A method for removing ring artifacts in CT images based on data fitting and statistical ranking

Abstract

This invention discloses a method for removing annular artifacts in CT images based on data fitting and statistical ranking. The method includes: first, acquiring original projection data and pre-correcting the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data; then, performing a secondary correction on the pre-corrected projection data using a one-dimensional filtering algorithm based on statistical ranking to obtain corrected projection data. This invention first utilizes the numerical abrupt change attribute of abnormal stripes in the projection data, employing a locally weighted data fitting approach to weaken these abrupt changes, thereby removing pseudo-texture information caused by abnormal stripes to a certain extent and providing a foundation for accurate correction of the projection data by subsequent filtering algorithms. Subsequently, a smoothing filtering algorithm based on data ranking is used to perform detail correction on the pre-corrected projection data, ultimately achieving fast and accurate correction of annular artifacts in CT images.

Description

Technical Field

[0001] This invention relates to the field of CT imaging technology, and in particular to a method for removing ring artifacts in CT images based on data fitting and statistical ranking. Background Technology

[0002] X-ray computed tomography (CT) scans acquire CT images of objects by reconstructing projection data. During this process, due to factors such as detector element defects and inconsistent responses, abnormal stripe elements that differ from surrounding elements may appear in the projection image, resulting in ring artifacts in the reconstructed CT image. As the name suggests, ring artifacts are a cluster of abnormal circular elements with varying radii centered at the reconstruction center. Ring artifacts degrade the quality of CT images, thus affecting subsequent image processing and analysis tasks, such as segmentation and quantitative analysis. Therefore, accurate and rapid correction of ring artifacts is of significant practical importance. Currently, based on the specific properties of ring artifacts in the projection data and reconstructed images, common techniques for removing ring artifacts from CT images can be divided into two categories: projection domain methods and reconstruction image domain methods.

[0003] Projection domain methods remove annular artifacts by correcting anomalous stripes in the projected data. The formation mechanism of annular artifacts shows that effectively removing anomalous linear elements from the projected data can reduce their impact on CT images. Transform domain concepts are used for annular artifact removal, primarily including wavelet transform and Fourier transform. Initially, researchers analyzed the properties of anomalous stripes in the frequency domain (obtained through Fourier transform) and designed corresponding numerical truncation methods with fixed thresholds to remove artifacts. Furthermore, techniques combining wavelet transform and frequency domain filtering have been used to correct anomalous stripes in the projected data. Discrete wavelet transform is used to extract feature images from the projected data, while Gaussian filters in the frequency domain smooth the anomalous line elements. In addition, some researchers have focused on studying the numerical function models corresponding to anomalous stripes and using these models to correct artifacts. In these algorithms, the mean data received by the detector unit is often used for model construction. Simultaneously, treating anomalous stripe elements as noise and designing appropriate filtering algorithms is also a possible approach to removing annular artifacts. If the projection domain method is successful, it can significantly remove ring artifacts in CT images. However, the complexity of ring artifacts is a significant challenge for these algorithms.

[0004] Image domain reconstruction methods correct annular artifacts by directly processing anomalous circular elements in CT images. On one hand, annular artifacts are "simple" because they possess a very significant geometric property in CT images: circularity. Based on this property, some works achieve attribute transformation of anomalous circular elements through the conversion between Cartesian and polar coordinate systems, and use efficient algorithms to correct the transformed anomalous linear elements. In addition, some research directly utilizes the circular property of annular artifacts to design related artifact removal algorithms. For example, the directional total variational method implemented using elliptic filters. Meanwhile, some deep learning methods have also been used for annular artifact removal. However, these methods heavily rely on the construction of training datasets, and their generalization ability is relatively poor. Furthermore, polar coordinate transformation methods perform poorly when annular artifacts are complex, mainly because the numerical values ​​of the reconstructed artifacts are generally quite complex.

[0005] Therefore, existing technologies still need improvement. Summary of the Invention

[0006] The technical problem to be solved by this invention is to provide a method for removing ring artifacts in CT images based on data fitting and statistical ranking, which addresses the shortcomings of existing technologies. This method aims to solve the problems of high complexity, low generalization, and unsatisfactory results of existing algorithms for removing ring artifacts in CT images.

[0007] The technical solution adopted by this invention to solve the problem is as follows:

[0008] In a first aspect, embodiments of the present invention provide a method for removing ring artifacts in CT images based on data fitting and statistical ranking, wherein the method includes:

[0009] The original projection data is obtained, and the original projection data is pre-corrected using a fitting algorithm based on mean projection to obtain pre-corrected projection data.

[0010] The pre-corrected projection data is then subjected to secondary correction using a one-dimensional filtering algorithm based on statistical sorting to obtain the corrected projection data.

[0011] In one implementation, the original projection data includes original sub-projection data received by a plurality of detector units, and the pre-correction of the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data includes:

[0012] Calculate the mean projection of the raw sub-projection data received by each detector unit;

[0013] The mean projection of the original sub-projection data received by each detector unit is used as the control point set to perform function fitting, thereby obtaining the fitting function;

[0014] The ideal mean projection of each detector unit is calculated using the fitting function.

[0015] The mean projection error of each detector unit is determined based on the mean projection and ideal mean projection data of the original sub-projection data of each detector unit.

[0016] The original sub-projection data is corrected based on the mean projection error of each detector unit to obtain the pre-corrected sub-projection data of each detector unit.

[0017] The pre-corrected projection data is obtained based on the pre-corrected sub-projection data of each detector unit.

[0018] In one implementation, the step of using a one-dimensional filtering algorithm based on statistical sorting to perform secondary correction on the pre-corrected projection data to obtain corrected projection data includes:

[0019] The pre-corrector projection data of each detector unit are sorted.

[0020] One-dimensional smoothing filtering is performed on the pre-corrector sub-projection data of each sorted detector unit to obtain the filtered sub-projection data of each detector unit.

[0021] The sorting transformation matrix is ​​used to recover the sub-projection data of all detector units to obtain the corrected projection data.

[0022] In one implementation, the method further includes, before performing one-dimensional smoothing filtering on the pre-corrector projection data of each sorted detector unit:

[0023] Obtain the raw index of the precorrector projection data for each detector unit;

[0024] A sorting transformation matrix is ​​generated based on the sorting results of the pre-corrector projection data and the original index of each detector unit, and the sorting transformation matrix is ​​saved.

[0025] In one implementation, the function fitting is a locally weighted linear function fitting.

[0026] In one implementation, the direction of the one-dimensional smoothing filter is the direction of the detector element.

[0027] In one implementation, the one-dimensional smoothing filter employs a one-dimensional Gaussian filter or a median filter.

[0028] Secondly, embodiments of the present invention also provide a CT image annular artifact removal system based on data fitting and statistical ranking, wherein the system includes:

[0029] The pre-correction module is used to acquire the original projection data and pre-correct the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data.

[0030] The secondary correction module is used to perform secondary correction on the pre-corrected projection data using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data.

[0031] In one implementation, the pre-correction module includes:

[0032] The mean projection calculation unit is used to calculate the mean projection of the raw sub-projection data received by each detector unit.

[0033] The function fitting unit is used to perform function fitting using the mean projection of the original sub-projection data received by each detector unit as the control point set, so as to obtain the fitting function;

[0034] A data calculation unit is used to calculate the ideal mean projection of each detector unit using the fitting function;

[0035] An error calculation unit is used to determine the mean projection error of each detector element based on the mean projection and ideal mean projection of the original sub-projection data of each detector element.

[0036] A data correction unit is used to correct the original sub-projection data based on the mean projection error of each detector unit to obtain pre-corrected sub-projection data for each detector unit.

[0037] A data generation unit is used to obtain the pre-corrected projection data based on the pre-corrected sub-projection data of each detector unit.

[0038] In one implementation, the secondary correction module includes:

[0039] A sorting unit is used to sort the pre-corrector projection data of each detector unit;

[0040] The filtering unit is used to perform one-dimensional smoothing filtering on the pre-corrector projection data of each detector unit after sorting, so as to obtain the filtered sub-projection data of each detector unit.

[0041] The data recovery unit is used to recover the filtered sub-projection data of all detector units using the sorting transformation matrix to obtain corrected projection data.

[0042] In one implementation, the secondary correction module further includes:

[0043] The index acquisition unit is used to acquire the original index of the pre-corrector projection data of each detector unit;

[0044] The sorting transformation matrix generation unit is used to generate a sorting transformation matrix based on the sorting result of the pre-corrector projection data and the original index of each detector unit, and to save the sorting transformation matrix.

[0045] Thirdly, embodiments of the present invention also provide a terminal device. The terminal device is a commercial display terminal, comprising a memory, a processor, and a program stored in the memory and executable on the processor for a CT image annular artifact removal method based on data fitting and statistical ranking. When the processor executes the program for the CT image annular artifact removal method based on data fitting and statistical ranking, it implements the steps of the CT image annular artifact removal method based on data fitting and statistical ranking described in any of the above-mentioned schemes.

[0046] Fourthly, embodiments of the present invention also provide a computer-readable storage medium. The computer-readable storage medium stores a program for a CT image annular artifact removal method based on data fitting and statistical ranking. When the program for the CT image annular artifact removal method based on data fitting and statistical ranking is executed by a processor, it implements the steps of the CT image annular artifact removal method based on data fitting and statistical ranking as described in any of the above-described solutions.

[0047] The beneficial effects of this invention are as follows: Compared with the prior art, this invention provides a method for removing annular artifacts in CT images based on data fitting and statistical ranking. First, the original projection data is acquired, and a fitting algorithm based on mean projection is used to pre-correct the original projection data to obtain pre-corrected projection data. Then, a one-dimensional filtering algorithm based on statistical ranking is used to perform secondary correction on the pre-corrected projection data to obtain corrected projection data. This invention first utilizes the numerical abrupt change attribute of abnormal stripes in the projection data, employing the idea of ​​local weighted linear function fitting to weaken abnormal abrupt changes, thereby removing pseudo-texture information in the projection data caused by abnormal stripes to a certain extent, and providing a foundation for subsequent filtering algorithms to accurately correct the projection data. Then, a smoothing filtering algorithm based on data ranking is used to perform detail correction on the pre-corrected projection data, ultimately achieving fast and accurate correction of annular artifacts in CT images. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments and the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1This is a schematic flowchart of the CT image ring artifact removal method based on data fitting and statistical sorting provided in an embodiment of the present invention.

[0050] Figure 2 This is a CT scanning and reconstruction process diagram of the CT image ring artifact removal method based on data fitting and statistical sorting provided in the embodiments of the present invention.

[0051] Figure 3 This is a projection diagram of a CT scan using a CT image ring artifact removal method based on data fitting and statistical sorting provided in an embodiment of the present invention.

[0052] Figure 4 This is a schematic diagram of the projection data and detector mean projection curve of the CT image annular artifact removal method based on data fitting and statistical sorting provided in the embodiments of the present invention.

[0053] Figure 5 This is an image showing the effect of ring artifact removal in a CT image based on data fitting and statistical sorting, as provided in an embodiment of the present invention.

[0054] Figure 6 This is a flowchart of the annular artifact correction process for the CT image annular artifact removal method based on data fitting and statistical sorting provided in this embodiment of the invention.

[0055] Figure 7 This is a block diagram illustrating the principle of a CT image ring artifact removal system based on data fitting and statistical sorting, as provided in an embodiment of the present invention.

[0056] Figure 8 This is a block diagram illustrating the internal structure of a CT image ring artifact removal device based on data fitting and statistical sorting, as provided in an embodiment of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0058] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0059] In existing technologies, the CT scanning and reconstruction process can be modeled as follows: Figure 2The content shown is as follows: X represents the object projection data obtained by the detector during the scanning process, and Y represents the reconstructed CT image obtained using the reconstruction algorithm P and the projection data X. This process can be specifically expressed as formula (1):

[0060] Y = P(X)(1)

[0061] However, during the projection process, due to inconsistent detector unit responses and other reasons, some abnormal elements exist in the projected data X, which in turn leads to ring artifacts in the reconstructed CT image Y, thus affecting the quality of the CT image. For example... Figure 3 The image shows the CT scan projection process of an object, with abnormal detector units (such as...) Figure 3 The data received by the detector (the dots in the image) exhibits a fixed pattern across all rotation angles. Therefore, this detector unit has problems throughout the scanning process, resulting in anomalous stripes along the rotation angle direction in the projected data. These elements have relatively obvious characteristics along the angular direction. Clearly, when the anomalous stripes in the projected data are relatively simple—for example, mainly consisting of a small difference between the values ​​of the anomalous stripes and their adjacent normal values, and a small number of anomalous stripes—these anomalous elements can be filtered out using filtering algorithms, thus removing weak artifacts. This is because the anomalous stripes are insufficient to form pseudo-structural information that affects the data structure, and there is sufficient effective information to ensure the normal operation of the filtering algorithm. However, when the anomalous stripes in the projected data are complex—for example, mainly consisting of a large difference between the values ​​of the anomalous stripes and their surrounding normal values, and a large number of anomalous stripes—the filtering algorithm typically cannot effectively correct the anomalous elements. This is because the anomalous stripes will result in a large amount of pseudo-texture information in the data, with relatively little effective data.

[0062] To address the problems in existing technologies, this embodiment provides a method for removing annular artifacts in CT images based on data fitting and statistical ranking. This method enables rapid and accurate removal of annular artifacts in CT images. In specific implementation, firstly, original projection data is acquired and pre-corrected using a fitting algorithm based on mean projection to obtain pre-corrected projection data. Then, a one-dimensional filtering algorithm based on statistical ranking is used to perform secondary correction on the pre-corrected projection data to obtain corrected projection data. This invention first utilizes the numerical abrupt change attribute of abnormal stripes in the projection data, employing the idea of ​​local weighted linear function fitting to weaken abnormal changes in the data. This, to a certain extent, removes the pseudo-texture information in the projection data caused by abnormal stripes and provides a foundation for accurate correction of the projection data by subsequent filtering algorithms. Then, a smoothing filtering algorithm based on data ranking is used to perform detail correction on the pre-corrected projection data, ultimately achieving rapid and accurate correction of annular artifacts in CT images.

[0063] For example, to remove ring artifacts from CT images, the original projection data is first obtained. A pre-correction algorithm is then used to initially correct the original projection data to remove abnormal stripes, resulting in pre-corrected projection data. This pre-correction algorithm is generated based on a locally weighted linear function fitting of the mean projection. After filtering out abnormal stripes through pre-correction, a one-dimensional filtering algorithm is used to perform a second correction on the pre-corrected projection data, i.e., detail correction, to obtain the final corrected projection data. This filtering algorithm is generated based on statistical sorting. This invention first utilizes the numerical abrupt change attribute of abnormal stripes in the original projection data, employing the idea of ​​locally weighted linear function fitting to weaken the abrupt changes in the data. This removes, to a certain extent, the pseudo-texture information in the original projection data caused by abnormal stripes, providing a foundation for the subsequent filtering algorithm to accurately correct the projection data. Then, a smoothing filtering algorithm based on data sorting is used to perform detail correction on the original projection data, ultimately achieving fast and accurate correction of ring artifacts in CT images.

[0064] Exemplary methods

[0065] This embodiment provides a method for removing ring artifacts in CT images based on data fitting and statistical ranking. This method can be applied to terminal devices for removing ring artifacts in CT images, specifically as follows: Figure 1 As shown, the method includes:

[0066] Step S100: Obtain the original projection data and pre-correct the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data.

[0067] In this embodiment, it is necessary to remove ring artifacts from CT images. Utilizing the geometric properties of ring artifacts in the projection domain, the abrupt change characteristics of the data received by the abnormal detector unit were studied. A pre-correction algorithm was designed to initially correct the acquired original projection data, resulting in pre-processed original projection data. In specific implementation, the original projection data includes original sub-projection data received by several detector units. The pre-correction of the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data includes: calculating the mean projection of the original sub-projection data received by each detector unit; performing function fitting using the mean projection of the original sub-projection data received by each detector unit as a control point set to obtain a fitting function; using the fitting function to calculate the ideal mean projection of each detector unit; determining the mean projection error of each detector unit based on the mean projection and ideal mean projection of the original sub-projection data of each detector unit; correcting the original sub-projection data based on the mean projection error of each detector unit to obtain pre-corrected sub-projection data for each detector unit; and obtaining the pre-corrected projection data based on the pre-corrected sub-projection data of each detector unit. The pre-correction algorithm can reduce the abrupt changes in data caused by abnormal stripes to a certain extent by utilizing the smoothing property of the fitting function, thereby initially realizing the correction of abnormal stripes in the projected image and obtaining the pre-processed corrected projection data, i.e., the pre-corrected projection data.

[0068] In practical implementation, because the anomalous data causing ring artifacts in the original projection data typically exhibits linear properties along the rotation angle, these lines visually appear to cover the data relatively uniformly based on the ideal values. Therefore, the average projection values ​​received by each detector unit at all scanning angles were studied, such as... Figure 4 The image shows the projection data from a CT scan and the mean projection curve of the detector. The background represents the projection data, the vertical axis represents the rotation angle, and the horizontal axis represents the detector unit. Figure 4 The curve in the figure represents the average projected value received by all detector units. It can be seen that the abnormal stripes in the projected data cause abrupt spikes in the average projected value curve of the detectors (e.g., ...). Figure 4(As shown by the arrow), this results in an uneven average projection curve, and the degree of abrupt changes in the peaks is positively correlated with the magnitude of the corresponding abnormal stripes. Furthermore, if the curve perturbation caused by abnormal peaks is disregarded, the change process of the average projection curve of the detector unit can well describe the complexity of the information received by each detector unit. Therefore, it is reasonable to weaken the pseudo-texture in the data caused by abnormal stripes to a certain extent before using filtering algorithms to correct the abnormal stripes in the original projection data. Meanwhile, based on the analysis of the influence of bar elements on the detector's average projection curve, it can be seen that pre-correction of the stripe elements can be achieved by compressing the abnormal peaks of this curve. To accurately and effectively remove abnormal peaks from the average projection curve and make the corrected average projection curve smoother, a locally weighted linear function can be fitted using the average projection values ​​of the detector units to obtain a function that is similar to and smoother than the average projection data. Then, the fitted function can be approximated as the ideal average projection curve, thereby calculating the difference between the abnormal average projection curve and the ideal average projection curve, and using this difference to remove abrupt peaks in the original average projection curve, thus achieving pre-correction of the abnormal stripes. For example, let the obtained raw projection data be set as follows: Where x represents the index of the detector unit, and y represents the angle during the scanning process of the detector unit, that is, the original projection data is composed of the original sub-projection data obtained by several detector units; calculate the mean projection data of the original sub-projection data of each detector unit at all angles, as shown in formula (2).

[0069]

[0070] Where, N y N represents the number of times the X-ray source rotates during the scan. x This represents the number of elements in the detector; after calculating the mean projection of the original sub-projection data of all detector elements, it can be used as a control point set, that is... As a set of control points, a locally weighted linear function fitting can be performed to obtain a smooth fitting function f(t), as shown in formula (3):

[0071]

[0072] Where Fit represents the locally weighted linear fitting function used in the data fitting process. The fitting of the mean projection curve in this application uses the mean projection of all detector units as control points. This fitting method can detect abnormal detector units and their corresponding perturbation values. The obtained fitting function can be used as the ideal mean projection curve to estimate the approximate ideal mean projection data of each detector unit. Based on the approximate ideal mean projection data and the mean projection of the original sub-projection data, the mean projection error of each detector unit can be obtained. That is, the potential error in the mean projection of the original sub-projection data is calculated using the fitting function f(t), as shown in formula (4).

[0073]

[0074] Where f(x) represents the value of the fitting function f(t) at point x, which is considered as approximately ideal projection data. The difference Res(x) reflects, to some extent, the difference between the abnormal stripes and the mean projection of the ideal data. Finally, the error of each detector unit is used to correct the original sub-projection data, as shown in formula (5).

[0075]

[0076] Where S′ represents the pre-corrected projection data, and the definition of S′(x, y) indicates that for a certain detector element x, at all angles y∈{0, ..., N}, y The data below are all corrected using the estimation error Res(x), thereby weakening the pseudo-texture properties caused by abnormal lines. This removes abrupt peaks in the original mean projection curve, giving it a smoother texture. Finally, all pre-corrected sub-projection data are combined to obtain pre-corrected projection data. Thus, the pre-correction process for abnormal stripes based on mean projection fitting is complete. This method can reduce the error information caused by complex abnormal stripes to a certain extent. However, pre-correction cannot completely correct ring artifacts; subsequent filtering algorithms are still needed to correct smaller anomalous elements.

[0077] In one implementation, the function fitting is a locally weighted linear function fitting. The fitting function f(t) is typically smooth, without sharp peaks, and can describe the variation of the control point set as accurately as possible. Simultaneously, it reduces the involvement of outlier data in the fitting process while increasing the involvement of normal data. Therefore, this fitting function can be considered an approximation of an ideal mean projection curve.

[0078] Step S200: The pre-corrected projection data is corrected a second time using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data.

[0079] In this embodiment, after preprocessing the original projection data to obtain preprocessed projection data, in order to further correct the small number of residual elements with small abnormal values ​​in the preprocessed projection data, this application utilizes a one-dimensional filtering algorithm based on statistical sorting to achieve the final correction of ring artifacts. Since the data fitting-based pre-correction algorithm performs a uniform supplementation or deletion operation on all angles received by each detector unit, the abnormal elements in the pre-correction data still exist in the form of lines, but the difference between the abnormal stripes and their adjacent data is not prominent. Because normal projection data usually contains a large amount of subtle detail information, in order to prevent the loss of data texture caused by the smoothing algorithm processing lines, a sorting-based one-dimensional filtering algorithm is used to perform the final correction on the pre-correction data, thereby effectively avoiding the loss of important data structure information caused by the smoothing operation. This filter can effectively recover the structural information in the image while removing ring artifacts. In specific implementation, the method of using a one-dimensional filtering algorithm based on statistical sorting to perform secondary correction on the pre-corrected projection data to obtain corrected projection data includes: sorting the pre-corrected sub-projection data of each detector unit; performing one-dimensional smoothing filtering on the sorted pre-corrected sub-projection data of each detector unit to obtain filtered sub-projection data of each detector unit; and using a sorting transformation matrix to recover the filtered sub-projection data of all detector units to obtain corrected projection data. The specific processing steps of the one-dimensional filtering algorithm based on statistical sorting are as follows: First, sort the pre-corrected sub-projection data of each detector unit at all angles in the manner shown in the following formula (6).

[0080] S′ s (x, y)=Sort(S′(X, y), 2), x∈{0,…,N x} (6)

[0081] Among them, S′ s The sorted pre-corrected projection data is represented by Sort(S′(x,y),2), which is the formal expression of the sorting operation. Here, 2 indicates that only the data in the second dimension of S′ is sorted, that is, the column vectors of matrix S′ (corresponding to the data under all angles) are sorted respectively. By sorting the pre-processed projection data under all angles received by the detector unit, similar data can be clustered. Through clustering, more similar data can participate in the filtering process, thereby effectively avoiding data loss and structural ambiguity caused by the presence of abnormal data. Then, a one-dimensional smoothing filter is performed on each pre-corrected sub-projection data. In this application, a one-dimensional filter along the detector direction is used to perform a one-dimensional smoothing filter on the pre-corrected sub-projection data to obtain the filtered sub-projection data of each detector unit, as shown in formula (7).

[0082]

[0083] Among them, S′ sf The filtered sorted projection data is Cov1, a one-dimensional filter with a radius only in the direction of the detector unit. This step reduces the involvement of abnormal stripe data in data recovery and improves the data restoration effect. Finally, the filtered sorted data is restored by using the sorting transformation matrix to recover the filtered sub-projection data of all detector units, resulting in the final corrected projection data. The sorting transformation matrix is ​​denoted as Rank, as shown in formula (8).

[0084]

[0085] Where S” represents the corrected projection data with normal order, Rank -1 It is a formal expression, representing the data S′ sf Restore according to the original Rank order.

[0086] In one implementation, before performing one-dimensional smoothing filtering on the pre-corrector projection data of each sorted detector unit, the method further includes: obtaining the original index of the pre-corrector projection data of each detector unit; generating a sorting transformation matrix based on the sorting result of the pre-corrector projection data and the original index of each detector unit, and saving the sorting transformation matrix. In a specific implementation, when sorting the pre-corrector projection data of each detector unit, the original index corresponding to each detector unit is simultaneously obtained, and a sorting transformation matrix is ​​generated from the original index according to the sorting result, and saved using Rank to facilitate the restoration of the order of the sorted data for subsequent filtering kernels.

[0087] In one implementation, the direction of the one-dimensional smoothing filter is the detector element direction. The main reasons for using only a one-dimensional filter along the detector element direction to correct anomalous data are: firstly, data in the angular direction may be corrupted by anomalous data; secondly, the sorting operation results in more similar data existing along the detector element direction. Therefore, performing one-dimensional smoothing filtering along the detector element direction can effectively correct anomalous elements in that one-dimensional region.

[0088] In one implementation, the one-dimensional smoothing filter employs a one-dimensional Gaussian filter and a median filter, both of which achieve good results.

[0089] In one implementation, due to the potential for overfitting caused by an excessive number of control points under certain conditions, a subset of continuous detector units' mean projections can be selected as control points, and piecewise fitting can be performed to ultimately fit and correct the entire mean projection curve. This approach is suitable for situations where the scanned object has complex information. Piecewise fitting allows for more accurate fitting of the mean projection data and effectively avoids fitting errors near drastically changing detector units. Therefore, when the projection data is relatively simple, the global mean projection data described in this application can be used as control points for fitting; when the projection data is more complex, piecewise fitting can improve the pre-correction effect.

[0090] Figure 5 The experimental results of annular artifact correction achieved using this application are shown below. Figure 5 'a' is the original projection data containing the anomalous stripe elements. Figure 5 b is the effect after removing linear elements from the original projection data using this invention. Figure 5 a and Figure 5 As can be seen from the comparison of b, the present invention can effectively correct and remove abnormal stripe elements in the projection data, and other effective information in the projection data is also effectively preserved and restored. Figure 5 c uses the original projection data ( Figure 5 a) Reconstructed CT image Figure 5 d is the CT image reconstructed using the corrected projection data. The comparison shows that this invention can effectively remove ring artifacts in CT images, and the structural information in the image is also effectively restored.

[0091] In this invention, Figure 6 This is a flowchart illustrating the overall framework, combining a pre-correction algorithm for anomalous stripes based on data fitting with a one-dimensional filtering algorithm based on statistical ranking. The pre-correction algorithm based on data fitting provides a more reasonable data foundation for the one-dimensional filtering algorithm based on statistical ranking. It utilizes the abrupt changes in data caused by anomalous stripe elements to effectively weaken the pseudo-image structures in the projection data caused by these anomalous data, thus providing more reliable and effective data for subsequent filtering algorithms. The one-dimensional filtering algorithm based on statistical ranking uses more similar data to filter out anomalous elements through ranking, effectively avoiding the loss of image structural information during the filtering process. By combining these techniques, effective removal of ring artifacts in CT images can be achieved.

[0092] The method disclosed in this invention designs a reasonable fringe element pre-correction algorithm based on the geometric properties of annular artifacts in projection data and the smoothing properties of the fitting function. This solves the problems of poor results and data loss caused by directly using filtering algorithms to correct artifacts when annular artifacts are severe. This method is applicable to various CT image reconstruction methods and can achieve fast and accurate correction of annular artifacts in CT images.

[0093] Exemplary System

[0094] Based on the above embodiments, the present invention also provides a CT image ring artifact removal system based on data fitting and statistical ranking, such as... Figure 7 As shown, the system in this embodiment includes a pre-correction module 10 and a secondary correction module 20. Specifically, the pre-correction module 10 is used to acquire the original projection data and pre-correct the original projection data using a pre-correction algorithm based on mean projection fitting to obtain pre-corrected projection data; the secondary correction module 20 is used to perform secondary correction on the pre-corrected projection data using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data.

[0095] In one implementation, the pre-correction module includes:

[0096] The mean projection calculation unit is used to calculate the mean projection of the raw sub-projection data received by each detector unit.

[0097] The function fitting unit is used to perform function fitting using the mean projection of the original sub-projection data received by each detector unit as the control point set, so as to obtain the fitting function;

[0098] A data calculation unit is used to calculate the ideal mean projection of each detector unit using the fitting function;

[0099] An error calculation unit is used to determine the mean projection error of each detector element based on the mean projection and ideal mean projection of the original sub-projection data of each detector element.

[0100] A data correction unit is used to correct the original sub-projection data based on the mean projection error of each detector unit to obtain pre-corrected sub-projection data for each detector unit.

[0101] A data generation unit is used to obtain the pre-corrected projection data based on the pre-corrected sub-projection data of each detector unit.

[0102] In one implementation, the secondary correction module includes:

[0103] A sorting unit is used to sort the pre-corrector projection data of each detector unit;

[0104] The filtering unit is used to perform one-dimensional smoothing filtering on the pre-corrector projection data of each detector unit after sorting, so as to obtain the filtered sub-projection data of each detector unit.

[0105] The data recovery unit is used to recover the filtered sub-projection data of all detector units using the sorting transformation matrix to obtain corrected projection data.

[0106] In one implementation, the secondary correction module further includes:

[0107] The index acquisition unit is used to acquire the original index of the pre-corrector projection data of each detector unit;

[0108] The sorting transformation matrix generation unit is used to generate a sorting transformation matrix based on the sorting result of the pre-corrector projection data and the original index of each detector unit, and to save the sorting transformation matrix.

[0109] The working principle of each module in the CT image ring artifact removal system based on data fitting and statistical sorting in this embodiment is the same as that of each step in the above method embodiment, and will not be repeated here.

[0110] Based on the above embodiments, the present invention also provides a terminal device, the schematic diagram of which can be as follows: Figure 8 As shown. The terminal device may include one or more processors 100 ( Figure 8 (Only one is shown in the image), memory 101, and computer program 102 stored in memory 101 and executable on one or more processors 100, such as a program for a CT image annular artifact removal method based on data fitting and statistical ranking. When one or more processors 100 execute computer program 102, they can implement various steps in the method embodiment of the CT image annular artifact removal method based on data fitting and statistical ranking. Alternatively, when one or more processors 100 execute computer program 102, they can implement the functions of various modules / units in the apparatus embodiment of the CT image annular artifact removal method based on data fitting and statistical ranking, which is not limited here.

[0111] In one embodiment, the processor 100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0112] In one embodiment, memory 101 may be an internal storage unit of an electronic device, such as a hard drive or RAM. Memory 101 may also be an external storage device of the electronic device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc. Furthermore, memory 101 may include both internal and external storage units. Memory 101 is used to store computer programs and other programs and data required by the terminal device. Memory 101 can also be used to temporarily store data that has been output or will be output.

[0113] Those skilled in the art will understand that Figure 8 The block diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal device to which the present invention is applied. Specific terminal devices may include, but are not limited to, those shown above. Figure 8 The diagram shows more or fewer components, or combinations of certain components, or different component arrangements.

[0114] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, operational databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual operating data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0115] In summary, this invention discloses a method for removing annular artifacts in CT images based on data fitting and statistical sorting. The method includes: first, acquiring original projection data and pre-correcting the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data; then, performing a secondary correction on the pre-corrected projection data using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data. This invention first utilizes the numerical abrupt change attribute of abnormal stripes in the projection data, using the idea of ​​data fitting to weaken the abrupt changes in the data, thereby removing pseudo-texture information in the projection data caused by abnormal stripes to a certain extent, and providing a data foundation for the subsequent filtering algorithm to accurately correct the projection data. Then, a smoothing filtering algorithm based on data sorting is used to perform detail correction on the original projection data, ultimately achieving fast and accurate correction of annular artifacts in CT images.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for removing ring artifacts in CT images based on data fitting and statistical ranking, characterized in that, The method includes: The original projection data is obtained, and the original projection data is pre-corrected using a fitting algorithm based on mean projection to obtain pre-corrected projection data. A one-dimensional filtering algorithm based on statistical sorting is used to perform secondary correction on the pre-corrected projection data to obtain the corrected projection data. The original projection data includes original sub-projection data received by several detector units. The pre-correction of the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data includes: Calculate the mean projection of the raw sub-projection data received by each detector unit; The mean projection of the original sub-projection data received by each detector unit is used as the control point set to perform function fitting, thereby obtaining the fitting function; The ideal mean projection of each detector unit is calculated using the fitting function. The mean projection error of each detector element is determined based on the mean projection and ideal mean projection of the raw sub-projection data of each detector element. The original sub-projection data is corrected based on the mean projection error of each detector unit to obtain the pre-corrected sub-projection data of each detector unit. The pre-corrected projection data is obtained based on the pre-corrected sub-projection data of each detector unit; The pre-corrected projection data is then subjected to secondary correction using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data, including: The pre-corrector projection data of each detector unit are sorted. One-dimensional smoothing filtering is performed on the pre-corrector sub-projection data of each sorted detector unit to obtain the filtered sub-projection data of each detector unit. The sorting transformation matrix is ​​used to recover the filtered sub-projection data of all detector units to obtain the corrected projection data; The process includes, prior to performing one-dimensional smoothing filtering on the pre-corrector projection data of each sorted detector unit: Obtain the raw index of the precorrector projection data for each detector unit; A sorting transformation matrix is ​​generated based on the sorting results of the pre-corrector projection data and the original index of each detector unit, and the sorting transformation matrix is ​​saved.

2. The method for removing ring artifacts in CT images based on data fitting and statistical ranking according to claim 1, characterized in that, The function fitting is a locally weighted linear function fitting.

3. The method for removing ring artifacts in CT images based on data fitting and statistical ranking according to claim 1, characterized in that, The direction for performing one-dimensional smoothing filtering is the direction of the detector unit.

4. The method for removing ring artifacts in CT images based on data fitting and statistical ranking according to claim 3, characterized in that, The one-dimensional smoothing filter uses a one-dimensional Gaussian filter or a median filter.

5. A CT image annular artifact removal system based on data fitting and statistical ranking, characterized in that, The system includes: The pre-correction module is used to acquire the original projection data and pre-correct the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data. The secondary correction module is used to perform secondary correction on the pre-corrected projection data using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data; The original projection data includes original sub-projection data received by several detector units. The pre-correction of the original projection data using a fitting algorithm based on mean projection to obtain pre-corrected projection data includes: Calculate the mean projection of the raw sub-projection data received by each detector unit; The mean projection of the original sub-projection data received by each detector unit is used as the control point set to perform function fitting, thereby obtaining the fitting function; The ideal mean projection of each detector unit is calculated using the fitting function. The mean projection error of each detector element is determined based on the mean projection and ideal mean projection of the raw sub-projection data of each detector element. The original sub-projection data is corrected based on the mean projection error of each detector unit to obtain the pre-corrected sub-projection data of each detector unit. The pre-corrected projection data is obtained based on the pre-corrected sub-projection data of each detector unit; The pre-corrected projection data is then subjected to secondary correction using a one-dimensional filtering algorithm based on statistical sorting to obtain corrected projection data, including: The pre-corrector projection data of each detector unit are sorted. One-dimensional smoothing filtering is performed on the pre-corrector sub-projection data of each sorted detector unit to obtain the filtered sub-projection data of each detector unit. The sorting transformation matrix is ​​used to recover the filtered sub-projection data of all detector units to obtain the corrected projection data; The process includes, prior to performing one-dimensional smoothing filtering on the pre-corrector projection data of each sorted detector unit: Obtain the raw index of the precorrector projection data for each detector unit; A sorting transformation matrix is ​​generated based on the sorting results of the pre-corrector projection data and the original index of each detector unit, and the sorting transformation matrix is ​​saved.

6. A terminal device, characterized in that, The terminal device includes a memory, a processor, and a CT image annular artifact removal method program based on data fitting and statistical sorting stored in the memory and executable on the processor. When the processor executes the CT image annular artifact removal method program based on data fitting and statistical sorting, it implements the steps of the CT image annular artifact removal method based on data fitting and statistical sorting as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program for a CT image annular artifact removal method based on data fitting and statistical sorting. When the program is executed by a processor, it implements the steps of the CT image annular artifact removal method based on data fitting and statistical sorting as described in any one of claims 1-4.

Images