A rapid reconstruction method and system for cardiac CT images

Abstract

The present invention relates to the field of image enhancement technology, and in particular to a method and system for rapid reconstruction of cardiac CT images. The method obtains an initial cardiac CT image; obtains the degree of cardiac possibility according to the grayscale value and grayscale superposition of the pixel points in the initial cardiac CT image, and then screens out suspected cardiac pixels; corrects the degree of cardiac possibility of the suspected cardiac pixel points according to the distance between the suspected cardiac pixel points and the centroid of the heart, the grayscale value difference between the suspected cardiac pixel points and the pixels of its preset neighborhood, and the distribution of the grayscale value in the preset window of the suspected cardiac pixel points and the fluctuation of the grayscale value in the background area, obtains the final degree of cardiac possibility of the suspected cardiac pixel points, determines the cardiac pixel points, and reconstructs the cardiac CT image. The present invention removes the halo blurring effect in the reconstructed cardiac CT image by determining the cardiac pixel points in the suspected cardiac pixel points, thereby improving the quality of the reconstructed cardiac CT image and the accuracy of the cardiac analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of image enhancement, and in particular to a method and system for rapid reconstruction of cardiac CT images. Background Art

[0002] The rapid reconstruction technology of cardiac CT images is an important branch in the field of medical imaging, which involves the evaluation of cardiac structure and function. In the existing method, a CT scanning device is used to scan the patient's heart, obtain the X-ray projection data of the patient's heart at different angles and perform back-projection processing, and then obtain the cardiac CT back-projection images at different angles, and superimpose the cardiac CT back-projection images at all angles to complete the reconstruction of the cardiac CT image.

[0003] In actual situations, when cardiac CT back-projection images at different angles are superimposed, halos will be generated in the obtained cardiac CT images, resulting in blurring around the heart, making the quality of the rapidly reconstructed cardiac CT images poor and making it impossible to accurately evaluate the cardiac structure and function. Summary of the invention

[0004] In order to solve the technical problem that the halo generated in the reconstructed cardiac CT image leads to blurring around the heart and inability to accurately evaluate the cardiac structure and function, the object of the present invention is to provide a method and system for rapid reconstruction of cardiac CT images. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for rapid reconstruction of cardiac CT images, the method comprising the following steps:

[0006] Superimposing cardiac CT back-projection images at different angles to obtain an initial cardiac CT image;

[0007] According to the gray value and gray superposition of each pixel in the initial cardiac CT image, as well as the gray change of two adjacent superpositions, the cardiac possibility degree of each pixel is obtained;

[0008] Screen out suspected heart pixels based on the possible degree of the heart, and obtain the close combination degree of each suspected heart pixel according to the distance between each suspected heart pixel and the centroid of the heart and the gray value difference between each suspected heart pixel and its preset neighboring pixels;

[0009] The signal strength of each suspected heart pixel is obtained according to the distribution of grayscale values ​​in a preset window of each suspected heart pixel and the fluctuation of grayscale values ​​in the background area;

[0010] According to the close combination degree, signal strength and cardiac possibility degree of each suspected cardiac pixel point, a final cardiac possibility degree of each suspected cardiac pixel point is obtained;

[0011] The cardiac pixels are determined based on the final cardiac probability and the cardiac probability, and the cardiac CT image is reconstructed.

[0012] Furthermore, the method for obtaining the possible degree of the heart is:

[0013] According to the grayscale superposition of each pixel in the initial cardiac CT image and the grayscale change of two adjacent superpositions, the regular growth degree of each pixel is obtained;

[0014] The result of normalizing the product of the gray value of each pixel and the regular growth degree is taken as the possible degree of the heart of each pixel.

[0015] Furthermore, the method for obtaining the regular growth degree is:

[0016] For any pixel point in the initial cardiac CT image, the number of times the grayscale accumulated value of the pixel point increases during the superposition process is obtained as the number of grayscale increases of the pixel point;

[0017] The ratio of the grayscale growth times to the superposition times is used as a first regular growth reference value of the pixel point;

[0018] Obtain the ratio of the grayscale accumulated value of the pixel after each superposition to the grayscale accumulated value after the previous superposition as the grayscale change value of the corresponding superposition;

[0019] The average of all grayscale change values ​​is used as the second law growth reference value of the pixel point;

[0020] The product of the first regular growth reference value and the second regular growth reference value is taken as the regular growth degree of the pixel point.

[0021] Furthermore, the method for acquiring the suspected heart pixel point is:

[0022] When the heart possibility level is less than or equal to a preset heart possibility level threshold, the corresponding pixel point is regarded as a suspected heart pixel point.

[0023] Furthermore, the method for obtaining the heart centroid is:

[0024] Normalizing the grayscale value of each pixel in the initial cardiac CT image, and when the normalized grayscale value is greater than a first preset grayscale threshold, using the corresponding pixel as a cardiac reference pixel;

[0025] The area formed by the heart reference pixel points is taken as the heart range area, and the centroid of the heart range area is obtained as the heart centroid.

[0026] Furthermore, the method for obtaining the degree of tight binding is:

[0027] For any suspected heart pixel point, obtain the Euclidean distance between the suspected heart pixel point and the heart centroid as the first distance;

[0028] The gradient direction of the suspected heart pixel is obtained by using a Sobel operator, and a preset number of pixels closest to the suspected heart pixel on the straight line where the gradient direction is located are used as preset neighborhood pixels of the suspected heart pixel;

[0029] Obtaining the grayscale value difference between the suspected heart pixel and each of its preset neighborhood pixels, all as the first difference;

[0030] The mean value of the first difference is used as the local contrast of the suspected heart pixel;

[0031] The close combination degree of the suspected heart pixel point is obtained according to the first distance and the local contrast; wherein the first distance and the close combination degree are negatively correlated, and the local contrast and the close combination degree are positively correlated.

[0032] Furthermore, the signal strength is obtained by:

[0033] For any suspected heart pixel, obtaining the mean grayscale value of the suspected heart pixel in a preset window as the reference grayscale value of the suspected heart pixel;

[0034] The pixels in the initial cardiac CT image whose grayscale values ​​are less than a second preset grayscale threshold are all used as background reference pixels;

[0035] The signal strength of the suspected heart pixel is obtained according to the standard deviation of the reference grayscale value and the grayscale value of the background reference pixel; the reference grayscale value is positively correlated with the signal strength, and the standard deviation of the grayscale value of the background reference pixel is negatively correlated with the signal strength.

[0036] Furthermore, the method for obtaining the final cardiac probability level is:

[0037] For any suspected heart pixel, the product of the close combination degree of the suspected heart pixel and the signal strength is normalized as the correction weight of the suspected heart pixel;

[0038] The product of the correction weight of the suspected heart pixel and the heart possibility degree is used as the final heart possibility degree of the suspected heart pixel.

[0039] Furthermore, the method for acquiring the heart pixel points is:

[0040] When the heart probability level is greater than a preset heart probability level threshold, the corresponding pixel point is regarded as a heart pixel point;

[0041] When the final heart possibility degree is greater than the preset final heart possibility degree threshold, the corresponding suspected heart pixel point is taken as the heart pixel point.

[0042] In a second aspect, another embodiment of the present invention provides a rapid reconstruction system for cardiac CT images, the system comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of any one of the above methods are implemented.

[0043] The present invention has the following beneficial effects:

[0044] The present invention obtains the cardiac possibility degree of each pixel point according to the grayscale value and grayscale superposition of each pixel point in the initial cardiac CT image, as well as the grayscale change of two adjacent superpositions, and preliminarily reflects the possibility that each pixel point in the initial cardiac CT image is a cardiac pixel point; in order to avoid the halo blurring effect in the reconstructed cardiac CT image, the suspected cardiac pixels are screened out based on the cardiac possibility degree, and then according to the distance between each suspected cardiac pixel point and the heart's centroid and the grayscale value difference between each suspected cardiac pixel point and its preset neighborhood pixels, the close combination degree of each suspected cardiac pixel point is obtained, and preliminarily reflects the possibility that each suspected cardiac pixel point is a cardiac pixel point; in order to more accurately analyze the possibility that each suspected cardiac pixel point is a cardiac pixel point, the suspected cardiac pixel point is further screened out according to each suspected cardiac pixel point. The signal strength of each suspected cardiac pixel is obtained based on the distribution of grayscale values ​​within the preset window of the pixel and the fluctuation of grayscale values ​​in the background area, which further reflects the possibility that each suspected cardiac pixel is a cardiac pixel; in order to determine the possibility that each suspected cardiac pixel is a cardiac pixel, the final cardiac possibility degree of each suspected cardiac pixel is obtained based on the close combination degree, signal strength and cardiac possibility degree of each suspected cardiac pixel, accurately reflecting the possibility that each suspected cardiac pixel is a cardiac pixel; the cardiac pixels are accurately determined based on the final cardiac possibility degree and the cardiac possibility degree, the halo blur in the initial cardiac CT image is eliminated, so that the quality of the reconstructed cardiac CT image is better, and the structure and function of the heart can be more accurately analyzed through the reconstructed cardiac CT image. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A schematic flow chart of a method for rapid reconstruction of cardiac CT images provided by one embodiment of the present invention;

[0047] Figure 2 A transverse CT image of the heart without halo effect provided by an embodiment of the present invention;

[0048] Figure 3 A flow chart of a method for obtaining a possible degree of heart disease provided by an embodiment of the present invention;

[0049] Figure 4 A flow chart of a method for obtaining a degree of close integration provided by an embodiment of the present invention;

[0050] Figure 5 A structural diagram of a rapid reconstruction system for cardiac CT images provided by an embodiment of the present invention;

[0051] Figure 6 A schematic diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a method and system for rapid reconstruction of cardiac CT images proposed by the present invention, its specific implementation, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0053] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0054] The specific scheme of a rapid reconstruction method and system for cardiac CT images provided by the present invention is described in detail below with reference to the accompanying drawings.

[0055] Embodiment 1:

[0056] The specific scenario of this embodiment is: in order to avoid the presence of halos in the reconstructed cardiac CT images, resulting in blurring around the heart and making it impossible to accurately evaluate the cardiac structure and function, this embodiment obtains an initial cardiac CT image, analyzes the grayscale value and grayscale superposition of each pixel in the initial cardiac CT image, and the grayscale change of two adjacent superpositions, obtains the cardiac possibility degree of each pixel, and then determines the suspected cardiac pixel based on the cardiac possibility degree, analyzes the position and local grayscale value of the suspected cardiac pixel, obtains the tight combination degree and signal strength of each suspected cardiac pixel, and then corrects the cardiac possibility degree of each suspected cardiac pixel to obtain the final cardiac possibility degree of each suspected cardiac pixel, and finally determines the cardiac pixel in the initial cardiac CT image according to the final cardiac possibility degree and the cardiac possibility degree, quickly reconstructs the cardiac CT image, avoids the generation of halos around the heart, improves the quality and integrity of the cardiac CT image, and ensures accurate evaluation of the structure and function of the heart.

[0057] The present invention proposes a rapid reconstruction method for cardiac CT images. Figure 1 , which shows a schematic flow chart of a method for rapid reconstruction of cardiac CT images provided by an embodiment of the present invention, the method comprising the following steps:

[0058] Step S1: Superimposing cardiac CT back-projection images at different angles to obtain an initial cardiac CT image.

[0059] Specifically, this embodiment uses a patient's heart as an example for analysis, and the heart that appears later refers to the patient's heart. The heart is scanned by a CT scanning device to obtain X-ray projection data of the heart at different angles, wherein the X-ray projection data is the X-ray intensity information after attenuation by the patient's body tissue, that is, the X-rays at different angles are received by the detector after passing through the patient's body tissue. These signals represent the degree of absorption of X-rays by different tissues of the patient's body. In order to facilitate better subsequent analysis, this embodiment uses the Filtered Back-Projection (FBP) algorithm to perform back-projection processing on the X-ray projection data at different angles to obtain heart CT back-projection images at different angles. Among them, the Filtered Back-Projection (FBP) algorithm is a well-known technology and will not be described in detail. Then, the heart CT back-projection images at different angles are superimposed to obtain the initial heart CT image, that is, the initial reconstructed heart CT image. Among them, there is relatively obvious halo interference around the heart in the initial cardiac CT image. This embodiment mainly analyzes and processes the halo interference in the initial cardiac CT image, so that the final cardiac CT image does not have halo interference, ensuring accurate evaluation of the patient's heart through the cardiac CT image.

[0060] It should be noted that the patient needs to remain still during the scanning time required to generate the cardiac CT image, and the electric table is used to move the patient's position, and the X-ray scanning is stopped during the movement, and the patient is required to hold his breath during the data acquisition process. Among them, different angles are different angles of the patient's heart in a cross section, and the reconstructed cardiac CT image is the heart cross-sectional image of the corresponding cross section.

[0061] Since the reflected projection images separated by 180° are mirror images of each other, when obtaining all the cardiac CT back-projection images required to generate the reconstructed cardiac CT image, only the angle increase of half a circle needs to be considered, that is, the angle range of the X-ray projection data at different angles is For all the desired rotation angles within the range, this embodiment sets each rotation angle to 1 degree. The implementer can set the size of each rotation angle of the X-ray according to the actual situation, which is not limited here.

[0062] Step S2: Obtain the cardiac probability of each pixel point according to the grayscale value and grayscale superposition of each pixel point in the initial cardiac CT image, as well as the grayscale change between two adjacent superpositions.

[0063] In actual situations, the heart as a whole has a high absorption characteristic for X-rays, and appears brighter in cardiac CT images. However, other tissues around the heart absorb X-rays less strongly, and appear darker in cardiac CT images. Figure 2 The figure shows a transverse CT image of the heart without the influence of halo. Therefore, in the initial cardiac CT image, if the gray value of a certain pixel point is larger, the pixel point is more likely to be a cardiac pixel point. In order to more accurately analyze the possibility of each pixel point being a cardiac pixel point, this embodiment further analyzes the grayscale superposition of each pixel point in the initial cardiac CT image and the grayscale change of two adjacent superpositions. It is known that the grayscale value of the cardiac pixel point in the cardiac CT back-projection image at different angles is also relatively large, and then the grayscale cumulative value of the cardiac pixel point must increase during each superposition of the cardiac CT back-projection image, and the grayscale cumulative value of each superposition corresponding to the cardiac pixel point must be greater than the grayscale cumulative value of the adjacent previous superposition. Then, this embodiment obtains the cardiac possibility degree of each pixel point according to the grayscale value and grayscale superposition of each pixel point in the initial cardiac CT image, as well as the grayscale change of two adjacent superpositions. Among them, the smaller the cardiac possibility degree, the less likely the corresponding pixel point is to be a cardiac pixel point.

[0064] Preferably, in one possible implementation of this embodiment, the method for obtaining the possible degree of heart disease is as follows: Figure 3 , which shows a flow chart of a method for obtaining the possible degree of heart disease provided by this embodiment, the method comprising the following steps:

[0065] Step S301: Obtain the regular growth degree of each pixel point according to the grayscale superposition of each pixel point in the initial cardiac CT image and the grayscale change between two adjacent superpositions.

[0066] It is known that the grayscale value of the heart pixel in each heart CT back-projection image is relatively large. Then, when a certain pixel in the initial heart CT image is in the process of superimposing the heart CT back-projection image, the more the number of times the grayscale cumulative value of the pixel increases is consistent with the number of times the heart CT back-projection image is superimposed, it means that the pixel is more likely to be a heart pixel. Among them, in the process of superimposing the heart CT back-projection image, the heart CT back-projection images are superimposed one by one in sequence. At the same time, when the grayscale cumulative value of the pixel after each superposition is greater than the grayscale cumulative value after the adjacent previous superposition, it means that the change law of the grayscale cumulative value of the pixel is more consistent with the change law of the grayscale cumulative value of the heart pixel. Then, this embodiment obtains the regular growth degree of each pixel according to the grayscale superposition of each pixel in the initial heart CT image and the grayscale change of two adjacent superpositions. The greater the regular growth degree, the more the change law of the grayscale cumulative value of the corresponding pixel is consistent with the change law of the grayscale cumulative value of the heart pixel, and the corresponding pixel is more likely to be a heart pixel.

[0067] Preferably, in a method that can be implemented in the present embodiment, the method for obtaining the degree of regular growth is: for any pixel point in the initial cardiac CT image, the number of times the grayscale cumulative value of the pixel point increases during the superposition process is obtained as the grayscale growth number of the pixel point; the ratio of the grayscale growth number to the superposition number is used as the first regular growth reference value of the pixel point; the larger the first regular growth reference value, the more the change pattern of the grayscale cumulative value of the pixel point conforms to the change pattern of the grayscale cumulative value of the cardiac pixel point, and the more likely the pixel point is a cardiac pixel point; in order to further analyze whether the change of the grayscale cumulative value of the pixel point during the accumulation of the cardiac CT back projection image conforms to the change of the grayscale cumulative value of the cardiac pixel point, and then obtain the image The ratio of the grayscale accumulated value of the pixel after each superposition to the grayscale accumulated value after the previous adjacent superposition is used as the grayscale change value of the corresponding superposition; the larger the grayscale change value, the more the change law of the grayscale accumulated value of the corresponding pixel point conforms to the change law of the grayscale accumulated value of the heart pixel point. In order to comprehensively analyze the change law of the grayscale accumulated value of the pixel point, the mean of all grayscale change values ​​is used as the second regular growth reference value of the pixel point; the larger the second regular growth reference value, the more the change law of the grayscale accumulated value of the pixel point conforms to the change law of the grayscale accumulated value of the heart pixel point, and the more likely the pixel point is a heart pixel point; and the product of the first regular growth reference value and the second regular growth reference value is used as the regular growth degree of the pixel point.

[0068] The calculation formula for the regular growth degree is: ; In the formula, is the regular growth degree of the ath pixel in the initial cardiac CT image; B is the number of superpositions of the cardiac CT back-projection image; is the grayscale growth times of the ath pixel in the initial cardiac CT image; is the grayscale accumulation value of the a-th pixel in the initial cardiac CT image after the b-th superposition; is the grayscale accumulation value of the a-th pixel in the initial cardiac CT image after the b-1-th superposition; is the first regular growth reference value of the a-th pixel in the initial cardiac CT image; is the grayscale change value of the b-th superposition of the a-th pixel in the initial cardiac CT image; is the second regular growth reference value of the ath pixel in the initial cardiac CT image.

[0069] It should be noted that if the grayscale cumulative value of the a-th pixel in the initial cardiac CT image after the first superposition is 0, the a-th pixel is not analyzed, and the regular growth degree of the a-th pixel is directly assumed to be 0, because the grayscale value of the cardiac pixel in the cardiac CT back-projection image must be greater than 0. The grayscale cumulative value can be greater than 255.

[0070] At this point, the regular growth degree of each pixel in the initial cardiac CT image is obtained.

[0071] Step S302: normalize the product of the gray value of each pixel and the regular growth degree, and use the result as the possible degree of the heart of each pixel.

[0072] It is known that the greater the regular growth degree, the more likely the corresponding pixel is a heart pixel; at the same time, in the initial heart CT image, when the gray value of a certain pixel is greater, the pixel is more likely to be a heart pixel, and then this embodiment normalizes the product of the gray value of each pixel in the initial heart CT image and the regular growth degree as the heart possibility degree of each pixel. The greater the heart possibility degree, the more likely the corresponding pixel is a heart pixel.

[0073] The calculation formula for the possible degree of heart disease is: ; In the formula, is the cardiac probability degree of the ath pixel in the initial cardiac CT image; is the regular growth degree of the ath pixel in the initial cardiac CT image; is the gray value of the ath pixel in the initial cardiac CT image; norm is the normalization function.

[0074] At this point, the cardiac probability level of each pixel in the initial cardiac CT image is obtained.

[0075] Step S3: Filter out suspected heart pixels based on the possible degree of the heart, and obtain the close connection degree of each suspected heart pixel according to the distance between each suspected heart pixel and the heart centroid and the gray value difference between each suspected heart pixel and its preset neighborhood pixels.

[0076] Specifically, since the cardiac CT back-projection image is acquired at different angles, the cardiac CT back-projection image will produce halo blur around the heart during the superposition process, wherein the gray value of the halo blur is relatively small. If the pixels with smaller gray values ​​in the initial cardiac CT image are directly removed to eliminate the halo blur effect, the rapidly reconstructed cardiac CT image will be incomplete, affecting the accurate assessment of the cardiac structure and function, and losing the significance of rapidly reconstructing the cardiac CT image. It is known that the greater the degree of possibility of the heart, the more likely the corresponding pixel is to be a cardiac pixel. Thus, this embodiment first screens out suspected cardiac pixels based on the degree of possibility of the heart, wherein the area formed by the suspected cardiac pixels must contain a halo blur area. In order to analyze the possibility that each suspected cardiac pixel is a cardiac pixel and avoid the incompleteness of the rapidly reconstructed cardiac CT image, this embodiment obtains the distance between each suspected cardiac pixel and the centroid of the heart. Because the heart is concentrated, the closer the distance between a suspected cardiac pixel and the centroid of the heart is, the more likely the suspected cardiac pixel is to be a cardiac pixel.

[0077] It is known that the halo blur is located around the heart, and there is an obvious brightness difference between the heart and the background area, which indirectly reflects that there is an obvious grayscale value difference between the heart pixel and the background pixel. It is known that the gradient direction of each suspected heart pixel is the most obvious reflection of the grayscale change direction of each suspected heart pixel. Then, this embodiment obtains the gradient direction of each suspected heart pixel in the initial heart CT image through the Sobel operator. When the grayscale value difference between a suspected heart pixel and the preset number of pixels closest to the straight line where the gradient direction is located is greater, it means that the suspected heart pixel is more likely to be a heart pixel. For better description, this embodiment uses the preset number of pixels closest to the suspected heart pixel on the straight line where the gradient direction of the suspected heart pixel is located as the preset neighborhood pixel of the suspected heart pixel. In this embodiment, the preset number is set to 20, wherein the preset neighborhood pixel of the suspected heart pixel must be located on both sides of the suspected heart pixel on the straight line where the gradient direction of the suspected heart pixel is located, and each side is 10 preset neighborhood pixel points. The implementer can set the size of the preset number according to the actual situation, which is not limited here. Among them, the Sobel operator is a well-known technology and will not be repeated. Furthermore, this embodiment obtains the degree of close binding of each suspected heart pixel point based on the distance between each suspected heart pixel point and the heart's centroid, and the difference in grayscale values ​​between each suspected heart pixel point and its preset neighborhood pixels. The greater the degree of close binding, the more likely the corresponding suspected heart pixel point is a heart pixel point.

[0078] Preferably, in one possible implementation of this embodiment, the method for obtaining the suspected heart pixel point is: when the heart possibility degree is less than or equal to the preset heart possibility degree threshold, the corresponding pixel point is used as the suspected heart pixel point. This embodiment sets the preset heart possibility degree threshold to 0.7, and the implementer can set the preset heart possibility degree threshold value according to the actual situation, which is not limited here.

[0079] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the heart's centroid is: normalize the grayscale value of each pixel in the initial heart CT image, and when the normalized grayscale value is greater than a first preset grayscale threshold, use the corresponding pixel as a heart reference pixel; this embodiment sets the first preset grayscale threshold to 0.8, and the implementer can set the size of the first preset grayscale threshold according to actual conditions, which is not limited here. The area formed by the heart reference pixel is used as the heart range area, and the centroid of the heart range area is obtained as the heart's centroid. Among them, the method for obtaining the regional centroid is a well-known technology and will not be repeated.

[0080] Preferably, in one possible implementation of this embodiment, the method for obtaining the degree of close bonding is as follows: Figure 4, which shows a flow chart of a method for obtaining the degree of close binding provided by this embodiment, the method comprising the following steps:

[0081] Step S401: Acquire a first distance.

[0082] For any suspected heart pixel point, the Euclidean distance between the suspected heart pixel point and the heart centroid is obtained as the first distance; the smaller the first distance, the more likely the suspected heart pixel point is located in the heart area, and the more likely the suspected heart pixel point is a heart pixel point. The method for obtaining the Euclidean distance is a well-known technology and will not be described in detail.

[0083] At this point, the first distance of each suspected heart pixel is obtained.

[0084] Step S402: obtaining local contrast.

[0085] The greater the difference in grayscale value between a suspected heart pixel and its preset neighborhood pixels, the more obvious the grayscale difference between the suspected heart pixel and its preset neighborhood pixels, that is, the greater the local contrast of the suspected heart pixel, the more likely the suspected heart pixel is to be a heart pixel.

[0086] Among them, the method for obtaining local contrast is: for any suspected heart pixel point, the absolute value of the difference between the grayscale value of the suspected heart pixel point and each of its preset neighborhood pixels is obtained, and each is used as the first difference; in order to comprehensively analyze the grayscale value difference between the suspected heart pixel point and its preset neighborhood pixels, the mean value of the first difference is used as the local contrast of the suspected heart pixel point.

[0087] At this point, the local contrast of each suspected heart pixel is obtained.

[0088] Step S403: Obtain the degree of close bonding.

[0089] It is known that the smaller the first distance is, the more likely the corresponding suspected heart pixel is to be a heart pixel; the larger the local contrast is, the more likely the corresponding suspected heart pixel is to be a heart pixel. Furthermore, this embodiment obtains the close combination degree of the suspected heart pixel based on the first distance and the local contrast; wherein the first distance is negatively correlated with the close combination degree, and the local contrast is positively correlated with the close combination degree.

[0090] The calculation formula for the tight binding degree is: ; In the formula, is the close integration degree of the cth suspected heart pixel; is the first distance of the cth suspected heart pixel; is the first preset constant, greater than 0; is the number of preset neighborhood pixels of the cth suspected heart pixel; is the gray value of the cth suspected heart pixel; is the gray value of the i-th preset neighborhood pixel of the c-th suspected heart pixel; is the absolute value function; is the first difference; is the local contrast of the cth suspected heart pixel.

[0091] This embodiment sets 0.1 to avoid the denominator being 0. The implementer can set it according to the actual situation. The size is not limited here.

[0092] At this point, the close integration degree of each suspected heart pixel is obtained.

[0093] Step S4: acquiring the signal strength of each suspected heart pixel according to the distribution of the grayscale value in the preset window of each suspected heart pixel and the fluctuation of the grayscale value in the background area.

[0094] In order to avoid accidental interference in the grayscale value of a suspected heart pixel, this embodiment constructs a preset window for each suspected heart pixel, and indirectly determines the reference grayscale value of each suspected heart pixel according to the distribution of the grayscale value in the preset window of each suspected heart pixel, wherein the larger the reference grayscale value, the more likely the corresponding suspected heart pixel is a heart pixel, because the grayscale value of the heart pixel is relatively large. It should be noted that if a suspected heart pixel is a boundary pixel, the part of the preset window corresponding to the suspected heart pixel is not in the initial heart CT image, and this embodiment only analyzes the grayscale value in the preset window part existing in the initial heart CT image.

[0095] Considering that there may be noise influence in the initial cardiac CT image, resulting in inaccurate reference grayscale value of each suspected cardiac pixel, and thus unable to accurately analyze the possibility of each suspected cardiac pixel being a cardiac pixel, in order to improve the accuracy of the reference grayscale value of each suspected cardiac pixel, this embodiment measures the influence of noise according to the fluctuation of grayscale value in the background area, and further analyzes the credibility of the reference grayscale value of each suspected cardiac pixel. Therefore, this embodiment obtains the signal strength of each suspected cardiac pixel according to the distribution of grayscale value in the preset window of each suspected cardiac pixel and the fluctuation of grayscale value in the background area. Among them, the greater the signal strength, the greater and more credible the reference grayscale value of the corresponding suspected cardiac pixel, and the more likely the corresponding suspected cardiac pixel is to be a cardiac pixel.

[0096] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the signal strength is: for any suspected heart pixel, obtain the mean value of the grayscale value in the preset window of the suspected heart pixel as the reference grayscale value of the suspected heart pixel; it is known that the grayscale value of the background area in the initial heart CT image is relatively small, and thus the second preset grayscale threshold is set to 10 in this embodiment, and the implementer can set the size of the second preset grayscale threshold according to actual conditions, which is not limited here. The pixels in the initial heart CT image whose grayscale value is less than the second preset grayscale threshold are all used as background reference pixels; in the absence of noise interference, the grayscale values ​​of the background reference pixels are similar, therefore, this embodiment quantifies the degree of noise influence in the initial heart CT image by the standard deviation of the grayscale value of the background reference pixel, and then corrects the reference grayscale value of the suspected heart pixel by the standard deviation of the grayscale value of the background reference pixel, and accurately obtains the signal strength of the suspected heart pixel. Therefore, this embodiment obtains the signal strength of the suspected heart pixel based on the standard deviation of the reference grayscale value and the grayscale value of the background reference pixel; wherein the reference grayscale value is positively correlated with the signal strength, and the standard deviation of the grayscale value of the background reference pixel is negatively correlated with the signal strength.

[0097] The calculation formula of signal strength is: ; In the formula, is the signal intensity of the cth suspected heart pixel; is the reference gray value of the cth suspected heart pixel; is the standard deviation of the grayscale value of the background reference pixel; It is the second preset constant, which is greater than 0.

[0098] This embodiment sets 1 to avoid the denominator being 0. The implementer can set it according to the actual situation. The size is not limited here.

[0099] At this point, the signal strength of each suspected heart pixel is obtained.

[0100] Step S5: According to the close combination degree, signal strength and cardiac possibility degree of each suspected cardiac pixel point, the final cardiac possibility degree of each suspected cardiac pixel point is obtained.

[0101] It is known that the greater the degree of close binding and the greater the signal strength, both indicate that the corresponding suspected heart pixel is more likely to be a heart pixel. Therefore, this embodiment corrects the heart possibility degree of the corresponding suspected heart pixel according to the degree of close binding and the signal strength of each suspected heart pixel, and then accurately obtains the final heart possibility degree of each suspected heart pixel. The greater the final heart possibility degree, the more likely the corresponding suspected heart pixel is to be a heart pixel. Then, based on the final heart possibility degree, the possibility of each suspected heart pixel being a heart pixel is accurately judged to avoid incompleteness of the reconstructed heart CT image.

[0102] Preferably, in a method that can be implemented in the present embodiment, the method for obtaining the final heart possibility degree is: for any suspected heart pixel point, the product of the tight binding degree of the suspected heart pixel point and the signal strength is normalized as the correction weight of the suspected heart pixel point; the larger the correction weight, the more accurate the heart possibility degree of the suspected heart pixel point; and then the product of the correction weight of the suspected heart pixel point and the heart possibility degree is used as the final heart possibility degree of the suspected heart pixel point.

[0103] Among them, the calculation formula for the final cardiac probability is: ; In the formula, is the final possible degree of heart of the cth suspected heart pixel; is the close integration degree of the cth suspected heart pixel; is the signal intensity of the cth suspected heart pixel; is the possible degree of heart of the cth suspected heart pixel; norm is the normalization function; To correct the weight.

[0104] At this point, the final heart probability degree of each suspected heart pixel is obtained.

[0105] Step S6: Determine the cardiac pixel points based on the final cardiac likelihood and the cardiac likelihood, and reconstruct the cardiac CT image.

[0106] For pixels in the initial cardiac CT image, cardiac pixels and suspected cardiac pixels are first determined preliminarily through the cardiac possibility degree, and then cardiac pixels in the suspected cardiac pixels are further determined according to the final cardiac possibility degree of the suspected cardiac pixels. Finally, the cardiac pixels in the initial cardiac CT image are accurately determined, and the grayscale values ​​of the cardiac pixels are kept unchanged, and the grayscale values ​​of the non-cardiac pixels are set to the mean grayscale values ​​of the non-cardiac pixels, so that a cardiac CT image with better quality can be accurately and quickly reconstructed, while avoiding the influence of halo blur, which is conducive to accurate evaluation of cardiac structure and function.

[0107] Preferably, in an achievable manner of this embodiment, the method for acquiring heart pixel points is: when the heart possibility degree is greater than a preset heart possibility degree threshold, the corresponding pixel point is used as a heart pixel point; when the final heart possibility degree is greater than the preset final heart possibility degree threshold, the corresponding suspected heart pixel point is used as a heart pixel point. In this embodiment, the preset final heart possibility degree threshold is set to 0.7, and the implementer can set the size of the preset final heart possibility degree threshold according to actual conditions, which is not limited here.

[0108] In summary, the present embodiment obtains an initial cardiac CT image; obtains the cardiac possibility degree according to the grayscale value and grayscale superposition of the pixel points in the initial cardiac CT image, and then screens out the suspected cardiac pixel points; corrects the cardiac possibility degree of the suspected cardiac pixel points according to the distance between the suspected cardiac pixel points and the centroid of the heart, the grayscale value difference between the suspected cardiac pixel points and the preset neighborhood pixel points, and the distribution of the grayscale value within the preset window of the suspected cardiac pixel points and the fluctuation of the grayscale value within the background area, obtains the final cardiac possibility degree of the suspected cardiac pixel points, determines the cardiac pixel points, and reconstructs the cardiac CT image. The present invention removes the halo blurring effect in the reconstructed cardiac CT image by determining the cardiac pixel points in the suspected cardiac pixel points, thereby improving the quality of the reconstructed cardiac CT image and the accuracy of the cardiac analysis.

[0109] Embodiment 2:

[0110] The present invention also proposes a rapid reconstruction system for cardiac CT images, see Figure 5 , which shows a structural diagram of a rapid reconstruction system for cardiac CT images provided by an embodiment of the present invention, the system comprising: an acquisition module 10, a cardiac possibility degree acquisition module 20, a close connection degree acquisition module 30, a signal strength acquisition module 40, a final cardiac possibility degree acquisition module 50 and a processing module 60.

[0111] The acquisition module 10 is used to superimpose the cardiac CT back-projection images at different angles to acquire an initial cardiac CT image.

[0112] The cardiac probability acquisition module 20 is used to acquire the cardiac probability of each pixel point in the initial cardiac CT image according to the gray value and gray superposition of each pixel point and the gray change between two adjacent superpositions.

[0113] The close connection degree acquisition module 30 is used to screen out suspected heart pixels based on the possible degree of the heart, and obtain the close connection degree of each suspected heart pixel according to the distance between each suspected heart pixel and the heart's centroid and the grayscale value difference between each suspected heart pixel and its preset neighborhood pixels.

[0114] The signal strength acquisition module 40 is used to acquire the signal strength of each suspected heart pixel according to the distribution of the grayscale value in the preset window of each suspected heart pixel and the fluctuation of the grayscale value in the background area.

[0115] The final cardiac likelihood degree acquisition module 50 is used to acquire the final cardiac likelihood degree of each suspected cardiac pixel point according to the close combination degree, signal strength and cardiac likelihood degree of each suspected cardiac pixel point.

[0116] The processing module 60 is used to determine the cardiac pixel points based on the final cardiac probability degree and the cardiac probability degree, and reconstruct the cardiac CT image.

[0117] It should be noted that: the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the rapid reconstruction system of cardiac CT images and the rapid reconstruction method of cardiac CT images provided in the above embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0118] Embodiment 3:

[0119] The present invention also proposes a device for rapid reconstruction of cardiac CT images, which includes a memory and a processor, wherein the memory stores executable program code, and the processor is used to call and execute the executable program code to perform a rapid reconstruction method of cardiac CT images provided in an embodiment of the present application. The device can be a chip, a component or a module, and the chip can include a connected processor and a memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute a rapid reconstruction method of cardiac CT images provided in the above embodiment.

[0120] In addition, the present application embodiment also protects a computer device, see Figure 6 The computer device includes a memory 401, a processor 402, and a computer program 403 stored in the memory 401 and running on the processor 402, wherein when the processor 402 executes the computer program 403, the computer device can execute any one of the aforementioned methods for rapid reconstruction of cardiac CT images.

[0121] Embodiment 4:

[0122] This embodiment also provides a computer-readable storage medium, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement a method for rapid reconstruction of cardiac CT images provided in the above embodiment.

[0123] Embodiment 5:

[0124] This embodiment further provides a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the above-mentioned related steps to implement a method for rapid reconstruction of cardiac CT images provided in the above-mentioned embodiment.

[0125] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment is used to execute the corresponding method provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method provided above, and will not be repeated here.

[0126] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0127] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for rapid reconstruction of cardiac CT images, characterized in that: The method comprises the following steps: Superimposing cardiac CT back-projection images at different angles to obtain an initial cardiac CT image; According to the gray value and gray superposition of each pixel in the initial cardiac CT image, as well as the gray change between two adjacent superpositions, the cardiac possibility degree of each pixel is obtained; Screen out suspected heart pixels based on the possible degree of the heart, and obtain the close combination degree of each suspected heart pixel according to the distance between each suspected heart pixel and the centroid of the heart and the gray value difference between each suspected heart pixel and its preset neighboring pixels; The signal strength of each suspected heart pixel is obtained according to the distribution of grayscale values ​​in a preset window of each suspected heart pixel and the fluctuation of grayscale values ​​in the background area; According to the close combination degree, signal strength and cardiac possibility degree of each suspected cardiac pixel point, a final cardiac possibility degree of each suspected cardiac pixel point is obtained; Determine cardiac pixels based on the final cardiac probability and the cardiac probability, and reconstruct the cardiac CT image; The method for obtaining the possible degree of the heart is: According to the grayscale superposition of each pixel in the initial cardiac CT image and the grayscale change of two adjacent superpositions, the regular growth degree of each pixel is obtained; The result of normalizing the product of the gray value of each pixel and the regular growth degree is taken as the possible degree of the heart of each pixel; The method for obtaining the regular growth degree is: For any pixel point in the initial cardiac CT image, the number of times the grayscale accumulated value of the pixel point increases during the superposition process is obtained as the number of grayscale increases of the pixel point; The ratio of the grayscale growth times to the superposition times is used as a first regular growth reference value of the pixel point; Obtain the ratio of the grayscale accumulated value of the pixel after each superposition to the grayscale accumulated value after the previous superposition as the grayscale change value of the corresponding superposition; The average of all grayscale change values ​​is used as the second law growth reference value of the pixel point; The product of the first regular growth reference value and the second regular growth reference value is taken as the regular growth degree of the pixel point.

2. A method for rapid reconstruction of cardiac CT images as claimed in claim 1, characterized in that: The method for obtaining the suspected heart pixel point is: When the heart possibility level is less than or equal to a preset heart possibility level threshold, the corresponding pixel point is regarded as a suspected heart pixel point.

3. The method for rapid reconstruction of cardiac CT images according to claim 1, characterized in that: The method for obtaining the heart mass center is: Normalizing the grayscale value of each pixel in the initial cardiac CT image, and when the normalized grayscale value is greater than a first preset grayscale threshold, using the corresponding pixel as a cardiac reference pixel; The area formed by the heart reference pixel points is taken as the heart range area, and the centroid of the heart range area is obtained as the heart centroid.

4. The method for rapid reconstruction of cardiac CT images according to claim 1, characterized in that: The method for obtaining the degree of tight binding is: For any suspected heart pixel point, obtain the Euclidean distance between the suspected heart pixel point and the heart centroid as the first distance; The gradient direction of the suspected heart pixel is obtained by using a Sobel operator, and a preset number of pixels closest to the suspected heart pixel on the straight line where the gradient direction is located are used as preset neighborhood pixels of the suspected heart pixel; Obtaining the grayscale value difference between the suspected heart pixel and each of its preset neighborhood pixels, all as the first difference; The mean value of the first difference is used as the local contrast of the suspected heart pixel; The close combination degree of the suspected heart pixel point is obtained according to the first distance and the local contrast; wherein the first distance and the close combination degree are negatively correlated, and the local contrast and the close combination degree are positively correlated.

5. The method for rapid reconstruction of cardiac CT images according to claim 1, characterized in that: The method for obtaining the signal strength is: For any suspected heart pixel, obtaining the mean grayscale value of the suspected heart pixel in a preset window as the reference grayscale value of the suspected heart pixel; The pixels in the initial cardiac CT image whose grayscale values ​​are less than a second preset grayscale threshold are all used as background reference pixels; The signal strength of the suspected heart pixel is obtained according to the standard deviation of the reference grayscale value and the grayscale value of the background reference pixel; the reference grayscale value is positively correlated with the signal strength, and the standard deviation of the grayscale value of the background reference pixel is negatively correlated with the signal strength.

6. The method for rapid reconstruction of cardiac CT images according to claim 1, characterized in that: The method for obtaining the final cardiac probability is: For any suspected heart pixel, the product of the close combination degree of the suspected heart pixel and the signal strength is normalized as the correction weight of the suspected heart pixel; The product of the correction weight of the suspected heart pixel and the heart possibility degree is used as the final heart possibility degree of the suspected heart pixel.

7. The method for rapid reconstruction of cardiac CT images according to claim 1, characterized in that: The method for obtaining the heart pixel point is: When the heart probability level is greater than a preset heart probability level threshold, the corresponding pixel point is regarded as a heart pixel point; When the final heart possibility degree is greater than the preset final heart possibility degree threshold, the corresponding suspected heart pixel point is taken as the heart pixel point.

8. A rapid reconstruction system for cardiac CT images, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When executing the computer program, the processor implements the steps of a method for rapid reconstruction of cardiac CT images as described in any one of claims 1 to 7.

Images