A method and system for clear imaging of CT angiography of diabetic foot patients

Abstract

The present invention relates to the field of image processing technology, and proposes a method and system for clear imaging of CT angiography of diabetic foot patients, including: collecting CT angiography images and converting them into CT angiography grayscale images, dividing suspected blood vessel edges; marking target suspected blood vessel edges, obtaining grayscale similarity weights of target suspected blood vessel edges; obtaining extended feature similarity, obtaining interference object confidences of two different suspected blood vessel edges, and screening all blood vessel pixels; interpolating the CT angiography grayscale image according to the position distribution of all blood vessel pixels in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image, and obtaining clear imaging results of CT angiography of diabetic foot patients. The present invention solves the problem that interference objects such as bones are mixed with blood vessels, reducing the image enhancement efficiency of CT angiography images.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and system for clear imaging of CT angiography of diabetic foot patients. Background Art

[0002] When examining the degree of vascular lesions in a patient's lower limbs, CT angiography is generally used in combination with CT scanning and contrast agents to obtain CT angiography images of the patient, and the CT angiography images clearly show the vascular details of the patient's lower limbs. However, the blood vessels in the patient's lower limbs are relatively small, and the blood in the blood vessels keeps flowing during the CT angiography process, so CT angiography images often have artifacts and blur. In order to improve the clarity of CT angiography images and ensure the accuracy of vascular detail observation, it is necessary to enhance the acquired CT angiography images.

[0003] Image enhancement algorithms such as histogram equalization and local contrast enhancement are generally used to improve the clarity of CT angiography images. However, in CT angiography images, blood vessels are often mixed with the patient's bones. The mixed bone positions do not require fine image processing, and fine calculations at the bone positions often lead to reduced image enhancement efficiency. Summary of the invention

[0004] The present invention provides a method and system for clear imaging of CT angiography of diabetic foot patients, so as to solve the problem that when enhancing CT angiography images, interference objects such as bones are mixed with blood vessels, thereby reducing the image processing efficiency. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for clear imaging of CT angiography of a diabetic foot patient, the method comprising the following steps:

[0006] Acquiring CT angiography images and converting them into CT angiography grayscale images, and dividing suspected blood vessel edges of suspected blood vessel areas in the CT angiography grayscale images according to edge information;

[0007] Any suspected blood vessel edge is recorded as a target suspected blood vessel edge, and any pixel point included in the target suspected blood vessel edge is recorded as a target pixel point. According to the position distribution of the pixel point in the local range of the target pixel point and the target suspected blood vessel edge, and the gray value difference between the pixel point in the local range of the target pixel point and the pixel point included in the target suspected blood vessel edge, the gray similarity weight of the target suspected blood vessel edge is obtained;

[0008] According to the difference in the extension directions of two different suspected blood vessel edges and the difference in the grayscale values ​​of the pixels included, the similarity of the extension features of the two different suspected blood vessel edges is obtained, and the interference object confidences of the two different suspected blood vessel edges are obtained by combining the grayscale similarity weights of the two different suspected blood vessel edges, and all blood vessel pixels in the CT angiography grayscale image are screened out according to the interference object confidences;

[0009] According to the position distribution of all blood vessel pixels in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image, the CT angiography grayscale image is interpolated to obtain clear CT angiography imaging results of diabetic foot patients.

[0010] Further, the grayscale similarity weight of the target suspected blood vessel edge is obtained according to the position distribution of the pixel points within the local range of the target pixel point and the target suspected blood vessel edge, and the grayscale value difference between the pixel points within the local range of the target pixel point and the pixel points included in the target suspected blood vessel edge, including the specific method of:

[0011] The mean of the grayscale values ​​of all pixels contained in the target suspected blood vessel edge is recorded as the characteristic grayscale mean of the target suspected blood vessel edge;

[0012] A local grayscale window with a side length of a first preset threshold value is established with the target pixel point as the center, and the grayscale value means of the pixels contained in the two areas divided by the local grayscale window by the suspected blood vessel edge are calculated respectively, and the grayscale value mean with the smallest difference from the characteristic grayscale mean value of the target suspected blood vessel edge is recorded as the first grayscale characteristic value of the target pixel point, and the value of the first preset threshold value is 3;

[0013] The grayscale similarity weight of the target suspected blood vessel edge is obtained according to the difference between the characteristic grayscale mean value of the target suspected blood vessel edge and the first grayscale characteristic value of all pixels included in the target suspected blood vessel edge.

[0014] Furthermore, the grayscale similarity weight of the target suspected blood vessel edge is obtained according to the difference between the characteristic grayscale mean value of the target suspected blood vessel edge and the first grayscale characteristic value of all pixels included in the target suspected blood vessel edge, including the specific method of:

[0015] The average of the first grayscale eigenvalues ​​of all pixels included in the target suspected blood vessel edge is recorded as the local grayscale eigenvalue of the target suspected blood vessel edge;

[0016] The ratio between the characteristic grayscale mean value of the target suspected blood vessel edge and the local grayscale characteristic value is recorded as the first ratio of the target suspected blood vessel edge;

[0017] The absolute value of the difference between the number 1 and the first ratio of the target suspected blood vessel edge is recorded as the grayscale similarity weight of the target suspected blood vessel edge.

[0018] Furthermore, the method of obtaining the similarity of the extension features of the two different suspected blood vessel edges according to the difference in the extension directions of the two different suspected blood vessel edges and the difference in the grayscale values ​​of the pixels included therein includes:

[0019] The variance of the grayscale values ​​of all pixels included in the target suspected blood vessel edge is recorded as the characteristic grayscale difference of the target suspected blood vessel edge;

[0020] Fitting the coordinates of all pixel points included in the target suspected blood vessel edge to obtain a coordinate fitting line of the target suspected blood vessel edge, and recording the inclination angle of the coordinate fitting line as the angle characteristic value of the target suspected blood vessel edge;

[0021] According to the difference between the characteristic grayscale differences of two different suspected blood vessel edges and the difference between the angle characteristic values, the similarity of the extended characteristics of the two different suspected blood vessel edges is obtained.

[0022] Further, the method of obtaining the similarity of the extended features of two different suspected blood vessel edges according to the difference between the feature grayscale differences of the two different suspected blood vessel edges and the difference between the angle feature values ​​includes the following specific methods:

[0023] Any suspected blood vessel edge that is different from the target suspected blood vessel edge is recorded as a comparative suspected blood vessel edge;

[0024] The ratio between the characteristic grayscale difference of the target suspected blood vessel edge and the characteristic grayscale difference of the comparative suspected blood vessel edge is recorded as a second ratio between the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the target suspected blood vessel edge and the second ratio is recorded as a first absolute value between the target suspected blood vessel edge and the comparative suspected blood vessel edge;

[0025] The ratio between the angle characteristic value of the target suspected blood vessel edge and the angle characteristic value of the comparative suspected blood vessel edge is recorded as a third ratio between the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the third ratio is recorded as a second absolute value between the target suspected blood vessel edge and the comparative suspected blood vessel edge;

[0026] The sum of the first absolute value and the second absolute value is recorded as the extension feature similarity between the target suspected blood vessel edge and the comparison suspected blood vessel edge.

[0027] Furthermore, the specific method for obtaining the interfering object confidences of the two different suspected blood vessel edges is as follows:

[0028] The product of the mean of the grayscale similarity weights of two different suspected blood vessel edges and the extended feature similarity is recorded as the interference object confidence of the two different suspected blood vessel edges.

[0029] Furthermore, the method of screening out all blood vessel pixels in the CT angiography grayscale image according to the interference object confidence level includes the following specific methods:

[0030] When the interfering object confidence of two suspected blood vessel edges is less than or equal to a first judgment threshold, the pixels included in the two suspected blood vessel edges are marked as blood vessel pixels, and the value of the first judgment threshold is 0.1.

[0031] Further, the interpolation processing is performed on the CT angiography grayscale image according to the position distribution of all blood vessel pixels in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image to obtain a clear imaging result of CT angiography of diabetic foot patients, including the specific method of:

[0032] An arbitrary image column in the CT angiography grayscale image is recorded as a target image column, an image column adjacent to the left of the target image column is recorded as a left adjacent image column of the target image column, an image column adjacent to the right of the target image column is recorded as a right adjacent image column of the target image column, the number of blood vessel pixels included in the target image column and the number of all pixels included in the target image column are recorded as a blood vessel ratio of the target image column, and the cumulative sum of the blood vessel ratios of the target image column, the left adjacent image column and the right adjacent image column is recorded as the interpolation necessity of the target image column;

[0033] When the necessity of interpolation of the target image column is greater than the second judgment threshold, it is determined that the target image column needs to be interpolated, and the target image column that needs to be interpolated is interpolated to obtain clear imaging results of CT angiography of diabetic foot patients. The value of the second judgment threshold is 0.5.

[0034] Furthermore, the interpolation processing is performed on the image sequence that needs to be interpolated to obtain clear imaging results of CT angiography of diabetic foot patients, including the specific method of:

[0035] For a target image column that needs to be interpolated, an image column corresponding to the maximum value of the blood vessel ratio between the left adjacent image column and the right adjacent image column of the target image column is recorded as an interpolation reference column of the target image column;

[0036] For an image column that needs to be interpolated, the column vector obtained by arranging the pixel values ​​of all pixels included in the target image column in order from top to bottom is recorded as the first vector of the target image column, and the column vector obtained by arranging the pixel values ​​of all pixels included in the interpolation reference column of the target image column in order from top to bottom is recorded as the second vector of the target image column; the cosine similarity between the first vector and the second vector of the target image column is recorded as the first coefficient of the target image column, the product of the square of the modulus of the first vector of the target image column and the first coefficient of the target image column is recorded as the second coefficient of the target image column, and the ratio of the second coefficient of the target image column to the first vector of the target image column is recorded as the third vector of the target image column;

[0037] The rounded values ​​of the numerical values ​​contained in the third vector of the target image column are arranged in order from top to bottom, and inserted as an image column between the target image column and the next column of the target image column, so as to complete the interpolation processing of the target image column;

[0038] Interpolation processing is performed on all image columns that need to be interpolated, and the interpolation processing result is a clear imaging result of CT angiography of diabetic foot patients.

[0039] In a second aspect, an embodiment of the present invention further provides a clear imaging system for CT angiography of diabetic foot patients, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods when executing the computer program.

[0040] The beneficial effects of the present invention are:

[0041] The present application improves the accuracy of blood vessel recognition in CT angiography images based on the characteristics that venous blood flows slowly and the venous vessels are thin, resulting in significantly poor display effects of veins, and the characteristics that veins and bones have similar display effects on contrast agents in CT angiography images, performs image enhancement on the areas corresponding to the blood vessels, and improves the processing efficiency and accuracy of image enhancement. Specifically, first, based on the characteristics of uniform and consistent grayscale value distribution at the blood vessel position, the grayscale similarity weight of the suspected blood vessel edge is determined, and then, based on the characteristics that the extension directions of the blood vessel walls on both sides of the blood vessel are consistent, the extension feature similarities of two different suspected blood vessel edges are obtained, and the two different The grayscale similarity weights of the same suspected blood vessel edges are used to obtain the interference object confidences of two different suspected blood vessel edges. The interference object confidence is the confidence evaluation of the suspected blood vessel edge corresponding to non-vascular objects such as bones. All blood vessel pixels in the CT angiography grayscale image are screened out according to the interference object confidence. Finally, according to the position distribution of the screened blood vessel pixels and the pixel values ​​of the pixels of all image columns in the CT angiography grayscale image, the CT angiography grayscale image is interpolated to obtain clear imaging results of CT angiography for diabetic foot patients, so as to solve the problem that interference objects such as bones are mixed with blood vessels, which reduces the image enhancement efficiency of CT angiography images. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be 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 labor.

[0043] Figure 1 A schematic flow chart of a method for clear imaging of CT angiography of diabetic foot patients provided by one embodiment of the present invention;

[0044] Figure 2 A flowchart for obtaining grayscale similarity weights provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1, which shows a flow chart of a method for clear imaging of CT angiography of diabetic foot patients provided by an embodiment of the present invention, the method comprising the following steps:

[0047] Step S001 : acquiring a CT angiography image and converting it into a CT angiography grayscale image, and dividing suspected blood vessel edges of suspected blood vessel areas in the CT angiography grayscale image according to edge information.

[0048] Under the premise of ensuring that the patient remains still, CT angiography technology is used to obtain the patient's CT angiography image.

[0049] The CT angiography image is converted into a grayscale image to obtain the CT angiography grayscale image, and the CT angiography grayscale image is subjected to denoising and threshold segmentation to divide the suspected blood vessel area in the CT angiography grayscale image.

[0050] Preferably, in one embodiment of the present application, a median filtering algorithm is used to denoise the CT angiography grayscale image, and the maximum inter-class variance method is used for threshold segmentation. In actual application, as other implementation methods, the implementer can adopt denoising algorithms such as wavelet denoising, and threshold segmentation algorithms such as entropy algorithm and Niblack threshold method according to actual conditions. This application does not impose any special restrictions.

[0051] The canny edge detection algorithm is used to detect the edge of the suspected blood vessel area to obtain the edge of the suspected blood vessel. The suspected blood vessel edge corresponds to the edge of objects such as blood vessels and bones in the CT angiography image.

[0052] At this point, the suspected blood vessel edge of the suspected blood vessel area in the CT angiography grayscale image is obtained.

[0053] Step S002, any suspected blood vessel edge is recorded as a target suspected blood vessel edge, and any pixel point included in the target suspected blood vessel edge is recorded as a target pixel point, and the grayscale similarity weight of the target suspected blood vessel edge is obtained according to the position distribution of the pixel points in the local range of the target pixel point and the target suspected blood vessel edge, and the grayscale value difference between the pixel points in the local range of the target pixel point and the pixel points included in the target suspected blood vessel edge.

[0054] Since venous blood flows slowly and venous blood vessels are thin, arteries and veins have different display effects on contrast agents in CT angiography images, and the display effect of veins is significantly poor. At the same time, the display effect of veins on contrast agents in CT angiography images is similar to that of bones, and veins and bones with similar display effects are mixed together. Therefore, when enhancing CT angiography images, if the CT angiography images are enhanced as a whole, the processing efficiency is too low. If only the blood vessels in the CT angiography images are enhanced, the processing efficiency can be improved, but veins and bones with similar display effects are not easy to distinguish, and bones are easily misjudged as veins, resulting in low processing efficiency, and veins are easily misjudged as bones, resulting in poor image enhancement effects. Therefore, it is necessary to accurately divide all blood vessels in the CT angiography images and then perform image enhancement on the blood vessels.

[0055] In order to accurately divide all blood vessels in the CT angiography image, each suspected blood vessel edge is analyzed separately. Any suspected blood vessel edge is taken as a target suspected blood vessel edge, and any pixel point included in the target suspected blood vessel edge is recorded as a target pixel point. This embodiment takes any pixel point included in any suspected blood vessel edge as an example for description.

[0056] The mean of the grayscale values ​​of all pixels included in the target suspected blood vessel edge is recorded as the characteristic grayscale mean of the target suspected blood vessel edge.

[0057] A local grayscale window with a side length of the first preset threshold is established with the target pixel as the center. The suspected blood vessel edge divides the local grayscale window of the target pixel into two areas, and the grayscale value means of the pixels contained in the two areas are calculated respectively. The grayscale value mean that is closest to the characteristic grayscale mean of the target suspected blood vessel edge is recorded as the first grayscale characteristic value of the target pixel.

[0058] It is understandable that when the suspected blood vessel edge does not divide the local grayscale window of the target pixel into two areas, the average grayscale value of all pixels included in the local grayscale window of the target pixel is recorded as the first grayscale feature value of the target pixel.

[0059] In this embodiment, the value of the first preset threshold is 3. In actual application, as other implementation methods, the implementer can decide the value of the first preset threshold according to the actual situation, and this application does not impose any special restrictions.

[0060] The same method may be used to obtain the first grayscale characteristic value of any pixel point included in the edge of the target suspected blood vessel.

[0061] The grayscale similarity weight of the target suspected blood vessel edge is obtained according to the difference between the characteristic grayscale mean value of the target suspected blood vessel edge and the first grayscale characteristic value of all pixels included in the target suspected blood vessel edge.

[0062] Preferably, as an embodiment of the present application, the mean of the first grayscale eigenvalues ​​of all pixels contained in the target suspected blood vessel edge is recorded as the local grayscale eigenvalue of the target suspected blood vessel edge, the ratio between the characteristic grayscale mean and the local grayscale eigenvalue of the target suspected blood vessel edge is recorded as the first ratio of the target suspected blood vessel edge, and the absolute value of the difference between the number 1 and the first ratio of the target suspected blood vessel edge is recorded as the grayscale similarity weight of the target suspected blood vessel edge.

[0063] Specifically, the calculation formula for the grayscale similarity weight of any suspected blood vessel edge is:

[0064]

[0065]

[0066] In the formula, Indicates Gray similarity weight of suspected blood vessel edges; Indicates The characteristic grayscale mean of the suspected blood vessel edge; Indicates The local grayscale feature value of the suspected blood vessel edge; Indicates The number of all pixels contained in the edge of a suspected blood vessel; Indicates The suspected blood vessel edge contains the The first grayscale eigenvalue of a pixel.

[0067] in, Indicates The first ratio of the suspected blood vessel edge.

[0068] When the first grayscale eigenvalue of the pixel points contained in the suspected blood vessel edge is closer to the characteristic grayscale mean, the local characteristics of the pixel points contained in the suspected blood vessel edge are closer to the overall grayscale distribution of the suspected blood vessel edge. At this time, the closer the local grayscale eigenvalue of the suspected blood vessel edge is to the characteristic grayscale mean of the suspected blood vessel edge, the smaller the grayscale similarity weight of the suspected blood vessel edge is, and the greater the possibility that the object corresponding to the surrounding of the suspected blood vessel edge and the object corresponding to the suspected blood vessel region where the suspected blood vessel edge is located are the same object.

[0069] The same method can be used to obtain the grayscale similarity weight of any suspected blood vessel edge. The grayscale similarity weight acquisition flow chart is as follows: Figure 2 shown.

[0070] At this point, the grayscale similarity weights of all suspected blood vessel edges of all suspected blood vessel regions in the CT angiography grayscale image are obtained.

[0071] Step S003, based on the difference in extension directions of two different suspected blood vessel edges and the difference in grayscale values ​​of the pixels included, the similarity of extension features of the two different suspected blood vessel edges is obtained, and the interfering object confidence of the two different suspected blood vessel edges is obtained in combination with the grayscale similarity weights of the two different suspected blood vessel edges, and all blood vessel pixels in the CT angiography grayscale image are screened out according to the interfering object confidence.

[0072] The variance of the grayscale values ​​of all pixels included in the target suspected blood vessel edge is recorded as the characteristic grayscale difference of the target suspected blood vessel edge.

[0073] The leftmost and lowermost pixel in the CT angiography grayscale image is taken as the origin. In the CT angiography grayscale image, the horizontal right direction is taken as the positive direction of the x-axis, and the vertical upward direction is taken as the positive direction of the y-axis. A plane rectangular coordinate system is established, and the coordinates of all pixel points contained in the target suspected blood vessel edge are fitted with a straight line to obtain the coordinate fitting line of the target suspected blood vessel edge. The inclination angle of the coordinate fitting line of the target suspected blood vessel edge is recorded as the angle characteristic value of the target suspected blood vessel edge.

[0074] The inclination angle of a straight line is a well-known definition of the extension direction of a straight line in the mathematical field and will not be described in detail.

[0075] The same method can be used to obtain the characteristic grayscale difference and angle characteristic value of any suspected blood vessel edge.

[0076] According to the difference between the characteristic grayscale differences of two different suspected blood vessel edges and the difference between the angle characteristic values, the similarity of the extended characteristics of the two different suspected blood vessel edges is obtained.

[0077] Any suspected blood vessel edge different from the target suspected blood vessel edge is recorded as a comparative suspected blood vessel edge.

[0078] Preferably, as an embodiment of the present application, the ratio between the characteristic grayscale difference of the target suspected blood vessel edge and the characteristic grayscale difference of the comparative suspected blood vessel edge is recorded as the second ratio of the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the second ratio is recorded as the first absolute value of the target suspected blood vessel edge and the comparative suspected blood vessel edge. The ratio between the angle characteristic value of the target suspected blood vessel edge and the angle characteristic value of the comparative suspected blood vessel edge is recorded as the third ratio of the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the third ratio is recorded as the second absolute value of the target suspected blood vessel edge and the comparative suspected blood vessel edge. The sum of the first absolute value and the second absolute value is recorded as the extended characteristic similarity between the target suspected blood vessel edge and the comparative suspected blood vessel edge.

[0079] In the process of calculating the ratio, in order to avoid the situation where the denominator is zero, a preset value needs to be added to the denominator, and an embodiment of the preset value is 0.01.

[0080] Specifically, the calculation formula for the similarity of the extension features of two different suspected blood vessel edges is:

[0081]

[0082] In the formula, Indicates The edge of the suspected blood vessel Similarity of extension features of suspected blood vessel edges; Indicates The angle characteristic value of the suspected blood vessel edge; Indicates The angle characteristic value of the suspected blood vessel edge; Indicates The characteristic grayscale difference of the suspected blood vessel edge; Indicates The characteristic grayscale difference of the suspected blood vessel edge; represents the first preset parameter. In this embodiment, the value of the first preset parameter is 0.01.

[0083] in, Indicates The edge of the suspected blood vessel The second ratio of the suspected blood vessel edge, Indicates The edge of the suspected blood vessel The first absolute value of the suspected blood vessel edge, Indicates The edge of the suspected blood vessel The third ratio of the suspected blood vessel edge, Indicates The edge of the suspected blood vessel The second absolute value of the suspected blood vessel edge.

[0084] When the difference between the angle feature values ​​of two suspected blood vessel edges is smaller, the extension direction and contour features of the two suspected blood vessel edges are more consistent, and the possibility that the two suspected blood vessel edges correspond to the vascular walls on both sides of the same blood vessel is greater. When the difference in the characteristic grayscale of the two suspected blood vessel edges is smaller, the fluctuation of the grayscale values ​​of all pixels contained in the two suspected blood vessel edges is more similar, and the possibility that the two suspected blood vessel edges correspond to the vascular walls on both sides of the same blood vessel is greater. When the difference between the angle feature values ​​of the two suspected blood vessel edges is smaller and the difference in the characteristic grayscale is smaller, the possibility that the two suspected blood vessel edges correspond to the vascular walls on both sides of the same blood vessel is greater, and at this time, the similarity of the extension features of the two suspected blood vessel edges is smaller.

[0085] The same method can be used to obtain the similarity of the extension features of any two different suspected blood vessel edges.

[0086] According to the extension feature similarity and grayscale similarity weights of two different suspected blood vessel edges, the interference object confidences of the two different suspected blood vessel edges are obtained.

[0087] Preferably, as an embodiment of the present application, the mean of the grayscale similarity weights of two different suspected blood vessel edges is recorded as the first mean of the two different suspected blood vessel edges, and the product of the first mean of the two different suspected blood vessel edges and the extended feature similarity is recorded as the interfering object confidence of the two different suspected blood vessel edges.

[0088] When the extended feature similarity and grayscale similarity weights of the two suspected blood vessel edges are smaller, the possibility that the two suspected blood vessel edges correspond to the two sides of the vascular walls of the same blood vessel is greater. At this time, the interfering object confidence of the two suspected blood vessel edges is smaller.

[0089] A first judgment threshold is set, and when the interference object confidence of two suspected blood vessel edges is less than or equal to the first judgment threshold, it is determined that the two suspected blood vessel edges correspond to the blood vessel walls at both ends of the same blood vessel. In this embodiment, the value of the first judgment threshold is 0.1.

[0090] In particular, it can be understood that when the interference object confidences of a certain suspected blood vessel edge and multiple other suspected blood vessel edges are all less than or equal to the first judgment threshold, the two suspected blood vessel edges with the smallest interference object confidences are determined to correspond to the two end blood vessel walls of the same blood vessel.

[0091] All pixels included in the suspected blood vessel edges of the blood vessel walls at both ends of the same blood vessel when two suspected blood vessel edges are determined to correspond are marked as blood vessel pixels.

[0092] At this point, all blood vessel pixels in the CT angiography grayscale image are obtained.

[0093] Step S004, interpolating the CT angiography grayscale image according to the position distribution of all blood vessel pixels in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image, to obtain clear CT angiography imaging results of diabetic foot patients.

[0094] In order to improve the distinction between veins and bones with similar display effects and improve the efficiency of image enhancement for CT angiography images, the image columns that need to be interpolated in the CT angiography grayscale image are determined according to the position distribution of blood vessel pixels.

[0095] It can be understood that the image column is a pixel column composed of a column of pixel points in the CT angiography grayscale image.

[0096] Any image column in the CT angiography grayscale image is recorded as the target image column, the image column adjacent to the left of the target image column is recorded as the left adjacent image column of the target image column, and the image column adjacent to the right of the target image column is recorded as the right adjacent image column of the target image column.

[0097] Preferably, as an embodiment of the present application, the number of blood vessel pixel points contained in the target image column and the number of all pixel points contained in the target image column are recorded as the blood vessel ratio of the target image column, and the cumulative sum of the blood vessel ratios of the target image column, the left adjacent image column and the right adjacent image column is recorded as the interpolation necessity of the target image column.

[0098] It can be understood that the leftmost image column in the CT angiography grayscale image has no left adjacent image column, and the rightmost image column in the CT angiography grayscale image has no right adjacent image column, and the vascular ratio of the left adjacent image column of the leftmost image column and the vascular ratio of the right adjacent image column of the rightmost image column in the CT angiography grayscale image are both assigned a value of 0.

[0099] A second judgment threshold is set. When the interpolation necessity of the target image sequence is greater than the second judgment threshold, it is determined that the target image sequence needs to be interpolated; when the interpolation necessity of the target image sequence is less than or equal to the second judgment threshold, it is determined that the target image sequence does not need to be interpolated. In this embodiment, the value of the second judgment threshold is 0.5.

[0100] The same method can be used to obtain a determination result of whether any image column in the CT angiography grayscale image needs to be interpolated.

[0101] Furthermore, an interpolation direction is determined for the image sequence that needs to be interpolated.

[0102] Compare the vascular ratios of the left adjacent image column and the right adjacent image column of the target image column, record the direction of the image column with the largest vascular ratio relative to the target image column as the interpolation direction of the target image column that needs to be interpolated, and record the image column with the largest vascular ratio as the interpolation reference column of the target image column.

[0103] In particular, when the vascular ratio of the left adjacent image column of the target image column is equal to the vascular ratio of the right adjacent image column, the left side of the target image column is used as the interpolation direction of the target image column, and the left adjacent image column of the target image column is recorded as the interpolation reference column of the target image column.

[0104] It can be understood that, when the vascular ratio of the left adjacent image column of the target image column is greater than or equal to the vascular ratio of the right adjacent image column, the left side of the target image column is used as the interpolation direction of the target image column, and the left adjacent image column of the target image column is recorded as the interpolation reference column of the target image column; when the vascular ratio of the left adjacent image column of the target image column is less than the vascular ratio of the right adjacent image column, the right side of the target image column is used as the interpolation direction of the target image column, and the right adjacent image column of the target image column is recorded as the interpolation reference column of the target image column.

[0105] For the image column that needs to be interpolated, the column vector obtained by arranging the pixel values ​​of all the pixels in the target image column from top to bottom is recorded as the first vector of the target image column, and the column vector obtained by arranging the pixel values ​​of all the pixels in the interpolation reference column of the target image column from top to bottom is recorded as the second vector of the target image column. The cosine similarity between the first vector and the second vector of the target image column is recorded as the first coefficient of the target image column, the product of the square of the modulus of the first vector of the target image column and the first coefficient of the target image column is recorded as the second coefficient of the target image column, and the ratio of the second coefficient of the target image column to the first vector of the target image column is recorded as the third vector of the target image column.

[0106] The rounded values ​​of the numerical values ​​contained in the third vector of the target image column are arranged in sequence from top to bottom, and inserted as an image column between the target image column and the next column of the target image column.

[0107] The same method can be used to perform interpolation processing on all image columns that need to be interpolated, thereby implementing interpolation processing on the CT angiography grayscale image and completing image enhancement on the CT angiography image.

[0108] At this point, clear imaging results of CT angiography of patients with diabetic foot were obtained.

[0109] Based on the same inventive concept as the above method, an embodiment of the present invention also provides a CT angiography clear imaging system for patients with diabetic foot, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of any one of the above-mentioned methods for clear CT angiography imaging of patients with diabetic foot are implemented.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for clear imaging of CT angiography of diabetic foot patients, characterized in that: The method comprises the following steps: Acquiring CT angiography images and converting them into CT angiography grayscale images, and dividing suspected blood vessel edges of suspected blood vessel areas in the CT angiography grayscale images according to edge information; Any suspected blood vessel edge is recorded as a target suspected blood vessel edge, and any pixel point included in the target suspected blood vessel edge is recorded as a target pixel point. According to the position distribution of the pixel point in the local range of the target pixel point and the target suspected blood vessel edge, and the gray value difference between the pixel point in the local range of the target pixel point and the pixel point included in the target suspected blood vessel edge, the gray similarity weight of the target suspected blood vessel edge is obtained; According to the difference in the extension directions of two different suspected blood vessel edges and the difference in the grayscale values ​​of the pixels included, the similarity of the extension features of the two different suspected blood vessel edges is obtained, and the interference object confidences of the two different suspected blood vessel edges are obtained by combining the grayscale similarity weights of the two different suspected blood vessel edges, and all blood vessel pixels in the CT angiography grayscale image are screened out according to the interference object confidences; According to the position distribution of all blood vessel pixel points in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image, the CT angiography grayscale image is interpolated to obtain clear imaging results of CT angiography of diabetic foot patients; The grayscale similarity weight of the target suspected blood vessel edge is obtained according to the position distribution of the pixel points in the local range of the target pixel points and the target suspected blood vessel edge, and the grayscale value difference between the pixel points in the local range of the target pixel points and the pixel points included in the target suspected blood vessel edge, including the specific method of: The mean of the grayscale values ​​of all pixels contained in the target suspected blood vessel edge is recorded as the characteristic grayscale mean of the target suspected blood vessel edge; A local grayscale window with a side length of a first preset threshold value is established with the target pixel point as the center, and the grayscale value means of the pixels contained in the two areas divided by the local grayscale window by the suspected blood vessel edge are calculated respectively, and the grayscale value mean with the smallest difference from the characteristic grayscale mean value of the target suspected blood vessel edge is recorded as the first grayscale characteristic value of the target pixel point, and the value of the first preset threshold value is 3; Obtaining a grayscale similarity weight of the target suspected blood vessel edge according to a difference between a characteristic grayscale mean value of the target suspected blood vessel edge and a first grayscale characteristic value of all pixels included in the target suspected blood vessel edge; The specific method of obtaining the similarity of the extension features of two different suspected blood vessel edges according to the difference in the extension directions of the two different suspected blood vessel edges and the difference in the grayscale values ​​of the pixels included is as follows: The variance of the grayscale values ​​of all pixels included in the target suspected blood vessel edge is recorded as the characteristic grayscale difference of the target suspected blood vessel edge; Fitting the coordinates of all pixel points included in the target suspected blood vessel edge to obtain a coordinate fitting line of the target suspected blood vessel edge, and recording the inclination angle of the coordinate fitting line as the angle characteristic value of the target suspected blood vessel edge; According to the difference between the characteristic grayscale differences of two different suspected blood vessel edges and the difference between the angle characteristic values, the similarity of the extended characteristics of the two different suspected blood vessel edges is obtained.

2. The method for clear CT angiography imaging of diabetic foot patients according to claim 1, characterized in that: The specific method of obtaining the grayscale similarity weight of the target suspected blood vessel edge according to the difference between the characteristic grayscale mean value of the target suspected blood vessel edge and the first grayscale characteristic value of all pixels included in the target suspected blood vessel edge is as follows: The average of the first grayscale eigenvalues ​​of all pixels included in the target suspected blood vessel edge is recorded as the local grayscale eigenvalue of the target suspected blood vessel edge; The ratio between the characteristic grayscale mean value of the target suspected blood vessel edge and the local grayscale characteristic value is recorded as the first ratio of the target suspected blood vessel edge; The absolute value of the difference between the number 1 and the first ratio of the target suspected blood vessel edge is recorded as the grayscale similarity weight of the target suspected blood vessel edge.

3. The method for clear CT angiography imaging of diabetic foot patients according to claim 1, characterized in that: The specific method of obtaining the similarity of the extended features of two different suspected blood vessel edges according to the difference between the feature grayscale differences of the two different suspected blood vessel edges and the difference between the angle feature values ​​is as follows: Any suspected blood vessel edge that is different from the target suspected blood vessel edge is recorded as a comparative suspected blood vessel edge; The ratio between the characteristic grayscale difference of the target suspected blood vessel edge and the characteristic grayscale difference of the comparative suspected blood vessel edge is recorded as a second ratio between the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the target suspected blood vessel edge and the second ratio is recorded as a first absolute value between the target suspected blood vessel edge and the comparative suspected blood vessel edge; The ratio between the angle characteristic value of the target suspected blood vessel edge and the angle characteristic value of the comparative suspected blood vessel edge is recorded as a third ratio between the target suspected blood vessel edge and the comparative suspected blood vessel edge, and the absolute value of the difference between the number 1 and the third ratio is recorded as a second absolute value between the target suspected blood vessel edge and the comparative suspected blood vessel edge; The sum of the first absolute value and the second absolute value is recorded as the extension feature similarity between the target suspected blood vessel edge and the comparison suspected blood vessel edge.

4. The method for clear CT angiography imaging of diabetic foot patients according to claim 1, characterized in that: The specific method for obtaining the interfering object confidences of the two different suspected blood vessel edges is as follows: The product of the mean of the grayscale similarity weights of two different suspected blood vessel edges and the extended feature similarity is recorded as the interference object confidence of the two different suspected blood vessel edges.

5. The method for clear CT angiography imaging of diabetic foot patients according to claim 1, characterized in that: The specific method of screening out all blood vessel pixels in the CT angiography grayscale image according to the confidence of the interfering object includes: When the interfering object confidence of two suspected blood vessel edges is less than or equal to a first judgment threshold, the pixels included in the two suspected blood vessel edges are marked as blood vessel pixels, and the value of the first judgment threshold is 0.

1.

6. The method for clear CT angiography imaging of diabetic foot patients according to claim 1, characterized in that: The method of interpolating the CT angiography grayscale image according to the position distribution of all blood vessel pixels in the CT angiography grayscale image and the pixel values ​​of all image columns in the CT angiography grayscale image to obtain a clear imaging result of CT angiography of a diabetic foot patient includes the following specific methods: An arbitrary image column in the CT angiography grayscale image is recorded as a target image column, an image column adjacent to the left of the target image column is recorded as a left adjacent image column of the target image column, an image column adjacent to the right of the target image column is recorded as a right adjacent image column of the target image column, the number of blood vessel pixels included in the target image column and the number of all pixels included in the target image column are recorded as a blood vessel ratio of the target image column, and the cumulative sum of the blood vessel ratios of the target image column, the left adjacent image column and the right adjacent image column is recorded as the interpolation necessity of the target image column; When the necessity of interpolation of the target image column is greater than the second judgment threshold, it is determined that the target image column needs to be interpolated, and the target image column that needs to be interpolated is interpolated to obtain clear imaging results of CT angiography of diabetic foot patients. The value of the second judgment threshold is 0.

5.

7. The method for clear CT angiography imaging of diabetic foot patients according to claim 6, characterized in that: The specific method of performing interpolation processing on the image sequence that needs to be interpolated to obtain clear imaging results of CT angiography of diabetic foot patients includes: For a target image column that needs to be interpolated, an image column corresponding to the maximum value of the blood vessel ratio between the left adjacent image column and the right adjacent image column of the target image column is recorded as an interpolation reference column of the target image column; For an image column that needs to be interpolated, the column vector obtained by arranging the pixel values ​​of all pixels included in the target image column in order from top to bottom is recorded as the first vector of the target image column, and the column vector obtained by arranging the pixel values ​​of all pixels included in the interpolation reference column of the target image column in order from top to bottom is recorded as the second vector of the target image column; the cosine similarity between the first vector and the second vector of the target image column is recorded as the first coefficient of the target image column, the product of the square of the modulus of the first vector of the target image column and the first coefficient of the target image column is recorded as the second coefficient of the target image column, and the ratio of the second coefficient of the target image column to the first vector of the target image column is recorded as the third vector of the target image column; The rounded values ​​of the numerical values ​​contained in the third vector of the target image column are arranged in order from top to bottom, and inserted as an image column between the target image column and the next column of the target image column, so as to complete the interpolation processing of the target image column; Interpolation processing is performed on all image columns that need to be interpolated, and the interpolation processing result is a clear imaging result of CT angiography of diabetic foot patients.

8. A CT angiography clear imaging system for diabetic foot patients, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

Images