CT image artifact correction method, system and medical equipment

Abstract

The present invention relates to the technical field of image data processing, and in particular to a method, system and medical device for CT image artifact correction. A CT image is acquired, and a plurality of suspected arc segment sequences are determined according to the CT image; a chain code value set of each of the suspected arc segment sequences is determined, and a plurality of target suspected arc segment sequence pairs are extracted according to the chain code value set; a ring artifact region is determined according to the plurality of pairs of the target suspected arc segment sequence pairs; artifact correction is performed according to the ring artifact region; based on the variation characteristics of the chain code value on the ring artifact region, a chain code value set is introduced to finally determine the ring artifact region, and artifact correction is performed, so that the ring artifact region can be accurately located, thereby facilitating improving the correction effect and the quality of the CT image.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a CT image artifact correction method, system and medical equipment. Background Art

[0002] With the advancement of science and technology, CT imaging plays an increasingly important role in assisting doctors in analyzing the condition. CT imaging often produces ring artifacts due to various factors such as errors in the transmitter, thus affecting the image quality. There are many reasons for the occurrence of ring artifacts, such as detector generation, hardware failure, x-ray tube problems, sampling frequency and noise, etc. The appearance of these artifacts will reduce the image quality and interfere with the subsequent processing of the image. Therefore, it is necessary to correct the artifacts to improve the quality of CT images. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide a CT image artifact correction method, system and medical equipment, and the technical solutions adopted are as follows:

[0004] In a first aspect, an embodiment of the present application provides a CT image artifact correction method, comprising:

[0005] Acquire a CT image, and determine a plurality of suspected arc segment sequences based on the CT image;

[0006] Determine a chain code value set of each of the suspected arc sequence, and extract a plurality of target suspected arc sequence pairs according to the chain code value set;

[0007] Determine the annular artifact area according to a plurality of pairs of target suspected arc segment sequence pairs;

[0008] Artifact correction is performed according to the annular artifact area.

[0009] In one implementation, determining a plurality of suspected arc segment sequences according to the CT image includes:

[0010] Performing inverse Fourier transform on the CT image to obtain a transformed CT image, and performing edge detection on the transformed CT image to obtain an edge CT image;

[0011] Determining a grayscale histogram of the transformed CT image, and determining a plurality of target pixel points according to the grayscale histogram and a preset grayscale threshold;

[0012] Determine a plurality of suspected artifact regions according to the plurality of target pixels, determine a minimum bounding rectangle of each of the suspected artifact regions, and determine the possibility that each of the suspected artifact regions is an artifact region according to an aspect ratio of a length to a width of the minimum bounding rectangle and a pixel ratio of the target pixels within the minimum bounding rectangle;

[0013] Determine a candidate artifact region according to the possibility and the suspected artifact region;

[0014] A plurality of suspected arc segment sequences are determined according to the plurality of candidate artifact regions and the edge CT image.

[0015] In one embodiment, determining a plurality of suspected arc segment sequences according to the plurality of candidate artifact regions and the edge CT images comprises:

[0016] In the edge CT image, determining a pixel region corresponding to each of the candidate artifact regions;

[0017] In each of the pixel regions, first distances between two adjacent pixel points are calculated along a pixel arrangement direction, and a distance mean corresponding to each of the pixel regions is determined according to the first distances and the number of the first distances;

[0018] A measurement parameter is determined according to the distance mean, and the candidate artifact regions corresponding to the measurement parameter being greater than a measurement threshold are determined as suspected arc segment sequences, thereby obtaining a plurality of suspected arc segment sequences.

[0019] In one implementation, extracting a plurality of target suspected arc sequence pairs according to the chain code value set includes:

[0020] From the plurality of suspected arc segment sequences, along a preset direction, using a first suspected arc segment sequence, determine a first suspected arc segment sequence and determine a second suspected arc segment sequence other than the first suspected arc segment sequence;

[0021] Locate the first endpoint of the first suspected arc sequence and the second endpoint of the second suspected arc sequence along a preset direction, and determine a first chain code value of the first suspected arc sequence and a second chain code value of the second suspected arc sequence; wherein the first chain code value is the last chain code value in the chain code value set of the first suspected arc sequence along the preset direction, and the second chain code value is the first chain code value in the chain code value set of the second suspected arc sequence along the preset direction;

[0022] Connecting the first endpoint and the second endpoint to obtain a line segment, and determining a line segment chain code value of the line segment;

[0023] When the line segment chain code value is between the first chain code value and the second chain code value, first differences between two adjacent chain code values ​​in the chain code value set of the first suspected arc segment sequence are calculated respectively, and second differences between two adjacent chain code values ​​in the chain code value set of the second suspected arc segment sequence are calculated respectively;

[0024] Calculating a first difference mean according to each of the first differences and the number of the first differences, calculating a second difference mean according to each of the second differences and the number of the second differences, and determining a first mean of the first difference means and the second difference means;

[0025] Calculate a third difference between the line segment chain code value and the first difference mean, calculate a fourth difference between the line segment chain code value and the second difference mean, and determine a second mean according to the third difference and the fourth difference;

[0026] Determine, according to the first mean and the second mean, a probability value that the first suspected arc segment sequence and the second suspected arc segment sequence are the same suspected annular region;

[0027] Returning to the step of determining the second suspected arc segment sequence other than the first suspected arc segment sequence, until the first suspected arc segment sequence and each second suspected arc segment sequence are obtained to obtain a probability value corresponding to the first suspected arc segment sequence;

[0028] According to each of the possibility values, determining the target first suspected arc segment sequence and the target second suspected arc segment sequence with the greatest possibility as a pair of target suspected arc segment sequence pairs;

[0029] Determine a new first suspected arc segment sequence, and return to the step of determining a second suspected arc segment sequence other than the first suspected arc segment sequence according to the new first suspected arc segment sequence, until each of the suspected arc segment sequences is used as a first suspected arc segment sequence, and determine a plurality of target suspected arc segment sequence pairs using the first suspected arc segment sequence as a starting sequence;

[0030] Using the second suspected arc sequence, determine the first suspected arc sequence and determine the second suspected arc sequence other than the first suspected arc sequence, return to the step of locating the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence along the preset direction, until obtaining several pairs of target suspected arc sequence pairs with each suspected arc sequence as the starting sequence.

[0031] In one implementation, determining the most likely target first suspected arc segment sequence and target second suspected arc segment sequence according to each of the likelihood values ​​includes:

[0032] Determine the largest first likelihood value and the smallest likelihood value from among the likelihood values;

[0033] determining a fifth difference between the largest first likelihood magnitude value and the smallest likelihood magnitude value, and determining a sixth difference between each of the likelihood magnitude values ​​and the smallest likelihood magnitude value, respectively;

[0034] Calculating the ratio of each of the sixth difference values ​​to the fifth difference value respectively, and obtaining the matching value of the first suspected arc segment sequence and each of the second suspected arc segment sequences;

[0035] The first suspected arc segment sequence and the second suspected arc segment sequence corresponding to the maximum matching value or the matching value greater than the matching threshold are respectively used as the most likely target first suspected arc segment sequence and the target second suspected arc segment sequence.

[0036] In one implementation, determining the ring artifact area according to the plurality of pairs of target suspected arc segment sequences includes:

[0037] Calculate the matching values ​​and the number of matching values ​​of multiple pairs of target suspected arc sequence pairs corresponding to each suspected arc sequence as the starting sequence, and calculate the corresponding multiple matching average values;

[0038] Determine, according to the matching average value and the matching value, a probability value of each pair of the target suspected arc segment sequences being in the same annular area;

[0039] All target suspected arc sequence pairs whose likelihood values ​​are greater than the possible threshold are retained, and all target suspected arc sequence pairs retained by taking each suspected arc sequence as the starting sequence are connected along a preset direction to obtain multiple annular regions;

[0040] Select a candidate annular region from the annular region, draw a plurality of tangent lines at any position on the edge of the candidate annular region, and draw normal lines of the tangent lines respectively, and use the points where the plurality of normal lines intersect in the candidate annular region as candidate points, and return to the step of selecting a candidate annular region from the annular region until a candidate point corresponding to each of the annular regions is obtained, thereby obtaining a plurality of candidate points;

[0041] Determine a first candidate point and a second candidate point other than the first candidate point from the plurality of candidate points, and calculate a second distance between each of the first candidate points and the second candidate point and a sum of all the second distances;

[0042] When the inverse of the sum of the distances is greater than a preset threshold, the annular area corresponding to the first candidate point is taken as the target artifact area, and the step of determining the first candidate point and the second candidate point other than the first candidate point from the multiple candidate points is returned to determine multiple target artifact areas; the annular artifact area includes multiple target artifact areas.

[0043] In one implementation, performing artifact correction according to the annular artifact area includes:

[0044] determining a center of the annular artifact region;

[0045] Local median filtering is performed according to the center and each of the target artifact regions.

[0046] In one implementation, determining the center of the annular artifact area includes:

[0047] Determine respectively straight lines passing through candidate points in the target artifact region so that the reciprocal of the sum of the distances is greater than a preset threshold, the straight lines comprising a first segment from the edge of the target artifact region to the candidate point and a second segment from the candidate point to the edge of the target artifact region;

[0048] Determine the distance difference between the first segment and the second segment, and use all straight lines corresponding to the distance difference within a preset range as target straight lines;

[0049] All pending target artifact regions having corresponding target straight lines are determined, and the candidate point of the pending target artifact region having the largest reciprocal of the sum of distances among all pending target artifact regions is taken as the center of the annular artifact region.

[0050] In a second aspect, an embodiment of the present application provides a CT image artifact correction system, comprising:

[0051] An acquisition module, used for acquiring a CT image and determining a plurality of suspected arc segment sequences according to the CT image;

[0052] An extraction module, used to determine a chain code value set of each of the suspected arc sequence, and extract a plurality of target suspected arc sequence pairs according to the chain code value set;

[0053] A determination module, used for determining a ring artifact area according to a plurality of pairs of target suspected arc segment sequence pairs;

[0054] A correction module is used to perform artifact correction according to the annular artifact area.

[0055] In a third aspect, an embodiment of the present application provides a medical device, comprising: a processor and a memory, wherein the memory stores instructions, and the instructions are loaded and executed by the processor to implement a method in any one of the above-mentioned embodiments.

[0056] The present invention has the following beneficial effects:

[0057] By acquiring a CT image, and determining a plurality of suspected arc sequences according to the CT image; determining a chain code value set of each of the suspected arc sequences, extracting a plurality of target suspected arc sequence pairs according to the chain code value set, determining a ring artifact region according to the plurality of pairs of the target suspected arc sequence pairs, performing artifact correction according to the ring artifact region, introducing a chain code value set based on the variation characteristics of the chain code value in the ring artifact region to finally determine the ring artifact region, and performing artifact correction, the ring artifact region can be accurately located, thereby facilitating improving the correction effect and the quality of the CT image. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] 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.

[0059] Figure 1 A schematic flow chart of the steps of a CT image artifact correction method provided by one embodiment of the present invention;

[0060] Figure 2 A schematic diagram of transforming a CT image provided by an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of an edge CT image provided by an embodiment of the present invention;

[0062] Figure 4 A schematic diagram of a grayscale histogram provided by an embodiment of the present invention;

[0063] Figure 5 A schematic diagram of a process of obtaining 72 chain codes by subdividing 8 chain codes provided by an embodiment of the present invention;

[0064] Figure 6 A schematic diagram of determining a chain code value based on a 72 chain code provided by an embodiment of the present invention;

[0065] Figure 7 A structural block diagram of a CT image artifact correction system provided by an embodiment of the present invention.

[0066] Figure 8 This is a structural block diagram of a medical device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0067] 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 CT image artifact correction method, system and medical device proposed by the present invention, its specific implementation, structure, features and effects, in combination 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.

[0068] 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.

[0069] In the related art, the problem faced by removing the annular artifact area is that the image shape of the background organ area is relatively complex, and the grayscale gradient of the pixel points is relatively large, which makes the displayed artifact feature annular feature weak. The method of the embodiment of the present application can eliminate the interference caused by the artifact area to the medical CT image to a certain extent, so that the specific condition of the patient presented by the CT image is more accurate, thereby helping doctors to formulate a more accurate treatment plan. At the same time, the annular artifact area on the CT image is expressed as an arc segment pixel sequence. Since the image of the background organ area of ​​the medical CT image with annular artifacts is relatively complex and the grayscale gradient of the pixel points is relatively large, the arc segment feature is not obvious enough, and most of them are blocked by the background area, which makes it impossible to extract it by the existing technology. In the embodiment of the present application, the chain code value of the line connecting the endpoints of the arc sequence of pixel points in the artifact area on the same ring has an increasing (decreasing) trend and a changing gradient with the chain codes of the arc segments on both sides thereof, to judge the possibility of whether the two arc segments are on the same ring, thereby changing from screening of one ring area to screening of multiple ring areas, thereby determining the ring-shaped artifact area, avoiding the influence of the background organ area on the artifact area screening to a certain extent and improving the accuracy of screening the ring-shaped artifact area.

[0070] The specific scheme of a CT image artifact correction method, system and medical device provided by the present invention is described in detail below with reference to the accompanying drawings.

[0071] See also Figure 1 , which shows a flow chart of a CT image artifact correction method provided by an embodiment of the present invention. The CT image artifact correction method may at least include steps S100-S400:

[0072] S100: Acquire a CT image, and determine a plurality of suspected arc segment sequences based on the CT image.

[0073] S200, determining a chain code value set for each suspected arc sequence, and extracting a plurality of target suspected arc sequence pairs according to the chain code value set.

[0074] S300: Determine a ring artifact region according to multiple pairs of target suspected arc segment sequences.

[0075] S400: performing artifact correction according to the annular artifact area.

[0076] The technical solution of the embodiment of the present application obtains a CT image, and determines a plurality of suspected arc segment sequences based on the CT image; determines a chain code value set for each suspected arc segment sequence, extracts a plurality of target suspected arc segment sequence pairs based on the chain code value set, determines a ring artifact region based on the plurality of pairs of target suspected arc segment sequence pairs, performs artifact correction based on the ring artifact region, introduces a chain code value set based on the variation characteristics of the chain code value on the ring artifact region to finally determine the ring artifact region, and performs artifact correction, which can accurately locate the ring artifact region, thereby facilitating improving the correction effect and the quality of the CT image.

[0077] In one embodiment, the X-rays and detectors of a CT scanner are used to perform projection scanning on the patient's diseased area at multiple angles, and the scanned information is output in the form of a picture, thereby obtaining a CT image.

[0078] In one embodiment, determining multiple suspected arc segment sequences according to the CT image in step S100 includes steps S101-S105:

[0079] S101, performing inverse Fourier transform on the CT image to obtain a transformed CT image, and performing edge detection on the transformed CT image to obtain an edge CT image.

[0080] like Figure 2 and Figure 3 As shown, optionally, in order to enable the computer to more clearly identify its annular features, the CT image is subjected to inverse Fourier transform to obtain a transformed CT image, and the transformed CT image is subjected to edge detection to obtain an edge CT image (i.e. Figure 3 edge detection image). As you can see, Figure 2 In the transformed CT image, it can be observed that the ring artifact appears as a ring shape, but due to the occlusion of the background area, the ring area is interrupted or blocked over a large area, then the overall feature appears as multiple arc segments, and in the edge CT image, these arc segments are relatively weak compared to the features of the background area.

[0081] S102: Determine a grayscale histogram of the transformed CT image, and determine a plurality of target pixel points according to the grayscale histogram and a preset grayscale threshold.

[0082] In addition, to achieve targeted analysis of arc segments, it is necessary to preliminarily determine the suspected artifact area and background interference area. Figure 2 It can be found that the grayscale values ​​of the pixels in the suspected artifact area are higher and distributed in a curve shape, while there are also places with higher grayscale values ​​in the interference area, but these places are mostly block-shaped. Therefore, the grayscale value and shape characteristics of the suspected artifact area can be combined to distinguish other background areas. Therefore, the grayscale histogram of the transformed CT image is determined, and multiple target pixels are determined based on the grayscale histogram and the preset grayscale threshold, such as Figure 4 The grayscale histogram, with the horizontal axis being the number of pixels and the vertical axis being the grayscale value, shows that the number of pixels with a grayscale value of 200 increases suddenly. Combined with the pixel grayscale value characteristics and the regional area characteristics (number) of the annular artifact area in the grayscale image, the preset grayscale threshold is set to 200 (empirical value) here, and the pixels greater than the preset grayscale threshold are taken as target pixels, and subsequent extraction is performed to preliminarily generate a suspected artifact area.

[0083] S103, determining a plurality of suspected artifact regions based on a plurality of target pixel points, and determining a minimum bounding rectangle of each suspected artifact region, and determining the possibility that each suspected artifact region is an artifact region based on an aspect ratio of a length to a width of the minimum bounding rectangle and a pixel ratio of the target pixel points within the minimum bounding rectangle.

[0084] Optionally, different target pixel points may be distributed in a linear or block-like manner and have regional characteristics. Therefore, several suspected artifact areas can be determined accordingly, and then a minimum bounding rectangle is generated for these suspected artifact areas. In the embodiment of the present application, it is stipulated that the length of the minimum bounding rectangle is parallel to the x-axis direction, and the width is perpendicular to the direction. The larger the aspect ratio of the length to the width of the minimum bounding rectangle and the smaller the pixel ratio of the target pixel points in the minimum bounding rectangle, the more obvious the linear features of the area in the minimum bounding rectangle are, that is, the more obvious the linear features of the corresponding suspected artifact area are. Here, a formula can be used to express the possibility that a suspected artifact area is an artifact area. :

[0085]

[0086] Where D is the aspect ratio of the minimum circumscribed rectangle. Indicates the proportion of target pixels within the minimum bounding rectangle (i.e., the proportion of target pixels to all pixels within the minimum bounding rectangle). The larger the value, the greater the possibility that the suspected artifact area corresponding to the minimum bounding rectangle is the artifact area.

[0087] S104: Determine candidate artifact regions according to the possibility and the suspected artifact regions.

[0088] Optionally, implement setting a likelihood threshold, if the likelihood If it is greater than the possibility threshold, the suspected artifact region is considered to be a candidate artifact region. Therefore, all eligible candidate artifact regions can be determined to achieve preliminary distinction.

[0089] S105 . Determine multiple suspected arc segment sequences according to multiple candidate artifact regions and edge CT images.

[0090] It should be noted that, by using the above characteristics for exclusion, there may be some non-curved areas in the results obtained, and these areas usually show flake characteristics with unclear edges. In this case, the edge CT images often appear as scattered points without continuity. For these areas, here we refer to the corresponding areas of the edge CT images to extract multiple continuous pixel point sequences that meet the conditions, thereby obtaining multiple suspected arc segment sequences.

[0091] First, in the edge CT image, the pixel region corresponding to each candidate artifact region is determined.

[0092] Secondly, in each pixel area, the first distances between two adjacent pixels are calculated along the pixel arrangement direction (also called the trend of the pixel points); and the sum of the first distances is determined according to the first distances and the number of the first distances, and then the sum is divided by the number of the first distances to obtain the distance mean corresponding to each pixel area. .

[0093] Furthermore, according to the mean distance, the measurement parameter is determined , measurement parameters The specific calculation formula is:

[0094]

[0095] Optionally, a measurement threshold of 0.8 is set in advance, and the corresponding measurement parameter The candidate artifact regions larger than the measurement threshold of 0.8 are determined as suspected arc segment sequences, thereby obtaining multiple suspected arc segment sequences.

[0096] It should be noted that the aforementioned steps distinguish suspected ring sequences based on the grayscale value and shape features of the ring artifact area, and determine multiple suspected arc segment sequences. However, distinguishing only by grayscale value features and shape features cannot exclude the influence of other grayscale value-approximate areas and background organ areas with similar shapes on the identification of suspected ring sequences. Therefore, the present application proposes an improved method for identifying ring artifacts based on the chain code numerical features of ring artifacts. Specifically, the improved method is based on the features presented by the artifact area in the transformed CT image, which are manifested in the image as multiple arc segment pixel point sequences, which together form multiple ring regions, and the formed ring regions have the feature of sharing a center; the arc segments in a ring artifact area have a closed directional trend feature between them, which is reflected in the increasing (decreasing) trend of the chain code values ​​formed by the pixel point sequence in the area where the ring artifact is located, and the smaller the gradient fluctuation of the values ​​between the chain codes, the smoother the arc segment performance in the area and the closer it is to a circular ring. Specifically, it is manifested as follows: the line segment formed by the line connecting the endpoints of two suspected arc segment pixel point sequences has a chain code value between the chain code values ​​of the suspected arc segment pixel point sequences on both sides, and the values ​​of these chain codes are constantly increasing (decreasing), and the smaller the gradient value between the chain code value and the chain code values ​​of the suspected arc segments on both sides, the greater the possibility that the two suspected arc segment pixel point sequences are in the same annular area.

[0097] Therefore, in the embodiment of the present application, it is necessary to make judgments and analyses based on the chain code data. First, it is necessary to determine the chain code value set of each suspected arc sequence, and the chain code value set includes multiple chain code values. It should be noted that the existing 8-chain code represents fewer direction categories. If the 8-chain code is used, the obtained chain code values ​​are only 0, 1, 2, 3, 4, 5, 6, and 7, which is a total of eight. This situation lacks specificity in the gradient difference of the reaction chain code value. In order to make the directional characteristics of the subsequent suspected arc segments more specific, the embodiment of the present application subdivides the 8-chain code here to form a 72-chain code, such as Figure 5 As shown, this chain code further subdivides each of the 8 45° angles formed by the original 8 chain codes into 9 5° angles, and uses auxiliary tools to refine the change direction of the suspected arc area for the subsequent chain code value gradient change representation; based on the new chain code structure, arbitrarily locate one of the suspected arc sequences in the previous steps, indicating that its corresponding chain code value constitutes a chain code value set, and then match other arcs along a preset direction (for example, counterclockwise), and obtain the chain code value of the subsequent arc and the chain code value of the straight line between the two chain code endpoints.

[0098] In one implementation, in step S200, multiple target suspected arc sequence pairs are extracted according to the chain code value set, including steps S201-S211:

[0099] S201. From a plurality of suspected arc segment sequences, along a preset direction, using a first suspected arc segment sequence, determine a first suspected arc segment sequence and determine a second suspected arc segment sequence other than the first suspected arc segment sequence.

[0100] Optionally, assuming that the preset direction is counterclockwise, a suspected arc sequence is randomly determined as the first suspected arc sequence, and the first suspected arc sequence is used to determine the first suspected arc sequence and the second suspected arc sequence other than the first suspected arc sequence along the counterclockwise direction. It should be noted that the chain code value set A of the first suspected arc sequence is a series of chain code values ​​generated by constructing a chain code according to the arrangement direction of its pixel points, such as Figure 6 As shown, in the counterclockwise direction, multiple chain code values ​​can be determined based on the 72 chain codes to form a chain code value set.

[0101] S202: Locate the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence along a preset direction, and determine the first chain code value of the first suspected arc sequence and the second chain code value of the second suspected arc sequence.

[0102] Optionally, the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence are located in the counterclockwise direction, and the first chain code value of the first suspected arc sequence and the second chain code value of the second suspected arc sequence are determined. It should be noted that the first chain code value is the last chain code value in the chain code value set of the first suspected arc sequence along the preset direction, that is, the chain code value of the sub-arc segment between the second to last pixel point and the first endpoint of the first suspected arc sequence in the counterclockwise direction; the second chain code value is the first chain code value in the chain code value set of the second suspected arc sequence along the preset direction, that is, the chain code value of the sub-arc segment between the second endpoint and the second pixel point of the second suspected arc sequence in the counterclockwise direction.

[0103] S203, connecting the first endpoint and the second endpoint to obtain a line segment, and determining a line segment chain code value of the line segment.

[0104] Optionally, a line segment is obtained by connecting the first endpoint and the second endpoint, and a chain code value of the line segment is determined based on the 72 chain code, which is recorded as the line segment chain code value.

[0105] It should be noted that the direction trend of the suspected arc sequence is represented by the chain code value. If the chain code in the same annular area has an ascending (descending) order feature, the counterclockwise direction is specified as the positive direction in the embodiment of the present application, then the chain code in the same annular area should show an increasing feature. Since the establishment of the 72 chain code in this study is based on the structural basis of the 8 chain code, and since the direction of the starting direction value 0 of the 8 chain code is horizontal to the right, the chain code direction represented by the suspected arc pixel point sequence may include this direction. This situation is reflected in the increasing trend of the chain code value as a segmented increase feature. This feature has no effect when the pixel point feature is represented by the chain code value increase feature and the increase gradient, so here it is defaulted to 72+1=0.

[0106] At the same time, due to the performance differences of one or more detector channels in the CT system, concentric circular or arc artifacts will appear in the image. These annular artifact areas cannot be fully identified by existing technologies because of the complex shape of the background area and the large fluctuation of the grayscale value of the pixel points. They are only manifested by the characteristics of the arc segment pixel point sequence. These arc segments are composed of multiple concentric annular areas. The edge contours of the arc segments on the same annular area have annular characteristics, that is, the extension direction of the pixels formed by these annular contours is in a closed state. This closed state is reflected in the chain code as a trend of increasing or decreasing chain code values. Among them, the above steps obtain the corresponding chain code values ​​of the suspected arc segment sequence by constructing 72 chain codes. The change characteristics and gradients of these values ​​can be used as indicators to determine whether the pixel point sequences of two suspected arc segment sequences are on a suspected annular artifact. Therefore, a possibility analysis is performed based on these data features, and the suspected arc segment sequences that meet the conditions are connected according to the analysis results. It should be noted that the change gradient of the chain code value can be reflected by the change amount of the value increase, but there are also chain code values ​​that do not follow this increase feature; if the direction of the line connecting the suspected arc sequence identified by any suspected arc sequence is significantly different from the direction of the chain code at both ends, and the value of the formed chain code exists outside the chain code values ​​of the two sequence endpoints, then the pixel point sequences of the two suspected arcs may not be on the same suspected ring area. Therefore, while performing chain code value analysis, the suspected arc sequence can be preliminarily screened according to this feature, and then the relationship between the difference in the gradient of the chain code value change in each suspected arc sequence at both ends and the difference in the gradient of the chain code value change between the endpoints can be used to determine the possibility that the two suspected arc sequences are in the same suspected ring area.

[0107] S204. When the line segment chain code value is between the first chain code value and the second chain code value, first differences between two adjacent chain code values ​​in the chain code value set of the first suspected arc segment sequence are calculated respectively, and second differences between two adjacent chain code values ​​in the chain code value set of the second suspected arc segment sequence are calculated respectively.

[0108] Optionally, when the line segment chain code value is between the first chain code value m and the second chain code value n, that is, within the range of [m, n], the line segment is further extracted. Specifically, the first difference between the two adjacent chain code values ​​in the chain code value set of the first suspected arc sequence is calculated, and the second difference between the two adjacent chain code values ​​in the chain code value set of the second suspected arc sequence is calculated. It should be noted that since the direction of pixel point changes in the first suspected arc sequence and the second suspected arc sequence along the counterclockwise direction generally conforms to the feature of the increasing chain code value, there will be a very small number of points with inconsistent directions. Here, when calculating the numerical difference, the influence of individual abnormal points can be ignored to maximize the basic characteristics of the data.

[0109] S205. Calculate a mean of the first differences according to each first difference and the number of the first differences, calculate a mean of the second differences according to each second difference and the number of the second differences, and determine a first mean of the means of the first differences and the means of the second differences.

[0110] Optionally, the first difference values ​​are summed and then divided by the number of first differences to obtain the first difference mean, and the second difference values ​​are summed and then divided by the number of second differences to obtain the second difference mean; then, the first mean of the first difference means and the second difference means is determined. .

[0111] S206, calculating a third difference between the line segment chain code value and the first difference mean, calculating a fourth difference between the line segment chain code value and the second difference mean, and determining a second mean according to the third difference and the fourth difference.

[0112] Optionally, a third difference between the segment chain code value e and the first difference mean is calculated, and a fourth difference between the segment chain code value e and the second difference mean is calculated, and the second mean is determined by averaging the third difference and the fourth difference. .

[0113] S207: Determine, based on the first mean value and the second mean value, a probability value that the first suspected arc segment sequence and the second suspected arc segment sequence are the same suspected annular region.

[0114] Optionally, by first averaging and the second mean The difference between the line segment and the first suspected arc segment sequence and the second suspected arc segment sequence on both sides is used to express the relationship between the direction change, thereby indicating the possibility that the first suspected arc segment sequence and the second suspected arc segment sequence are the same suspected annular area. , the formula is as follows:

[0115]

[0116] In the formula, The larger the value is, the smaller the difference between the increase in the chain code values ​​of the first suspected arc sequence and the second suspected arc sequence and the corresponding increase in the chain code values ​​of the connecting area is. In other words, the greater the possibility that the first suspected arc sequence and the second suspected arc sequence are in the same annular area.

[0117] S208, returning to the step of determining the second suspected arc sequence other than the first suspected arc sequence, until the first suspected arc sequence and each second suspected arc sequence are obtained to obtain a probability value corresponding to the first suspected arc sequence.

[0118] Optionally, return to the step of determining the second suspected arc sequence other than the first suspected arc sequence in step S201 above, until the first suspected arc sequence and each second suspected arc sequence are obtained to obtain a probability value, that is, to obtain multiple probability values. .

[0119] S209: Determine the target first suspected arc segment sequence and the target second suspected arc segment sequence with the greatest probability according to each probability value, as a pair of target suspected arc segment sequence pairs.

[0120] First, from each possible size value , determine the maximum first possibility size value and the minimum possible size value .

[0121] Secondly, determine the fifth difference between the largest first likelihood magnitude value and the smallest likelihood magnitude value , and determine the value of each possibility separately With the minimum possible size value The sixth difference .

[0122] Furthermore, the ratio of each sixth difference to the fifth difference is calculated to obtain the matching value of the first suspected arc sequence and each second suspected arc sequence. , the specific formula is:

[0123]

[0124] Then, you can put the matching value When the matching value is the largest, the corresponding first suspected arc segment sequence and the second suspected arc segment sequence are respectively used as the target first suspected arc segment sequence and the target second suspected arc segment sequence with the greatest possibility, or the matching value is Compare the match value with the matching threshold (for example, 0.8) When it is greater than the matching threshold, the corresponding first suspected arc segment sequence and second suspected arc segment sequence are respectively used as the most likely target first suspected arc segment sequence and target second suspected arc segment sequence.

[0125] S210, determine a new first suspected arc sequence, and return to the step of determining a second suspected arc sequence other than the first suspected arc sequence according to the new first suspected arc sequence, until each suspected arc sequence is used as the first suspected arc sequence, and determine several pairs of target suspected arc sequence pairs using the first suspected arc sequence as the starting sequence.

[0126] Optionally, a new first suspected arc sequence is determined from the second suspected arc sequence other than the first suspected arc sequence, and the step of determining the second suspected arc sequence other than the first suspected arc sequence is returned based on the new first suspected arc sequence, i.e., the step of determining the second suspected arc sequence other than the first suspected arc sequence in step S201 is returned based on the new first suspected arc sequence, until each suspected arc sequence is used as the first suspected arc sequence, and several pairs of target suspected arc sequence pairs are determined using the first suspected arc sequence as the starting sequence.

[0127] S211. Using the second suspected arc sequence, determine the first suspected arc sequence and determine the second suspected arc sequence other than the first suspected arc sequence, return to the step of locating the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence along the preset direction, until obtaining several pairs of target suspected arc sequence pairs in which each suspected arc sequence is used as the starting sequence.

[0128] It should be noted that since the characteristic performance of the artifact area has a ring feature and is not limited to two suspected arc segment sequences, the other second suspected arc segment sequences determined according to the above method are sequentially used as the first suspected arc segment sequence, that is, as the starting sequence, and re-matched in the counterclockwise direction, and each time matching, the corresponding multiple Value, by these By analyzing the values, all the target first suspected arc segment sequences and target second suspected arc segment sequences that are most likely to be in a ring area are extracted as a pair of target suspected arc segment sequence pairs, and finally several pairs of target suspected arc segment sequence pairs are obtained in which each suspected arc segment sequence is used as the starting sequence.

[0129] Specifically, the second suspected arc sequence is used as the first suspected arc sequence, and then the second suspected arc sequence other than the first suspected arc sequence is determined, and the step of returning to step S202 to locate the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence along the preset direction is returned until several pairs of target suspected arc sequence pairs are obtained in which each suspected arc sequence is used as the starting sequence.

[0130] In one implementation, step S300 includes steps S301-S306:

[0131] S301, respectively calculating the matching values ​​and the number of matching values ​​of multiple pairs of target suspected arc segment sequences corresponding to each suspected arc segment sequence as a starting sequence, and calculating the corresponding multiple matching average values.

[0132] Optionally, the matching values ​​of multiple pairs of target suspected arc sequence pairs corresponding to each suspected arc sequence as the starting sequence are calculated respectively. And the number of matching values ​​K, calculate the corresponding multiple matching averages: .

[0133] S302: Determine the probability value of each pair of target suspected arc sequence pairs being in the same annular area according to the matching average value and the matching value.

[0134] Optionally, the probability value of each pair of target suspected arc sequence pairs in the same annular area The calculation formula is:

[0135]

[0136] Where K is the total number of matching values, For the The matching value of the target suspected arc sequence pair; where, The larger the value is, the greater the possibility that the pair of suspected arc sequence is in a ring area.

[0137] S303, retaining all target suspected arc sequence pairs whose likelihood values ​​are greater than a possible threshold, and connecting all target suspected arc sequence pairs retained by taking each suspected arc sequence as a starting sequence along a preset direction to obtain multiple annular areas.

[0138] Optionally, assuming that the possible threshold is set to 0.9, the possibility value All target suspected arc sequence pairs greater than the possible threshold of 0.9 are retained, and then all the target suspected arc sequence pairs retained by taking each suspected arc sequence as the starting sequence are connected along the preset counterclockwise direction to obtain multiple annular areas.

[0139] S304, select candidate annular areas from the annular areas, draw several tangents at any position on the edge of the candidate annular area, and draw normals of the tangents respectively, and use the points where the normals intersect in the candidate annular area as candidate points, and return to the step of selecting candidate annular areas from the annular areas until each annular area corresponds to a candidate point, and obtain multiple candidate points.

[0140] It should be noted that the mechanism of generating annular artifacts is closely related to the performance difference of the detector, that is, the inconsistency of the detector's probe response. This artifact appears as a straight line in the vertical direction on the projection sinusoidal diagram, and as a concentric ring or arc with the reconstruction center as the center on the CT image. By comparing the images, it is found that the artifact area presents multiple annular features, and these annular areas approximately share a center. In the embodiment of the present application, based on this feature, it is necessary to further analyze the multiple annular areas obtained to determine the possibility of whether they are annular artifact areas. Specifically, if these annular areas meet the feature that the annular areas presented by the annular artifact area approximately share a center, then these annular areas are considered to be annular artifact areas.

[0141] In an embodiment of the present application, a candidate annular area is selected from the annular area, and several tangents are made at any position on the edge of the candidate annular area, as well as normals Q of the tangents. The points where the several normals Q intersect in the candidate annular area are used as candidate points O. The step of selecting the candidate annular area from the annular area is returned until a candidate point corresponding to each annular area is obtained, and multiple candidate points O are obtained.

[0142] S305 : Determine a first candidate point and second candidate points other than the first candidate point from the plurality of candidate points, and calculate a second distance between each first candidate point and the second candidate point and a sum of all the second distances.

[0143] Optionally, a first candidate point and a second candidate point other than the first candidate point are determined from the plurality of candidate points O, and then a second distance between each first candidate point and the second candidate point is calculated. and the sum of all second distances ,in is the total number of candidate points O, For the The second distance of the candidate point O.

[0144] S306. When the inverse of the sum of the distances is greater than a preset threshold, the annular area corresponding to the first candidate point is taken as the target artifact area, and the process returns to the step of determining the first candidate point and the second candidate point other than the first candidate point from the multiple candidate points to determine multiple target artifact areas; the annular artifact area includes multiple target artifact areas.

[0145] Optionally, determine the inverse of the sum of distances , as the possibility that the ring area is a ring artifact area, when the inverse of the sum of distances is greater than a preset threshold, for example, assuming the preset threshold is 0.9, the annular area corresponding to the first candidate point is taken as the target artifact area, and the process returns to step S305 to determine the first candidate point and the second candidate point other than the first candidate point from multiple candidate points, and finally determine multiple target artifact areas; wherein the annular artifact area includes multiple target artifact areas.

[0146] In one implementation, step S400 includes steps S401-S402:

[0147] S401: Determine the center of the ring artifact area.

[0148] It should be noted that the determination of the position of the center (center point) of the annular artifact area needs to be based on the positional characteristics of the center of the annular artifact area, that is, the vector distances from the center point of the annular artifact area to the opposite directions of the edge area are approximately equal, which is equivalent to the two points in the edge area forming a straight line passing through the center point, and the center point is the midpoint of the straight line. Based on this principle, the distance difference between the distances corresponding to the two straight lines can be used as the value of the point for the area, and then the amplitude of the point for all annular areas can be calculated to select the best center.

[0149] First, determine the reciprocal of the sum of the distances through the target artifact area A straight line of candidate points greater than a preset threshold of 0.9; wherein the straight line includes a first segment from the edge of the target artifact area to the candidate point and a second segment from the candidate point to the edge of the target artifact area.

[0150] Secondly, determine the distance difference between the first segment and the second segment, that is, calculate the length of the first segment minus the length of the second segment and then remove the absolute value to get the distance difference between the first segment and the second segment. Then, all the straight lines corresponding to the distance difference between the preset range (for example, [0,1]) are taken as the target straight line. Because it is between [0,1]), the candidate point is likely to be the center point.

[0151] Furthermore, all pending target artifact regions with corresponding target straight lines are determined, and the reciprocal of the sum of distances in all pending target artifact regions is calculated. That is, the amplitude of the candidate point corresponding to point O as the center point, so the reciprocal of the sum of distances is selected The candidate point of the largest undetermined target artifact area is used as the center of the entire annular artifact area.

[0152] S402 : Perform local median filtering according to the center and each target artifact area.

[0153] It should be noted that the annular artifact area affects the presentation effect of the CT image. After the specific location of the annular artifact area is determined, the image of the artifact area can be corrected by image processing, thereby achieving the purpose of restoring the image.

[0154] Optionally, after the annular artifact area is determined by the above steps, correction is performed according to the position of the above center, local median filtering is performed on the center position, vertical stripes in the sinusoidal graph are eliminated, and the annular artifact area is eliminated. For example, based on the center and multiple target artifact areas identified in the above steps, a filter template is set, the selected filter template is covered on the pixel points of the target artifact area, and then all the pixels in the template are sorted according to the grayscale value. After sorting, the median of the grayscale values ​​of all pixels in the template is taken, and this median is assigned to the pixel points of these target artifact areas for replacement until all the pixel points of the target artifact areas of the annular artifact area are processed. In some embodiments, in order to avoid residual artifacts in the image, multiple processing can be performed, wherein the grayscale histogram of the CT image determined in step S100 is determined to be transformed to step S402 once, and multiple processing can be repeated; at the same time, the final filtered image is checked to ensure that all the target artifact areas of the annular artifact area are effectively corrected.

[0155] Reference Figure 7 , shows a structural block diagram of a CT image artifact correction system according to an embodiment of the present application, the system may include:

[0156] An acquisition module is used to acquire CT images and determine multiple suspected arc segment sequences based on the CT images;

[0157] An extraction module, used to determine a chain code value set of each suspected arc sequence, and extract multiple target suspected arc sequence pairs according to the chain code value set;

[0158] A determination module, used for determining the ring artifact area according to multiple pairs of target suspected arc segment sequence pairs;

[0159] The correction module is used to perform artifact correction according to the ring artifact area.

[0160] In the embodiment of the present application, the functions of each module in the system can be found in the corresponding description in the above method and will not be repeated here.

[0161] 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.

[0162] Reference Figure 8 , shows a block diagram of a medical device according to an embodiment of the present application, the medical device comprising: a memory 310 and a processor 320, the memory 310 stores instructions that can be run on the processor 320, the processor 320 loads and executes the instructions to implement the CT image artifact correction method in the above embodiment. The number of the memory 310 and the processor 320 can be one or more.

[0163] In one embodiment, the medical device further includes a communication interface 330 for communicating with external devices and performing data exchange transmission. If the memory 310, the processor 320 and the communication interface 330 are implemented independently, the memory 310, the processor 320 and the communication interface 330 can be interconnected through a bus and complete mutual communication. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0164] 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 CT image artifact correction method, characterized in that: The method comprises: Acquire a CT image, and determine a plurality of suspected arc segment sequences based on the CT image; Determine a chain code value set of each of the suspected arc sequence, and extract a plurality of target suspected arc sequence pairs according to the chain code value set; Determine the annular artifact area according to a plurality of pairs of target suspected arc segment sequence pairs; Performing artifact correction according to the annular artifact area; Among them, the method for obtaining the suspected arc segment sequence is as follows: performing inverse Fourier transform on the CT image to obtain a transformed CT image, and performing edge detection on the transformed CT image to obtain an edge CT image; determining the grayscale histogram of the transformed CT image, and determining a plurality of target pixel points according to the grayscale histogram and a preset grayscale threshold; determining a plurality of suspected artifact regions according to the plurality of target pixel points, and determining the minimum bounding rectangle of each of the suspected artifact regions, and determining the possibility that each of the suspected artifact regions is an artifact region according to the aspect ratio of the length to the width of the minimum bounding rectangle and the pixel proportion of the target pixel points in the minimum bounding rectangle; determining a candidate artifact region according to the possibility and the suspected artifact region; determining a plurality of suspected arc segment sequences according to the plurality of candidate artifact regions and the edge CT image; Among them, the method for extracting multiple target suspected arc segment sequence pairs is: From the plurality of suspected arc segment sequences, along a preset direction, using a first suspected arc segment sequence, determine a first suspected arc segment sequence and determine a second suspected arc segment sequence other than the first suspected arc segment sequence; Locate the first endpoint of the first suspected arc sequence and the second endpoint of the second suspected arc sequence along a preset direction, and determine a first chain code value of the first suspected arc sequence and a second chain code value of the second suspected arc sequence; wherein the first chain code value is the last chain code value in the chain code value set of the first suspected arc sequence along the preset direction, and the second chain code value is the first chain code value in the chain code value set of the second suspected arc sequence along the preset direction; Connecting the first endpoint and the second endpoint to obtain a line segment, and determining a line segment chain code value of the line segment; When the line segment chain code value is between the first chain code value and the second chain code value, first differences between two adjacent chain code values ​​in the chain code value set of the first suspected arc segment sequence are calculated respectively, and second differences between two adjacent chain code values ​​in the chain code value set of the second suspected arc segment sequence are calculated respectively; Calculating a first difference mean according to each of the first differences and the number of the first differences, calculating a second difference mean according to each of the second differences and the number of the second differences, and determining a first mean of the first difference means and the second difference means; Calculate a third difference between the line segment chain code value and the first difference mean, calculate a fourth difference between the line segment chain code value and the second difference mean, and determine a second mean according to the third difference and the fourth difference; Determine, according to the first mean and the second mean, a probability value that the first suspected arc segment sequence and the second suspected arc segment sequence are the same suspected annular region; Returning to the step of determining the second suspected arc segment sequence other than the first suspected arc segment sequence, until the first suspected arc segment sequence and each second suspected arc segment sequence are obtained to obtain a probability value corresponding to the first suspected arc segment sequence; According to each of the possibility values, determining the target first suspected arc segment sequence and the target second suspected arc segment sequence with the greatest possibility as a pair of target suspected arc segment sequence pairs; Determine a new first suspected arc segment sequence, and return to the step of determining a second suspected arc segment sequence other than the first suspected arc segment sequence according to the new first suspected arc segment sequence, until each of the suspected arc segment sequences is used as a first suspected arc segment sequence, and determine a plurality of target suspected arc segment sequence pairs using the first suspected arc segment sequence as a starting sequence; Using the second suspected arc sequence, determine the first suspected arc sequence and determine the second suspected arc sequence other than the first suspected arc sequence, return to the step of locating the first endpoint in the first suspected arc sequence and the second endpoint in the second suspected arc sequence along the preset direction, until obtaining a plurality of target suspected arc sequence pairs in which each suspected arc sequence is used as a starting sequence; The method for determining the ring artifact area is: respectively calculating the matching values ​​and the number of matching values ​​of multiple pairs of target suspected arc segment sequences corresponding to each suspected arc segment sequence as the starting sequence, and calculating the corresponding multiple matching average values; Determine, according to the matching average value and the matching value, a probability value of each pair of the target suspected arc segment sequences being in the same annular area; All target suspected arc sequence pairs whose likelihood values ​​are greater than the possible threshold are retained, and all target suspected arc sequence pairs retained by taking each suspected arc sequence as the starting sequence are connected along a preset direction to obtain multiple annular regions; Select a candidate annular region from the annular region, draw a plurality of tangent lines at any position on the edge of the candidate annular region, and draw normal lines of the tangent lines respectively, and use the points where the plurality of normal lines intersect in the candidate annular region as candidate points, and return to the step of selecting a candidate annular region from the annular region until a candidate point corresponding to each of the annular regions is obtained, thereby obtaining a plurality of candidate points; Determine a first candidate point and a second candidate point other than the first candidate point from the plurality of candidate points, and calculate a second distance between each of the first candidate points and the second candidate point and a sum of all the second distances; When the inverse of the sum of the distances is greater than a preset threshold, the annular area corresponding to the first candidate point is taken as the target artifact area, and the step of determining the first candidate point and the second candidate point other than the first candidate point from the multiple candidate points is returned to determine multiple target artifact areas; the annular artifact area includes multiple target artifact areas.

2. The CT image artifact correction method according to claim 1, characterized in that: The determining of a plurality of suspected arc segment sequences according to the plurality of candidate artifact regions and the edge CT images comprises: In the edge CT image, determining a pixel region corresponding to each of the candidate artifact regions; In each of the pixel regions, first distances between two adjacent pixel points are calculated along a pixel arrangement direction, and a distance mean corresponding to each of the pixel regions is determined according to the first distances and the number of the first distances; A measurement parameter is determined according to the distance mean, and the candidate artifact regions corresponding to the measurement parameter being greater than a measurement threshold are determined as suspected arc segment sequences, thereby obtaining a plurality of suspected arc segment sequences.

3. The CT image artifact correction method according to claim 1, characterized in that: Determining the target first suspected arc segment sequence and the target second suspected arc segment sequence with the greatest possibility according to each of the possibility magnitude values ​​includes: Determine the largest first likelihood value and the smallest likelihood value from among the likelihood values; determining a fifth difference between the largest first likelihood magnitude value and the smallest likelihood magnitude value, and determining a sixth difference between each of the likelihood magnitude values ​​and the smallest likelihood magnitude value, respectively; Calculating the ratio of each of the sixth difference values ​​to the fifth difference value respectively, and obtaining the matching value of the first suspected arc segment sequence and each of the second suspected arc segment sequences; The first suspected arc segment sequence and the second suspected arc segment sequence corresponding to the maximum matching value or the matching value greater than the matching threshold are respectively used as the most likely target first suspected arc segment sequence and the target second suspected arc segment sequence.

4. The CT image artifact correction method according to claim 3, characterized in that: The performing artifact correction according to the annular artifact area comprises: determining a center of the annular artifact region; Local median filtering is performed according to the center and each of the target artifact regions.

5. The CT image artifact correction method according to claim 4, characterized in that: Determining the center of the annular artifact area includes: Determine respectively straight lines passing through candidate points in the target artifact region so that the reciprocal of the sum of the distances is greater than a preset threshold, the straight lines comprising a first segment from the edge of the target artifact region to the candidate point and a second segment from the candidate point to the edge of the target artifact region; Determine the distance difference between the first segment and the second segment, and use all straight lines corresponding to the distance difference within a preset range as target straight lines; All pending target artifact regions having corresponding target straight lines are determined, and the candidate point of the pending target artifact region having the largest reciprocal of the sum of distances among all pending target artifact regions is taken as the center of the annular artifact region.

6. A CT image artifact correction system, characterized in that: The system is used to implement a CT image artifact correction method according to any one of claims 1 to 5, comprising: An acquisition module, used for acquiring a CT image and determining a plurality of suspected arc segment sequences according to the CT image; An extraction module, used to determine a chain code value set of each of the suspected arc sequence, and extract a plurality of target suspected arc sequence pairs according to the chain code value set; A determination module, used for determining a ring artifact area according to a plurality of pairs of target suspected arc segment sequence pairs; A correction module is used to perform artifact correction according to the annular artifact area.

7. A medical device, characterized in that: include: A processor and a memory, wherein the memory stores instructions, and the instructions are loaded and executed by the processor to implement the method according to any one of claims 1 to 5.

Images