresponding region to zero in the initial coronary three-dimensional image.

Exemplarily, FIG. 6 is a plan view of an exemplary deleted initial three-dimensional coronary artery image provided by an embodiment of the present invention. As shown in FIG. 6 , the left picture is a plan view of the 101st slice of the initial three-dimensional coronary artery image; The figure on the right is a plan view of the 101st slice of the initial coronary 3D image after clearing the aortic connected area, that is, the 101st slice plan of the deleted initial coronary 3D image.

It can be understood that the image processing device deletes the connected domain where the left ventricle is located in the initial three-dimensional coronary artery image, so that the atrial appendage, left atrium, and left ventricle in the initial three-dimensional coronary artery image will be cleared, and the subsequent coronary artery regrowth has been removed. The cleared atrial appendage, left atrium, and left ventricle will not grow again, and will no longer be affected by the atrial appendage and other parts that are easy to adhere to the coronary arteries. The image processing device deletes the connected domain where the coronary artery is located in the initial three-dimensional image of the coronary artery, so as to reduce the gray value in the subsequent coronary regrowth process to construct the connected domain, the connected domain that is connected to the coronary artery and the aorta will not be connected. Then there is a whole large connected domain, but many small connected domains, and these small connected domains are not connected with each other. Once a small connected domain has adhered to other tissues, this small connected domain can be chosen to be discarded, while other small connected domains can grow normally.

S104a3, performing coronary artery growth based on the gray value of the coronary artery growth and the deleted initial three-dimensional image of the coronary artery to obtain a three-dimensional image of the coronary artery growth.

In the embodiment of the present invention, after obtaining the gray value of coronary artery growth and the deleted initial three-dimensional image of coronary artery, the image processing device performs a cyclic operation so that the gray value of coronary artery growth decreases by a preset growth step each time, and The connected domain that grows after each reduction of the gray value is marked until the gray value of the coronary artery growth is reduced to the preset minimum growth gray value, and all the marked connected domains are counted, that is, the three-dimensional coronary artery growth is obtained. image.

Further, in the embodiment of the present invention, the image processing device in S103d performs coronary artery growth based on the grayscale value of the coronary artery growth and the deleted initial three-dimensional image of the coronary artery to obtain a three-dimensional image of the coronary artery growth, which specifically includes: an image processing device. According to the coronary artery growth gray value and the preset growth step size, the coronary artery generation is performed on the deleted initial coronary three-dimensional image until the coronary artery growth gray value is reduced to the preset minimum growth gray value, and the coronary artery growth three-dimensional image is obtained. Image, complete coronary growth.

It should be noted that the image processing device judges the number of pixels in each marked connected domain when growing the coronary arteries. When the number of pixels in the marked connected domain is greater than a preset number threshold (for example, 100,000) , the marked connected domain is discarded; and when the number of pixels in the marked connected domain is not greater than the preset number threshold, the marked connected domain is added to the three-dimensional image of coronary growth, and at the same time, the marked connected domain is Domains were removed in the initial coronary 3D image after removal. The image processing device repeats this operation until the coronary artery growth gray value is reduced to the preset minimum growth gray value, and the coronary artery growth is completed, and a three-dimensional image of the coronary artery growth is obtained, as shown in FIG. 7 .

S104b, using a preset algorithm to identify the aorta region on the three-dimensional image of coronary artery growth.

In the embodiment of the present invention, since the coronary artery three-dimensional image includes two parts, the coronary artery region and the aorta region, the image processing device obtains the coronary artery region only by removing the aortic region in the coronary artery growth three-dimensional image. The target image, namely the three-dimensional image of the coronary artery. Here, the image processing device performs the identification of the aorta region on the three-dimensional image of the coronary artery growth according to the preset algorithm.

Preferably, the preset algorithm is an algorithm of first eroding and then expanding, and the image processing device can remove the coronary artery region in the three-dimensional image of coronary artery growth by first eroding and then expanding the three-dimensional image of coronary artery growth.

It should be noted that the erosion algorithm uniformly shrinks the shape in the three-dimensional binary matrix according to the specified rules, which is equivalent to removing several layers of points near the outer layer of the shape. The expansion algorithm is the opposite of the erosion algorithm.

S104c, delete the aorta region from the three-dimensional image of the coronary artery growth to obtain a three-dimensional image of the coronary artery.

In the embodiment of the present invention, after the image processing device identifies the aortic region in the three-dimensional image of coronary artery growth, the aortic region is deleted from the three-dimensional image of coronary artery growth, for example, the three-dimensional image of coronary artery growth and the aortic region are combined If the difference is poor, a three-dimensional image of the coronary arteries with only two maximal connected domains remaining after removing the aorta is obtained, wherein the two maximal connected domains represent the left coronary artery and the right coronary artery.

Preferably, after obtaining the three-dimensional image of the coronary artery, the image processing device extracts isosurface data from the three-dimensional image of the coronary artery, and uses a preset grid algorithm (such as a Poisson surface reconstruction algorithm) to perform a curved surface on the isosurface data Reconstruction, and smooth the surface reconstruction results to obtain a three-dimensional mesh model of the coronary artery, as shown in Figure 8. That is to say, the three-dimensional image of the coronary artery is a solid stereoscopic image, and the image processing apparatus processes the solid stereoscopic image into a hollow stereoscopic image as a basis for medical diagnosis.

It can be understood that, because the image processing device accurately identifies the aorta and coronary arteries in the initial three-dimensional image of the coronary arteries according to the morphological structures of the aorta and the physiological structure of the aorta, and based on the identified aorta and left The location information of the ventricular boundary is used to automatically grow the coronary artery, thereby obtaining a three-dimensional image of the coronary artery, realizing the automatic and accurate extraction of the coronary artery, improving the intelligence of coronary artery acquisition, and improving the effect of coronary artery acquisition.

Embodiment 2

Based on the same inventive concept of the first embodiment, an embodiment of the present invention provides an image processing apparatus 30 , and FIG. 9 is a schematic structural diagram of an image processing apparatus provided by the first embodiment of the present invention. As shown in FIG. 9 , the image processing apparatus 30 includes:

The first identification unit 300 is configured to perform aortic identification in the initial three-dimensional image of the coronary artery based on the shape of the aorta when the initial three-dimensional image of the coronary artery is obtained to obtain the position information of the aorta;

a determining unit 301, configured to determine a three-dimensional image of aortic connectivity from the initial three-dimensional coronary artery image based on the aortic position information;

The second identification unit 302 is configured to perform cardiac sinus identification in the three-dimensional image of the aortic connected domain based on the aortic position information and the aortic physiological structure to obtain the boundary position information of the aorta and the left ventricle;

The regrowth unit 303 is configured to perform coronary artery regrowth based on the position information of the boundary between the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery.

Further, the image processing apparatus 30 further includes a superimposing unit 304, and the superimposing unit 304 is configured to, when acquiring at least two two-dimensional coronary angiography images, select from the at least two two-dimensional coronary angiography images. respectively acquire corresponding image grayscale correction information, where the image grayscale correction information includes initial grayscale values and format information corresponding to the at least two two-dimensional coronary angiography images; and according to the initial grayscale values and the format information to obtain the density values corresponding to the at least two two-dimensional coronary angiography images; and superimposing the at least two two-dimensional coronary angiography images to obtain the initial Three-dimensional image of coronary arteries.

Further, the first identifying unit 300 is specifically configured to determine a radius threshold value according to the shape of the aorta; and based on the preset image hierarchy order, according to the preset gray value to determine the initial coronary artery three-dimensional image. Perform binarization processing on the hierarchical coronary 3D image, and identify the aorta in the result image corresponding to the binarization of the hierarchical coronary 3D image according to the preset positioning algorithm and the radius threshold, until the obtained the aortic location information.

Further, the determining unit 301 is specifically configured to obtain an initial aortic gray value according to the aortic position information; and determine the aortic gray value according to the preset gray value and the initial aortic gray value and performing binarization processing on the initial three-dimensional coronary artery image according to the grayscale value of the aorta to obtain a binarized initial three-dimensional coronary artery image; and according to the aortic position information, from the The aortic connected area is determined in the initial three-dimensional image of the coronary artery after the binarization, and the three-dimensional image of the aortic connected area is obtained.

Further, the second identification unit 302 is specifically configured to determine a position determination condition according to the physiological structure of the aorta, and the position determination condition is used to identify the cardiac sinus; and a three-dimensional image of the aorta connected domain , the search plane is determined from the aorta position information according to the preset search step, until the aortic cross-sectional area corresponding to the search plane meets the position determination condition, the corresponding search plane is used as the main search plane. Information on the location of the arterial-left ventricular boundary.

Further, the regrowth unit 303 is specifically configured to perform coronary artery regrowth based on the boundary position information of the aorta and the left ventricle to obtain a three-dimensional image of coronary artery growth; Identifying an aortic region on the image; and deleting the aortic region from the three-dimensional image of the coronary artery growth to obtain the three-dimensional image of the coronary artery.

Further, the regrowth unit 303 is further specifically configured to determine the gray value of the coronary artery growth according to the boundary position information of the aorta and the left ventricle and the aorta position information; and according to the aorta and the left ventricle Ventricular boundary position information, deleting the three-dimensional image of the aortic connected area from the initial three-dimensional image of the coronary artery to obtain a three-dimensional image of the initial coronary artery after deletion; and based on the gray value of the coronary artery growth and the deleted three-dimensional image The coronary artery growth is performed on the initial three-dimensional image of the coronary artery to obtain the three-dimensional image of the coronary artery growth.

Further, the regrowth unit 303 is further specifically configured to obtain a coronary artery connected region according to the boundary position information of the aorta and the left ventricle; and when the volume corresponding to the coronary artery connected region is greater than a preset volume, in the On the basis of the aortic gray value corresponding to the aortic position information, the preset de-adhesion step is cyclically increased until the volume corresponding to the coronary connected domain is not greater than the preset volume, and the cyclically increased gray The degree value is taken as the gray value of the coronary artery growth.

Further, the regrowth unit 303 is further specifically configured to perform coronary artery generation on the deleted initial three-dimensional image of the coronary artery according to the gray value of the coronary artery growth and the preset growth step, until the coronary artery is generated. The arterial growth gray value is reduced to a preset minimum growth gray value to complete the growth of the coronary artery, and obtain the three-dimensional image of the coronary artery growth.

It should be noted that, in practical applications, the above-mentioned first identification unit 300 , determination unit 301 , second identification unit 302 and regeneration unit 303 may be implemented by the processor 304 located on the image processing device 30 , specifically a CPU (Central Processing Unit) , central processing unit), MPU (Microprocessor Unit, microprocessor), DSP (Digital Signal Processing, digital signal processor) or field programmable gate array (FPGA, Field Programmable Gate Array) and other implementations.

An embodiment of the present invention further provides an image processing apparatus 30 , as shown in FIG. 10 , the image processing apparatus 30 includes: a processor 304 , a memory 305 and a communication bus 306 , and the memory 305 communicates with the communication bus 306 through the communication bus 306 . The processor 304 communicates, and the memory 305 stores a program executable by the processor 304. When the program is executed, the image processing method according to the first embodiment is executed by the processor 305.

In practical applications, the above-mentioned memory 305 may be a volatile memory (volatile memory), such as a random-access memory (Random-Access Memory, RAM); or a non-volatile memory (non-volatile memory), such as a read-only memory ( Read-Only Memory (ROM), flash memory (flash memory), hard disk (Hard Disk Drive, HDD) or solid-state drive (Solid-State Drive, SSD); or a combination of the above types of memory, and provide the processor 304 with instructions and data.

An embodiment of the present invention provides a computer-readable storage medium on which a program is stored, and when the program is executed by the processor 304, the image processing method described in the first embodiment is implemented.

It can be understood that, because the image processing device accurately identifies the aorta and coronary arteries in the initial three-dimensional image of the coronary arteries according to the morphological structures of the aorta and the physiological structure of the aorta, and based on the identified aorta and left The location information of the ventricular boundary is used to automatically grow the coronary artery, thereby obtaining a three-dimensional image of the coronary artery, realizing the automatic and accurate extraction of the coronary artery, improving the intelligence of coronary artery acquisition, and improving the effect of coronary artery acquisition.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media having computer-usable program code embodied therein, including but not limited to disk storage, optical storage, and the like.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention.

Claims (10)

1. An image processing method, characterized in that the method comprises:

when an initial coronary artery three-dimensional image is obtained, carrying out aorta identification in the initial coronary artery three-dimensional image based on the aorta form to obtain aorta position information;

determining an aorta connected domain three-dimensional image from the initial coronary artery three-dimensional image based on the aorta position information;

based on the aorta position information and the aorta physiological structure, performing cardiac sinus recognition in the aorta connected domain three-dimensional image to obtain boundary position information of an aorta and a left ventricle;

performing coronary artery regrowth based on the boundary position information of the aorta and the left ventricle to obtain a coronary artery three-dimensional image;

wherein, the regrowth of the coronary artery based on the boundary position information of the aorta and the left ventricle to obtain a three-dimensional image of the coronary artery comprises the following steps:

performing regrowth of coronary arteries based on the boundary position information of the aorta and the left ventricle to obtain a coronary artery growth three-dimensional image;

identifying an aorta region on the coronary artery growth three-dimensional image by using a preset algorithm;

and deleting the aorta area from the coronary artery growth three-dimensional image to obtain the coronary artery three-dimensional image.

2. The method of claim 1, wherein before performing aorta identification in the initial coronary artery three-dimensional image based on aorta morphology when the initial coronary artery three-dimensional image is acquired to obtain aorta position information, the method further comprises:

when at least two coronary artery angiography two-dimensional images are obtained, corresponding image gray scale correction information is respectively obtained from the at least two coronary artery angiography two-dimensional images, and the image gray scale correction information comprises initial gray scale values and format information corresponding to the at least two coronary artery angiography two-dimensional images;

obtaining density values corresponding to the at least two coronary angiography two-dimensional images according to the initial gray value and the format information;

and superposing the at least two coronary angiography two-dimensional images according to the density value and a preset image layer sequence to obtain the initial coronary three-dimensional image.

3. The method according to claim 1, wherein the aorta recognition is performed in the initial coronary artery three-dimensional image based on the aorta morphology to obtain aorta position information, comprising:

determining a radius threshold from the aorta morphology;

and performing binarization processing on the hierarchical coronary artery three-dimensional image in the initial coronary artery three-dimensional image according to a preset gray value based on a preset image hierarchical sequence, and identifying the aorta in a result image corresponding to the hierarchical coronary artery three-dimensional image subjected to binarization processing according to a preset positioning algorithm and the radius threshold value until the aorta position information is obtained.

4. The method of claim 1, wherein determining an aorta connected domain three-dimensional image from the initial coronary artery three-dimensional image based on the aorta position information comprises:

obtaining an initial aorta gray value according to the aorta position information;

determining an aorta gray value according to a preset gray value and the initial aorta gray value;

according to the aorta gray value, carrying out binarization processing on the initial coronary artery three-dimensional image to obtain a binarized initial coronary artery three-dimensional image;

and determining an aorta communication region from the binarized initial coronary artery three-dimensional image according to the aorta position information to obtain the aorta communication region three-dimensional image.

5. The method according to claim 1, wherein the performing cardiac sinus recognition in the three-dimensional image of the connected domain of aorta based on the aortic position information and aortic physiological structures to obtain aortic and left ventricular boundary position information comprises:

determining a position judgment condition according to the aorta physiological structure, wherein the position judgment condition is used for identifying the cardiac sinus;

and determining a search plane from the aorta position information according to a preset search step length in the aorta connected domain three-dimensional image until the aorta section area corresponding to the search plane meets the position judgment condition, and taking the corresponding search plane as the aorta and left ventricle boundary position information.

6. The method of claim 5, wherein the re-growing of coronary arteries based on the aortic and left ventricular boundary location information, resulting in a three-dimensional image of coronary artery growth, comprises:

determining a coronary artery growth gray value according to the aorta and left ventricle boundary position information and the aorta position information;

deleting the three-dimensional image of the aorta communication domain from the initial coronary artery three-dimensional image according to the boundary position information of the aorta and the left ventricle to obtain a deleted initial coronary artery three-dimensional image;

and performing coronary artery growth based on the coronary artery growth gray value and the deleted initial coronary artery three-dimensional image to obtain the coronary artery growth three-dimensional image.

7. The method of claim 6, wherein determining a coronary artery growth gray scale value from the aorta and left ventricle boundary position information and the aorta position information comprises:

obtaining a coronary artery communication domain according to the boundary position information of the aorta and the left ventricle;

when the volume corresponding to the coronary artery communication domain is larger than a preset volume, circularly increasing a preset adhesion removing step length on the basis of the aorta gray value corresponding to the aorta position information until the volume corresponding to the coronary artery communication domain is not larger than the preset volume, and taking the gray value after circular increase as the coronary artery growth gray value.

8. The method of claim 6, wherein the performing coronary artery growth based on the coronary artery growth gray scale value and the deleted initial coronary artery three-dimensional image to obtain the coronary artery growth three-dimensional image comprises:

and according to the coronary artery growth gray value and a preset growth step length, performing coronary artery generation on the deleted initial coronary artery three-dimensional image until the coronary artery growth gray value is reduced to a preset minimum growth gray value, finishing coronary artery growth, and obtaining the coronary artery growth three-dimensional image.

9. An image processing apparatus, characterized in that the apparatus comprises: a processor, a memory and a communication bus, the memory in communication with the processor through the communication bus, the memory storing a program executable by the processor, the program, when executed, causing the processor to perform the method of any of claims 1-8.

10. A computer-readable storage medium, on which a program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1 to 8.

Images