WO2012055147A1 - 计算机断层成像(ct)图像重建方法及装置 - Google Patents

计算机断层成像(ct)图像重建方法及装置 Download PDF

Info

Publication number
WO2012055147A1
WO2012055147A1 PCT/CN2011/000830 CN2011000830W WO2012055147A1 WO 2012055147 A1 WO2012055147 A1 WO 2012055147A1 CN 2011000830 W CN2011000830 W CN 2011000830W WO 2012055147 A1 WO2012055147 A1 WO 2012055147A1
Authority
WO
WIPO (PCT)
Prior art keywords
projection data
projection
fdk
image reconstruction
filtering
Prior art date
Application number
PCT/CN2011/000830
Other languages
English (en)
French (fr)
Inventor
李亮
陈志强
张丽
康克军
邢宇翔
赵自然
肖永顺
黄志峰
Original Assignee
清华大学
同方威视技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 清华大学, 同方威视技术股份有限公司 filed Critical 清华大学
Priority to EP11743411.8A priority Critical patent/EP2469472A4/en
Priority to US13/140,761 priority patent/US8724889B2/en
Publication of WO2012055147A1 publication Critical patent/WO2012055147A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/002D [Two Dimensional] image generation
    • G06T11/003Reconstruction from projections, e.g. tomography
    • G06T11/006Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2211/00Image generation
    • G06T2211/40Computed tomography
    • G06T2211/421Filtered back projection [FBP]

Definitions

  • the present invention generally relates to the field of CT images, and more particularly to a CT image reconstruction method and apparatus. Background technique
  • CT has become one of the most important detection methods in the medical, biological, aerospace, and defense industries.
  • CT scanning modes and imaging methods are also constantly improving, and three-dimensional cone beam CT has become the mainstay of research and application.
  • X-ray cone beam CT has been widely used in medical clinical, safety inspection, non-destructive testing and other fields.
  • the cone beam CT system based on circular orbit scanning is widely used in clinical, security and industrial nondestructive testing because it is relatively simple in terms of mechanical and electrical control, and is easy to implement.
  • the FDK method can be seen as an approximate extension of the fan beam FBP (Filtered Backprojection) method in three dimensions.
  • the FDK method includes the following steps: first, performing weighting processing on the projection data; then performing one-dimensional filtering on the projection data of different projection angles in the horizontal direction; finally performing three-dimensional back projection along the reverse direction of the X-ray to obtain the final three-dimensional reconstructed image of the object. .
  • the voxel reconstruction value of the FDK method is the sum of the ray contributions of the voxel in the 360 ⁇ projection angle range, so the circular orbit cone beam FDK method is an approximation method with simple mathematical formula, fast and easy method calculation. It is characterized by engineering realization, and when the cone angle is relatively small (generally within ⁇ 5°), it can achieve good reconstruction effect, so it is widely recognized in practical engineering applications.
  • the FDK method also has certain problems: Since the circular orbital scanning itself does not satisfy the data completeness condition of the cone beam accurate reconstruction, there is a data loss problem of Radon; therefore, when the cone beam angle is increased, the FDK method The reconstruction results in severe cone beam artifacts. The FDK reconstruction value decreases rapidly in the plane away from the scanning orbit, which makes the application of the flat panel detector CT imaging system very limited.
  • the P-FDK method is to obtain the parallel fan beam projection data by rearranging the cone beam projection data, and then reconstruct the three-dimensional image of the object by the filtered back projection method.
  • Fig. 1 (a) it is a scanning schematic diagram of a circular orbital cone beam CT system using a flat panel detector, and SC ⁇ indicates the position of the X-ray source on a circular orbit, indicating the angular sampling position projected on the circular orbit.
  • ( ?, ⁇ , ⁇ ) represents the projection data on the flat panel detector,
  • W is a Cartesian coordinate system defined on the flat panel detector, used to indicate the position coordinates of each X-ray projection on the detector, R is The circular orbital radius;
  • Figure (b) is a fan beam projection that uses the P-FDK method to rearrange the cone beam projection into a parallel beam.
  • the black solid point is the X-ray source.
  • the rearranged projection data is the same. The distance from the parallel fan beam to the virtual detector is different at the angle, so the projection of the previous row of the original flat panel detector corresponds to the virtual detector being no longer horizontal, but on a curve that is convex along the center line of the virtual detector.
  • the data rearrangement formula of P-FDK is as shown in formula (1).
  • the subsequent derivation process of the present invention also takes a flat panel detector as an example, and other detector types such as cylindrical detectors can be based on the flat panel. The detector is deformed accordingly, and will not be described.
  • ( ⁇ ) is a Cartesian coordinate system on the central virtual detector after rearrangement, indicating the position coordinates of each X-ray on the virtual detector after rearrangement.
  • the difference between the P-FDK method and the FDK method is that the P-FDK method is first rearranged into parallel fan beams, so that the process of calculating the weighting coefficients is omitted in the back projection, and there is no difference in image quality from the FDK method.
  • the T-FDK method proposed by Grass et al. in 2000 is based on the P-FDK method.
  • T-FDK The projections are rearranged into parallel fan beams, and are also secondarily rearranged in the vertical fan beam plane, ie
  • T-FDK needs to rearrange the projection data in both horizontal and vertical directions, and finally the difference between T-FDK and P-FDK is that the filtering direction of the projection data by T-FDK is along the horizontal direction of the central virtual detector. Performed as shown in Figure 1 (c), rather than along the convex curve direction.
  • the data rearrangement formula of the T-FDK method is as follows.
  • T-FDK is similar to P-FDK in that it eliminates the weighting coefficient of back projection because it is rearranged into parallel fan beams, so it is more efficient in calculation; at the same time, because of T-FDK method
  • the filtering of the projection data is performed along the horizontal direction of the central virtual detector, which reduces the cone beam artifact caused by the increase of the cone angle, improves the quality of image reconstruction, and realizes the scanning through the circular orbit by using the large-area flat panel detector. Accurate three-dimensional imaging of large objects is possible.
  • Another way to improve the FDK method is to use the circular orbit to scan the conjugate rays in the projection, and to improve the reconstructed image quality by selecting different back projection weight coefficients for the conjugate projection, and also achieve better results.
  • the main technical problem to be solved by the present invention is to provide a CT image reconstruction method and apparatus capable of eliminating severe cone beam artifacts under a large cone angle.
  • the technical solution of the CT image reconstruction method of the present invention includes the following steps:
  • the filtered projection data is back-projected in three dimensions in the ray direction.
  • the step of selecting the projection data of the same height on the curve that approximates the curvature of the circular orbit includes selecting the projection data according to the following formula:
  • the step of weighting the selected projection data includes processing the selected projection data according to the following formula:
  • g e — "" ⁇ , ⁇ ) represents filtered projection data
  • ® represents convolution
  • the step of performing three-dimensional back projection on the filtered projection data in the ray direction comprises performing three-dimensional back projection according to the following formula:
  • f c ⁇ FDK ( , y,z) ⁇ g c ⁇ FDK ( ⁇ , t(x, y, ⁇ ), c(x, y, z, ⁇ )) ⁇
  • the dish ⁇ , ⁇ ) represents the reconstructed image in the x-axis, y-axis, and z- axis directions.
  • the technical solution of the CT image reconstruction device of the present invention includes:
  • a rearrangement unit for selecting projection data of the same height on a curve that approximates the curvature of the circular orbit
  • a weighting unit configured to perform weighting processing on the selected projection data
  • a filtering unit configured to filter the weighted processing data in a horizontal direction
  • a back projection unit to perform three-dimensional back projection on the filtered projection data in a ray direction.
  • the rearrangement unit selects projection data according to the following formula:
  • P - ( 'c) represents the selected projection data; represents the projection direction; represents the distance between parallel fan beams; c represents the angular sampling interval in the x-axis direction; R represents the radius of the circular orbit.
  • the weighting unit processes the selected projection data according to the following formula:
  • the back projection unit performs three-dimensional back projection according to the following formula:
  • the dish ⁇ , ⁇ ) represents a reconstructed image in the X-axis, y-axis, and Z- axis directions.
  • the beneficial effects of the CT image reconstruction method and apparatus of the present invention are as follows: First, the rearrangement of the projection data by the present invention uses projection data of the same height on a curve that approximates the curvature of the scanning circular orbit, thereby Sampling the concave curve along the virtual center detector along the virtual center detector in the plane parallel to the Z-axis, which greatly improves the numerical accuracy of the reconstruction method under the large cone angle, effectively suppressing the cone caused by the large cone angle Bunch of artifacts.
  • the present invention has high efficiency and stability.
  • Figure 1 (a) is a schematic view of a CT scan of a circular orbital cone beam of a flat panel detector
  • FIG. 1(b) is a schematic diagram showing the rearrangement of the projection data obtained in Fig. 1(a) by the P-FDK method
  • Fig. 1(c) is a rearrangement of the projection data obtained in Fig. 1(a) by the T-FDK method
  • FIG. 1(d) is a schematic diagram showing the rearrangement of the projection data obtained in FIG. 1(a) by using the CT image reconstruction method of the present invention
  • Figure 2 is a plan view of the l-axis along the axis of rotation of Figures l(a), 1(b), 1(c) and 1(d);
  • Figure 3 is a side elevational view of the vertical section of Figure 2 through SC;
  • FIG. 4 is a flow chart of a CT image reconstruction method of the present invention.
  • Figure 5(a) shows the projection data acquired by the circular orbit CT scan, which is a sinogram composed of the central layer data of the flat panel detector at various angles;
  • FIG. 5 (b) is a sinogram of the data of the central layer of the virtual detector at various angles after the projection data is rearranged by the C-FDK method of the present invention
  • FIG. 5(c) is a schematic diagram of projection data obtained by performing weighting processing on the projection data shown in FIG. 5(b);
  • Figure 6(a) is a schematic representation of an accurate image of a three-dimensional Shepp-logan head model in a vertical section;
  • Figure 6(b) is a schematic diagram of the reconstruction results obtained by the FDK method.
  • Figure 6(c) is a schematic diagram of the reconstruction result obtained by the T-FDK method
  • Fig. 6(d) is a schematic view showing the reconstruction result obtained by the C-FDK method of the present invention
  • Fig. 7 is a section taken along the lines of Fig. 6(a), Fig. 6(b), Fig. 6(c) and Fig. 6(d) Schematic diagram of the reconstruction results after the hatching of the vertical center;
  • Figure 8 is a schematic illustration of a CT image reconstruction apparatus of the present invention. Detailed ways
  • the CT image reconstruction method of the present invention includes the steps of:
  • the CT image reconstruction method of the present invention performs rearrangement by selecting projection data of the same height on a curve that approximates the curvature of the circular orbit, and then weighting the selected projection data, and then performing weighting processing.
  • the projection data is filtered in the horizontal direction, and finally the three-dimensional CT image of the scanned object is obtained by performing three-dimensional back projection on the filtered projection data along the ray direction.
  • the curvature of the scanning circular orbit is 1/R
  • the approximate average curvature may be 1 / 2R 2 2R.
  • the filtering core may be the most basic ramp filtering kernel, or may be a filtering kernel after the standard ramp filtering is smoothed in the frequency domain, such as The SL filter kernel, etc. (SL filter kernel was proposed by LAShepp and BFLogan in 1974).
  • a flat panel detector is taken as an example to describe the technical solution of the present invention.
  • the invention can be applied to other area array detectors such as cylindrical detectors.
  • Fig. 1 first, define the scan path of the in-plane cone beam X-ray source as a circle C ⁇ - R ⁇ osAsin); R is the radius of the circular orbit; represent the angular parameter corresponding to the source point; O is the coordinate origin and The center of the circular orbit, which is the center of rotation.
  • (?, ⁇ , ⁇ ) represents projection data acquired on the flat panel detector 1 after X-ray irradiation of the object to be scanned, wherein ( ⁇ , >) represents the horizontal and vertical coordinates of a position of a projection point on the two-dimensional flat panel detector 1.
  • the method selects the projection data of the cone beam projection data on the convex curve on the central virtual detector 2, as shown in Fig. 1 (b).
  • Fig. 2 is a top view from top to bottom along the Z axis of the rotation axis.
  • the area array detector becomes a straight line or a curve, and the solid point S on the circular track represents an X-ray source.
  • the process of rearranging the data by the P-FDK method can be regarded as combining the data of the same height projection on the curve passing through the OP in Fig.
  • the OP curve is a curve defined by the P-FDK rearrangement formula (1).
  • the process of rearranging data by the T-FDK method can be regarded as combining the data of the same height projection on the line passing through OO' in Fig. 2, that is, the data of T-FDK is obtained, that is, the result of formula (2).
  • the OO' curve is a curve defined by the T-FDK rearrangement formula (2).
  • the process of data rearrangement by the CT image reconstruction method (C-FDK for short) of the present invention can be regarded as combining the data of the same height projection on the curve passing through the OC in FIG. 2, that is, the data of the C-FDK is obtained.
  • the OC curve is an arc of a radius R, that is, an arc of the same curvature as the scanning circle, and the center of curvature of the OC is on a line passing through the O point and parallel to the SC.
  • the curvature of the 0C curve may be inaccurately equal to the curvature of the scanning circular orbital, and may be within the range of 50% to 200% of the curvature of the circular orbital, that is, the projection data within the range may also achieve the effect of the present invention.
  • c denotes the coordinate of the central virtual detector along the vertical direction after the rearrangement of the C-FDK method
  • S denotes the intersection position of the X-ray and the orbit to the center ray distance t
  • SC denotes the position at S The distance of the X-ray source point along the X-ray to the curve OC
  • SP denotes the distance of the X-ray source point located at S along the strip of X-rays to the curve OP.
  • Figure 3 is a vertical section taken along line SC of Figure 2, wherein the line at an angle to the axis of the SC is an X-ray.
  • the lengths of S (:, SP in Figure 3 can be calculated by the spatial geometric relationship, and the specific length is as follows:
  • t represents the distance between parallel fan beams
  • c represents the angular sampling interval in the Z-axis direction.
  • the next step is to weight the projection data, and the weighting coefficient may be cos, where each X-ray is projected onto the central plane along the Z-axis direction.
  • the angle between the projection lines as shown in Figure 3, according to the spatial geometry:
  • f c — FDK (x, y,z) ⁇ g c _ FDK ( ⁇ , t(x, y, ⁇ ), c(x, y, z, ⁇ )) ⁇ ( 11)
  • FIG. 5 (a) the projection data collected by the circular orbit CT scan is shown, which is composed of the central layer data of the flat panel detector at various angles.
  • Figure 5 (b) is a sinogram of the central layer data of the virtual detector at each angle after the projection data is rearranged by the C-FDK method of the present invention;
  • Figure 5 (c) is for Figure 5 ( b) Projection data obtained by performing weighting processing on the projection data shown. Then, the weighted processed projection data is subjected to one-dimensional ramp filtering, and finally, back projection is performed to obtain a CT image of the scanned object, as shown in Fig. 6 (d).
  • the CT image reconstruction apparatus of the present invention includes:
  • the rearrangement unit 1 is configured to select projection data of the same height on a curve that approximates the curvature of the circular orbit;
  • a weighting unit 2 configured to perform weighting processing on the selected projection data
  • a filtering unit 3 configured to filter the weighted projection data in a horizontal direction
  • the back projection unit 4 performs three-dimensional back projection on the filtered projection data in the ray direction.
  • the rearrangement unit 1 selects projection data according to the following formula: Wherein P - ( , c) represents the selected projection data; represents the projection direction; t represents the distance between the parallel fan beams; c represents the angle of the angle in the Z-axis direction; R represents the radius of the circular orbit.
  • the weighting unit 2 processes the selected projection data according to the following formula:
  • _ ( , represents filtered projection data
  • ® represents convolution
  • the back projection unit 4 performs three-dimensional back projection according to the following formula:
  • / - dish ( ⁇ represents the reconstructed image in the X-axis, y-axis, and x-axis directions.
  • the inventors of the present application conducted numerical simulation experiments using a three-dimensional Shepp-Logan head model, and conducted experimental comparisons with the FDK method and the T-FDK method to verify the technical solution of the present invention.
  • the three-dimensional head model is confined to a lmm sphere, and the center of the model is the rotation center of the CT scan.
  • the distance from the X source to the rotation center is 4 mm, and the distance from the X source to the detector is 8 mm.
  • the size is 4mmx4mm, the number of detection units is 256x256, 360 cone beam projections are evenly distributed in an angle of 360 degrees, and then 3D CT image reconstruction is performed.
  • the maximum cone angle of the cone beam X-ray is about 14 degrees, that is, the X-ray cone angle range of this experiment is ⁇ 14°.
  • the three-dimensional CT image reconstruction is performed by the FDK method, the T-FDK method and the C-FDK method of the present invention, respectively.
  • Figure 6 (b) It is the reconstruction result obtained by the FDK method;
  • Fig. 6 (c) is the reconstruction result obtained by the T-FDK method;
  • Fig. 6 (d) is the reconstruction result obtained by the C-FDK method of the present invention.
  • the C-FDK method of the present invention can be at ⁇ 14.
  • the image of the three-dimensional head model is well reconstructed, which overcomes the cone-beam artifact problem that is common in the large cone-angle circular orbit cone beam CT reconstruction, and solves the large cone-angle circular orbit CT well. Image reconstruction problem.
  • Figure 7 selects the section line along the vertical center of the section of Figure 6, where the abscissa indicates the length coordinate of the point on the section line, and the ordinate indicates the point on the section line.
  • the linear attenuation coefficient indicates the reconstruction result of the FDK method
  • the dotted line indicates the reconstruction result of the T-FDK method
  • the short dashed line indicates the reconstruction result of the C-FDK method of the present invention
  • the solid line indicates the accurate value of the head model.
  • the existing FDK and T-FDK methods have a rapid downward trend in the reconstruction value when the cone angle is increased, and are farther and farther away from the accurate value of the model, and the C-FDK of the present invention.
  • the method can reconstruct the accurate value of the original model more accurately even within the range of ⁇ 14° cone angle, which basically solves the problem of CT image reconstruction of large cone angle orbit.
  • the core of the C-FDK method of the present invention is the filtering of the projection data along the concave direction of the mid-line virtual detector.
  • the present invention is to X-ray and FIG.
  • the projections of the same height on the middle OC curve are rearranged into C-FDK data P c - (6 >, t, c ) of the same height c.
  • the OC curve is the same arc as the arc of the circular scanning track.
  • the selection of the OC curve is not unique, and a small range change can be made near the curvature, and it should also be within the scope of the claims of the present invention.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Image Processing (AREA)

Description

计算机断层成像 (CT) 图像重建方法及装置 技术领域
本发明总体上涉及 CT图像领域, 尤其涉及一种 CT图像重建方法及 装置。 背景技术
自从 1972年 Hounsfield发明了第一台 CT机, CT技术给医学诊断 和工业无损检测带来了革命性的影响, CT已经成为医疗、 生物、 航空 航天、 国防等行业重要的检测手段之一。 随着技术的进步, CT扫描模 式和成像方法也在不断地改进,三维锥束 CT已经成为研究和应用的主 流。 X射线锥束 CT已经在医学临床、 安全检查、 无损检测等领域得到 了广泛的应用。 特别是基于圆轨道扫描的锥束 CT系统由于其在机械、 电控等方面相对简单, 易于工程实现, 因此在临床、 安检和工业无损 检测应用非常广泛。 在圆轨道锥束 CT 重建方法中, 应用最广泛的是 1984年由 Feldkamp等人提出的 FDK方法 ( Feldkamp.L.A., L.C.Davis, and J.W.Kress. Practical cone-beam algorithm [J]. Journal of the Optical Society of America, 1984, (1):612-619. ) 。
FDK方法可以看作是扇束 FBP ( Filtered Backprojection, 滤波反投 影) 方法在三维情况下的近似扩展。 该 FDK方法 ¾括如下步骤: 首先 对投影数据进行加权处理; 然后对不同投影角度的投影数据进行水平 方向上的一维滤波; 最后沿 X射线逆方向进行三维反投影得到物体最 终的三维重建图像。
因此可知 FDK 方法的体素重建值是在 360·投影角度范围内通过 该体素的射线贡献之和,所以圆轨道锥束 FDK方法作为一种近似方法, 具有数学公式简单、 方法计算快速、 易于工程实现等特点, 并且当锥 角比较小的时候 (一般在 ± 5°以内) , 能够取得很好的重建效果, 因 此在实际工程应用中得到了广泛认可。
但是, FDK方法也存在一定的问题: 由于圆轨道扫描本身不满足 锥束精确重建的数据完备性条件, 存在雷当 (Radon )数据缺失问题; 因此, 当锥束锥角增大时, FDK方法重建结果会产生严重的锥束伪影, 在远离扫描轨道平面内 FDK重建数值快速下降, 使得该方法.在平板探 测器 CT成像系统中的应用受到了很大限制。
为了改善大锥角下圆轨道 FDK方法的图像重建质量, 在 FDK方 法的基础上,提出了多种改进的 FDK方法,例如 P-FDK( parallel FDK )、 T-FDK( tent-FDK )、 HT-FDK( hybrid tent-FDK )、 EFDK( extended FDK ) 等。 在这些改进的 FDK类方法中, 由于 P-FDK和 T-FDK方法简单, 易于在工程中实现, 因此得到了较为广泛的应用, 下面简单介绍一下 这两种方法。
P-FDK 方法是将锥束投影数据通过重排获得平行的扇束投影数 据, 然后再通过滤波反投影方法重建出物体的三维图像。 如图 1 ( a ) 所示, 其是使用平板探测器的圆轨道锥束 CT系统扫描示意图, SC ^表 示 X光源在圆轨道上的位置, 表示投影在圆轨道上的角度采样位置。
( ?, α,δ)表示平板探测器上的投影数据, (", W是定义在平板探测器上的 直角坐标系, 用于表示每条 X射线投影在探测器上的位置坐标, R是 圆轨道半径; 图 (b )是利用 P-FDK方法把锥束投影重排成平行束的扇 束投影, 黑色实心点为 X光源, 在中心虚拟探测器上, 重排后的投影 数据由于同一角度下平行扇束到虚拟探测器的距离不相同, 因此原平 板探测器上一行投影对应到该虚拟探测器不再是水平的, 而是在一段 沿虚拟探测器中心行外凸的曲线上。 以平板探测器为例, P-FDK 的数 据重排公式如公式( 1 )。 本发明后续推导过程也皆以平板探测器为例, 其他探测器类型例如柱面探测器等, 都可以根据平板探测器做相应变 形得到, 不做赘述。 ( 1 )
Figure imgf000004_0001
其中, 表示经过 P-FDK重排完后, 其重排投影数据在角度方向 的采样位置。 (^)是经过重排后在中心虚拟探测器上的直角坐标系, 表 示重排后每条 X射线在虛拟探测器上的位置坐标。
P-FDK方法和 FDK方法的不同之处, 只是由于 P-FDK方法先重 排成平行的扇束使得在反投影时省去了计算加权系数的过程, 在图像 质量上和 FDK方法没有区别。 不同于 P-FDK, 2000年由 Grass等人提 出的 T-FDK方法在 P-FDK方法基础上做了改进, T-FDK除了将锥束 投影重排成平行的扇束以外, 还在竖直扇束平面内进行二次重排, 即
T-FDK 需要在水平和竖直两个方向上对投影数据进行重排, 最终使得 T-FDK和 P-FDK的差别在于 T-FDK对投影数据的滤波方向是沿中心 虛拟探测器的水平方向进行的, 如图 1 ( c ) 所示, 而不是沿外凸的曲 线方向。 T-FDK方法的数据重排公式如下,
PT~FDK (θ, t, s) = Ρ(θ - arcsin -, . tR , Ύ) ( 2 ) 其中, 表示经过 T-FDK重排完后, 其重排投影数据在角度方向 的釆样位置。 (^)是经过 T-FDK重排后在中心虛拟探测器上的直角坐 标系, 表示重排后每条 X射线在虛拟探测器上的位置坐标。
T-FDK相比较 FDK来说, 一方面类似于 P-FDK由于重排成了平 行的扇束省去了反投影时的加权系数, 因此在计算方面更加有效率; 同时, 由于 T-FDK方法对投影数据的滤波是沿中心虚拟探测器的水平 方向进行的, 减轻了锥角增大时导致的锥束伪影, 提高了图像重建的 质量, 使得利用大面积平板探测器通过圆轨道扫描实现大物体的准确 三维成像成为可能。 另外还有一类改进 FDK方法的思路是利用圆轨道 扫描投影中的共轭射线, 通过对共轭投影选取不同的反投影权重系数 来改善重建图像质量, 也取得了较好了效果。
随着平板探测器的日益普及, 使用大面积平板探测器的新型锥束 CT系统越来越多, 对大锥角圆轨道锥束 CT图像重建方法的要求也越 来越高。以目前在牙科疾病诊断中应用较为广泛的牙科锥束 CT设备为 例, 目前多个厂家的三维牙科 CT使用了 20cm X 25cm平板探测器, X 光源到探测器的距离 设为 70cm , 此时圆轨道扫描对应的锥角大小为 ± 8.13。,由于医生主要依靠图像的 CT数即重建图像像素的数值大小进 行疾病诊断, 因此, 医用 CT对图像的重建数值要求非常高; 如此大锥 角的圆轨道扫描已经远远超出了 FDK方法能够重建的范围, T-FDK重 建结果也没法让人满意。 并且, 随着探测器技术的飞速发展, 更大面 积的平板探测器已经在临床得到了应用, 例如 30cm x 40cm、 43cm χ 43cm 等多种不同尺寸的平板探测器都已经在临床 DR ( Digital Radiography System, 数字式 X射线摄影技术)成像中应用了。 这些更 大面积的探测器能够大大提高锥束 X射线的有效探测面积, 扩大成像 视野,最重要的是能够减少甚至消除锥束投影数据截断导致的 CT数值 无法准确重建的问题; 因此, 大面积平板探测器在目前和未来的三维
CT影像设备中具有极为广阔的应用空间。 但是, 随着平板探测器面积 的增大, 三维 CT系统的锥角也相应增大, 如何消除大锥角下的严重的 锥束伪影成为一个必须解决的难题。 发明内容
本发明要解决的主要技术问题是提供一种能够消除大锥角下严重 锥束伪影的 CT图像重建方法及装置。
为了解决上述问题, 本发明 CT 图像重建方法的技术方案包括步 骤:
选取与扫描圆轨道近似曲率的曲线上的同高度的投影数据; 对所选取的投影数据进行加权处理;
对经过加权处理的投影数据沿水平方向进行滤波;
对滤波后的投影数据沿射线方向进行三维反投影。
所述步骤选取与扫描圆轨道近似曲率的曲线上的同高度的投影数 据包括按下列公式选取投影数据:
PC~FDK (Θ, t, c) = Ρ(θ
Figure imgf000006_0001
其中, ― )表示所选取的投影数据; 表示投影方向; t表示 平行扇束之间的距离; c表示在 Z轴方向上的角度釆样间隔; R表示圆 轨道的半径。
所述步骤对所选取的投影数据进行加权处理包括按下列公式对所 选取的投影数据进行处理:
Figure imgf000006_0002
其中, φ, ) 表示经过加权处理的投影数据。
所述步骤对经过加权处理的投影数据沿水平方向进行滤波包括按 下列公式进行滤波: gc-FDK(0,t,c) = Pc - FDK(e,t,c) h(t)
其中, ge— ""^,^)表示经过滤波的投影数据; ®表示卷积; 表 示滤波函数。
所述步骤对滤波后的投影数据沿射线方向进行三维反投影包括按 下列公式进行三维反投影:
fc~FDK ( , y,z)= \ gc~FDK (Θ, t(x, y, Θ), c(x, y, z, θ)) άθ
0
其中, 皿^,^)表示在 χ轴, y轴 ,z轴方向上的重建图像。
相应地, 本发明 CT图像重建装置的技术方案包括:
重排单元, 用于选取与扫描圆轨道近似曲率的曲线上的同高度的 投影数据;
加权单元, 用于对所选取的投影数据进行加权处理;
滤波单元, 用于对经过加权处理的投影数据沿水平方向进行滤波; 反投影单元, 对滤波后的投影数据沿射线方向进行三维反投影。 所述重排单元按下列公式选取投影数据:
Pc'FDK(0,i,c) = P(e
Figure imgf000007_0001
^C-FDK
其中, P - ( 'c)表示所选取的投影数据; 表示投影方向; 表示 平行扇束之间的距离; c表示在 ζ轴方向上的角度采样间隔; R表示圆 轨道的半径。
所述加权单元按下列公式对所选取的投影数据进行处理:
Figure imgf000007_0002
其中, ?c-舰 (θ, )表示经过加权处理的投影数据 ,
所述滤波单元按下列公式进行一维斜坡滤波: g"-^ (θ, t, c) = PC-FDK φ, t, c) ® hit) : ^ p^-^^ey tj C . h t - tr)dr 其中, 皿 ( ,e)表示经过滤波的投影数据; ®表示卷积; 表 示滤波函数。
所述反投影单元按下列公式进行三维反投影:
f-FDK{x, y, z) = ζ, θ))άθ
Figure imgf000008_0001
其中, 皿 <, ^ )表示在 X轴, y轴 ,Z轴方向上的重建图像。
与现有技术相比, 本发明 CT图像重建方法及装置的有益效果为: 首先, 本发明对投影数据进行重排是采用选取与扫描圆轨道近似 曲率的曲线上的同高度的投影数据, 从而在平行于 Z轴的扇束平面内 沿虚拟中心探测器向中心行内凹的曲线上采样, 这样大大提高了在大 锥角下重建方法的数值精度, 有效地抑制了大锥角带来的锥束伪影。
另外, 本发明执行效率高, 稳定性强。 附图说明
为了对本公开内容更透彻的理解, 下面参考结合附图所进行的下 列描述, 在附图中:
图 1(a)是平板探测器圆轨道锥束 CT扫描示意图;
图 1(b)是采用 P-FDK方法对图 1(a)所得投影数据重排后的示意图; 图 1(c)是采用 T-FDK方法对图 1(a)所得投影数据重排后的示意图; 图 1(d)是采用本发明 CT图像重建方法对图 1(a)所得投影数据重排 后的示意图;
图 2是图 l(a)、 1(b), 1(c) 和 1(d)沿旋转轴 Z轴的俯视图; 图 3是图 2过 SC的竖直切面的侧视图;
图 4是本发明 CT图像重建方法的流程图;
图 5(a)示出了圆轨道 CT扫描所采集的投影数据, 其是由在各个角 度下平板探测器中心层数据组成的正弦图;
图 5(b)是经过本发明 C-FDK方法对投影数据进行重排后, 各个角 度下虚拟探测器中心层数据组成的正弦图; 图 5(c)是对图 5(b)所示的投影数据进行加权处理后所得到的投影 数据的示意图;
图 6(a) 是三维 Shepp-logan头模型在一竖直截面的准确图像的示 意图;
图 6(b)是釆用 FDK方法所得到的重建结果的示意图 ·;
图 6(c)是采用 T-FDK方法所得到的重建結果的示意图;
图 6(d)是采用本发明 C-FDK方法所得到的重建结果的示意图; 图 7是选取沿图 6(a)、 图 6(b) 、 图 6(c) 和图 6(d)截面的竖直中心 的剖面线后重建结果的示意图;
图 8是本发明 CT图像重建装置的示意图。 具体实施方式
下面将详细描述本发明的具体实施例, 但本发明并不限于下述具 体实施例。
如图 4所示, 本发明 CT图像重建方法包括步骤:
1 )选取与扫描圆轨道近似曲率的曲线上的同高度的投影数据;
2 )对所选取的投影数据进行加权处理;
3 )对经过加权处理的投影数据沿水平方向进行滤波;
4 )对滤波后的投影数据沿射线方向进行三维反投影。
从上述可知,本发明 CT图像重建方法进行重排的方式是选取与扫 描圆轨道近似曲率的曲线上的同高度的投影数据, 接着对所选取的投 影数据进行加权处理, 然后将经过加权处理的投影数据沿水平方向进 行滤波, 最后再对滤波后的投影数据沿射线方向进行三维反投影即可 得到被扫描物体的三维 CT图像。
对于与扫描轨道曲率近似的曲率而言,假设扫描圆轨道曲率为 1/R , 该近似平均曲率范围可以是 1/ 2R〜2R。对于滤波,可以采用本领域技术 人员可以获知的任何滤波方法, 例如, 滤波核可以是最基本的斜坡滤 波核, 也可以是对标准斜坡滤波在频域进行过平滑处理后的滤波核, 比如常用的 S-L滤波核等 ( S-L滤波核由 L.A.Shepp和 B.F.Logan于 1974 年提出) 。
本说明书中是以平板探测器为例来描述本发明的技术方案的。 当 然, 本发明可以应用到诸如柱面探测器之类的其它面阵探测器。 如图 1 ) 所示, 首先, 定义平面内锥束 X射线源的扫描路径为 一个圆周 C^ - R^osAsin ) ; R是圆轨道半径; 表示光源点对应的角 度参量; O是坐标原点和圆轨道的中心, 也就是旋转中心。 ( ?,α, έ)表 示 X射线照射被扫描物体后在平板探测器 1上采集的投影数据, 其中 (α, >)表示二维平板探测器 1上某个投影点位置的横纵坐标。
根据 P-FDK方法的性质, 该方法对锥束投影数据进行重排后选取 的是在中心虛拟探测器 2上在外凸的曲线上的投影数据, 如图 1 ( b ) 所示。 下面请参见图 2 , 图 2是沿旋转轴 Z轴从上往下的俯视图, 此时 面阵列探测器成为一条直线或曲线, 圆轨道上的实心点 S表示 X射线 光源。 在图 2中, P-FDK方法重排数据的过程可以看成是将图 2中经 过 OP的曲线上的同高度投影数据进行组合, 即得到 P-FDK的数据, 也 就是公式( 1 )的结果, 其中 OP曲线即为由 P-FDK重排公式( 1 )定义 的一条曲线。 而 T-FDK方法重排数据的过程可以看成是将图 2中经过 OO'的直线上的同高度投影数据进行组合, 即得到了 T-FDK的数据, 也就是公式(2 ) 的结果, 其中 OO'曲线即为由 T-FDK重排公式 (2 ) 定义的一条曲线。
而本发明 CT 图像重建方法 (简称 C-FDK ) 进行数据重排的过程 则可以看成是将图 2中经过 OC的曲线上的同高度投影数据进行组合, 即得到了 C-FDK的数据, 本例中 OC曲线是以半径为 R的一段圆弧, 也 就是和扫描圆轨道相同曲率的一段弧线, 且 OC的曲率中心在过 O点且 平行于 SC的直线上。 当然, 0C曲线的曲率可以不准确等于扫描圆轨道 的曲率, 可以在圆轨道曲率的 50%〜200%范围之内, 也即在此范围内 的投影数据同样可以达到近似本发明的效果。
本发明 C-FDK方法对投影数据的重排是在中心虚拟探测器的水平 和竖直两个方向上完成, 在水平方向与 P-FDK以及 T-FDK是相同的; 在竖直方向上, 如图 3所示, 对重排后的数据: , , 、 c = b K 5 )
Figure imgf000010_0001
其中, c表示 C-FDK方法重排后中心虚拟探测器沿竖直方向的坐 标, S表示角度为 且到中心射线距离为 t的 X射线与轨道的交点位置, 因此, SC表示位于 S处的 X射线源点沿该条 X射线到曲线 OC的距离。 SP 表示位于 S处的 X射线源点沿该条 X射线到曲线 OP的距离。
图 3是沿图 2中过 SC的竖直切面, 其中, 与 SC所在的轴成 角的 线为 X射线。 根据图 2中 OC弧线的定义, 图 3 中 S (:、 SP的长度都可 以通过空间几何关系计算得到, 其具体长度如下:
Figure imgf000011_0001
因此, 本发明 C-FDK方法中对圆轨道锥束 CT投影数据的重排公 式:^下: ( 7 )
Figure imgf000011_0002
其中, 表示投影方向, t表示平行扇束之间的距离, c表示在 Z 轴方向上的角度采样间隔。
利用上面公式 (7 ) 完成 C-FDK 方法的数据重排后, 下一步是对 投影数据进行加权处理, 加权系数可以为 cos , 其中 是每条 X射线与 其沿 Z轴方向投影到中心平面上的投影线之间的夹角, 如图 3所示, 根据空间几何关系:
Figure imgf000011_0003
即对投影进行加权处理的公式为 ( 9 )
Figure imgf000011_0004
然后, 对完成上述处理的投影数据沿水平方向进行一维斜坡滤波: gc-mK φ, t, c) = PC-FDK (θ, t, c) ® h(t)
¾ ( in ) 上式中, ®表示卷积, 而; t)是滤波函数, 一般采用 Ramp滤波器。 最后, 将滤波后的投影数据沿射线方向进行三维反投影即可重建 出被扫描物体的三维 CT图像, 反投影公式如下:
fcFDK (x, y,z)= \ gc_FDK (Θ, t(x, y, Θ), c(x, y, z, θ)) άθ ( 11)
0
其中, 探测器上的投影位置计算公式如下 t(x, }>,θ) = y cos ^ - x sin ^ ( 12)
z-{2^R2 -f -R)
c(x,y,z,0) ( 13) 如图 5 (a) 所示, 示出了圆轨道 CT扫描所采集的投影数据, 其 疋由在各个角度下平板探测器中心层数据组成的正弦图; 图 5 (b) 则 是经过本发明 C-FDK方法对投影数据进行重排后, 各个角度下虚拟探 测器中心层数据组成的正弦图; 图 5 (c) 则是对图 5 (b) 所示的投影 数据进行加权处理后所得到的投影数据。 接着对经过加权处理的投影 数据进行一维斜坡滤波, 最后, 再进行反投影即可得到被扫描物体的 CT图像, 如图 6 (d) 所示。
相应地, 如图 8所示, 本发明 CT图像重建装置包括:
重排单元 1,用于选取与扫描圆轨道近似曲率的曲线上的同高度的 投影数据;
加权单元 2, 用于对所选取的投影数据进行加权处理;
滤波单元 3, 用于对经过加权处理的投影数据沿水平方向进行滤 波;
反投影单元 4, 对滤波后的投影数据沿射线方向进行三维反投影。 优选地, 所述重排单元 1按下列公式选取投影数据:
Figure imgf000012_0001
其中, P - ( ,c)表示所选取的投影数据; 表示投影方向; t表示 平行扇束之间的距离; c表示在 Z轴方向上的角度釆样间隔; R表示圆 轨道的半径。
优选地, 所述加权单元 2按如下公式对所选取的投影数据进行处 理:
PC— (θ, ί, C) = - ' - . pC-FDK φ, t, c)
V5R2 - 4t2 - 4R /R2 - t2 + c2
其中, ( , c)表示经过加权处理的投影数据。
优选地, 所述滤波单元 3按如下公式进行滤波: gd (Θ, t, c) = PC-FDK (Θ, t, c) ® hit)
其中, _ ( , 表示经过滤波的投影数据; ®表示卷积; 表 示滤波函数。
优选地, 所述反投影单元 4按如下公式进行三维反投影:
Figure imgf000013_0001
其中, /—皿(^ 表示在 X轴, y轴 ,ζ轴方向上的重建图像。
由于本发明 CT 图像重建装置的技术方案所包含的技术特征与本 发明 CT重建方法的技术方案所包含的技术特征相对应,因此不再对本 发明 CT图像重建装置的技术方案进行详细描述。
本申请的发明人利用三维 Shepp-Logan 头模型进行了数值模拟实 验,并与 FDK方法和 T-FDK方法进行了实验对比来验证本发明的技术 方案。 在数值模拟实验中, 三维头模型被限制在 lmm的球体内, 模型 中心即为 CT扫描的旋转中心,. X光源到旋转中心距离为 4mm, X光 源到探测器的距离为 8mm, 平板探测器尺寸为 4mmx4mm, 探测单元 数量为 256x256 , 在 360度范围内角度均匀釆集 360个锥束投影, 然后 进行三维 CT图像重建。 根据上述几何定义, 可以计算出在此实验中, 锥束 X射线的最大锥角约为 14度, 即本次实验的 X射线锥角范围为 ±14°。使用上述扫描条件下的圆轨道锥束 CT投影数据,分别采用 FDK 方法、 T-FDK方法和本发明 C-FDK方法进行三维 CT图像重建。
图 6 ( a )是三维 Shepp-logan头模型在一竖直截面的准确图像, 所 述竖直截面是三维头模型在 y=-0.25处,沿平行于 x-z平面予以选取的; 图 6 ( b )是采用 FDK方法所得到的重建结果; 图 6 ( c )是采用 T-FDK 方法所得到的重建结果; 图 6 ( d )是采用本发明 C-FDK方法所得到的 重建结果。
从上述所得到的各重建图像可以看出: 本发明 C-FDK方法能够在 ±14。条件下, 很好地重建出三维头模型的图像, 克服了现有方法在大 锥角圆轨道锥束 CT重建中普遍存在的锥束伪影难题,很好地解决了大 锥角圆轨道 CT图像重建问题。
为了更准确地分析上述重建结果的数值准确性, 图 7 选取沿图 6 截面的竖直中心的剖面线, 其中, 横坐标表示该剖面线上点的长度坐 标, 纵坐标表示该剖面线上点的线性衰减系数, 长虚线表示 FDK方法 的重建结果, 点划线表示 T-FDK方法的重建结果, 短虚线表示本发明 C-FDK方法的重建结果, 实线表示头模型的准确数值。 从该剖面线可 以更清楚地看到:现有的 FDK和 T-FDK方法当锥角增大时其重建数值 呈快速下降趋向, 离模型的准确值越来越远, 而本发明 C-FDK方法即 使在 ± 14°锥角范围内仍然能够较为准确地重建出原始模型的准确值, 基本解决了大锥角圆轨道 CT图像重建的难题。
最后, 需要特别说明的是: 本发明 C-FDK方法的核心之处在于沿 中线虛拟探测器内凹的方向上进行投影数据的滤波, 在上述推导过程 中, 本发明是将 X射线和图 2中 OC曲线上同高度的投影重排为同一高 度 c的 C-FDK数据 Pc-皿 (6>, t, c)。 而 OC曲线则是和圆扫描轨道弧度相同 的一段圆弧, 此处 OC曲线的选取不是唯一的, 可以在该曲率附近做小 范围的改变, 也同样应该在本发明专利的权利要求范围内。
虽然上述已经结合附图描述了本发明的具体实施例, 但是本领域 技术人员在不脱离本发明的精神和范围的情况下, 可以对本发明进行 各种改变、 修改和等效替代。 这些改变、 修改和等效替代都意为落入 随附的权利要求所限定的精神和范围之内。

Claims

权 利 要 求
1. 一种 CT图像重建方法, 其特征在于, 包括步骤:
选取与扫描圆轨道近似曲率的曲线上的同高度的投影数据; 对所选取的投影数据进行加权处理;
对经过加权处理的投影数据沿水平方向进行滤波;
对滤波后的投影数据沿射线方向进行三维反投影。
2. 如权利要求 1所述的 CT图像重建方法, 其特征在于, 所述步 骤选取与扫描圆轨道近似曲率的曲线上的同高度的投影数据包括按下 列公式选取投影数据:
PcmK (Θ, t, c) = Ρ{θ一 arcsin - , . tR , ι )
R R2- 2 2(R2-t2)- R2-/2
其中, 皿 ( ,c)表示所选取的投影数据; 表示投影方向; ,表示 平行扇束之间的距离; c表示在 Ζ轴方向上的角度采样间隔; R表示圆 轨道的半径。
3. 如权利要求 2所述的 CT图像重建方法, 其特征在于, 所述步 骤对所选取的投影数据进行加权处理包括按下列公式对所选取的投影 数据进行处理:
Figure imgf000015_0001
其中, 表示经过加权处理的投影数据。
4. 如权利要求 3所述的 CT图像重建方法, 其特征在于, 所述步 骤对经过加权处理的投影数据沿水平方向进行滤波包括按下列公式进 行滤波: gc-FOK(9,t,c) = Pc-FDK(e,t,c)®h(t)
Figure imgf000015_0002
其中, ^― W表示经过滤波的投影数据; ®表示卷积; 表 示滤波函数。
5. 如权利要求 4所述的 CT图像重建方法, 其特征在于, 所述步 骤对滤波后的投影数据沿射线方向进行三维反投影包括按下列公式进 行三维反投影:
广厘 (X, y, z) = \ g c -皿 (Θ, t(x, y, Θ), c(x, y, z, θ)) άθ
0
其中, fc—置 (x, y 表示在 X轴, y轴 ,Z轴方向上的重建图像。
6. 一种 CT图像重建装置, 其特征在于, 包括:
重排单元, 用于选取与扫描圆轨道近似曲率的曲线上的同高度的 投影数据;
加权单元, 用于对所选取的投影数据进行加权处理;
滤波单元, 用于对经过加权处理的投影数据沿水平方向进行滤波; 反投影单元, 对滤波后的投影数据沿射线方向进行三维反投影。
7. 如权利要求 6所述的 CT图像重建装置, 其特征在于, 所述重 排单元按下列公式选取投影数据:
Figure imgf000016_0001
其中, ―皿 )表示所选取的投影数据; 表示投影方向; t表 示平行扇束之间的距离「 c表示在 Z轴方向上的角度釆样间隔; R表示 圆轨道的半径。
8. 如权利要求 7所述的 CT图像重建装置, 其特征在于, 所述加 权单元按下列公式对所选取的投影数据进行处理: _ φ, t, c)
Figure imgf000016_0002
其中, 服 ( )表示经过加权处理的投影数据。
9. 如权利要求 8所述的 CT图像重建装置, 其特征在于, 所述滤 波单元按下列公式进行一维斜坡滤波:
gc-皿 c -皿 φ, t, c) ® h{t)
Figure imgf000016_0003
其中, 皿 表示经过滤波的投影数据; ®表示卷积; 表 示滤波函数。
10. 如权利要求 9所述的 CT图像重建装置, 其特征在于, 所述反 投影单元按下列公式进行三维反投影: fc-FDK(x,y,z) = ]gc-FDK {θ, χ,γ,θ) χ,γ,ζ,θ))άθ
0
其中, /c-皿 , ζ) 表示在 χ轴, y轴 ,ζ轴方向上的重建图像。
PCT/CN2011/000830 2010-10-28 2011-05-11 计算机断层成像(ct)图像重建方法及装置 WO2012055147A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP11743411.8A EP2469472A4 (en) 2010-10-28 2011-05-11 Computed tomography (ct) image reconstruction method and apparatus
US13/140,761 US8724889B2 (en) 2010-10-28 2011-05-11 Method and apparatus for CT image reconstruction

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201010529974.0A CN102456227B (zh) 2010-10-28 2010-10-28 Ct图像重建方法及装置
CN201010529974.0 2010-10-28

Publications (1)

Publication Number Publication Date
WO2012055147A1 true WO2012055147A1 (zh) 2012-05-03

Family

ID=45626044

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/000830 WO2012055147A1 (zh) 2010-10-28 2011-05-11 计算机断层成像(ct)图像重建方法及装置

Country Status (4)

Country Link
US (1) US8724889B2 (zh)
EP (1) EP2469472A4 (zh)
CN (1) CN102456227B (zh)
WO (1) WO2012055147A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9858690B2 (en) 2013-10-11 2018-01-02 National Yang-Ming University Computed tomography (CT) image reconstruction method

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102012217163B4 (de) 2012-09-24 2022-06-02 Siemens Healthcare Gmbh Verfahren zur Rekonstruktion von CT-Bilddaten mit gewichteter Rückprojektion, einschließlich Recheneinheit und CT-System für dieses Verfahren
CN103366389A (zh) * 2013-04-27 2013-10-23 中国人民解放军北京军区总医院 Ct图像重建方法
CN103871064B (zh) * 2014-03-25 2017-02-08 中国石油大学(华东) 一种火山岩ct图像的预处理和确定分割阈值的方法
CN105096357A (zh) * 2014-05-19 2015-11-25 锐珂(上海)医疗器材有限公司 层析成像的重建方法
CN105488823B (zh) * 2014-09-16 2019-10-18 株式会社日立制作所 Ct图像重建方法、ct图像重建装置及ct系统
CN106204679B (zh) * 2016-07-18 2019-04-09 上海交通大学 基于可分离足迹函数技术的投影方法、装置及系统
WO2019128660A1 (zh) * 2017-12-29 2019-07-04 清华大学 训练神经网络的方法和设备、图像处理方法和设备以及存储介质
CN108283502B (zh) * 2018-02-12 2021-05-25 沈阳晟诺科技有限公司 一种焦点移动式ct机、扫描方法及图像重建方法
CN108283503B (zh) * 2018-02-12 2021-08-20 沈阳晟诺科技有限公司 一种ct机、扫描方法及图像重建方法
CN108720863B (zh) * 2018-02-12 2021-06-01 沈阳晟诺科技有限公司 一种焦点切换式ct机、扫描方法及图像重建方法
CN110660123B (zh) * 2018-06-29 2021-08-31 清华大学 基于神经网络的三维ct图像重建方法和设备以及存储介质
WO2020210941A1 (zh) * 2019-04-15 2020-10-22 清华大学 分布式光源ct图像重建方法与系统
CN111265231B (zh) * 2019-04-15 2021-08-31 清华大学 分布式光源ct图像重建方法与系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030161444A1 (en) * 2001-08-16 2003-08-28 Alexander Katsevich Method of reconstructing images for spiral and non-spiral computer tomography
US20080013676A1 (en) * 2004-06-09 2008-01-17 Koninklijke Philips Electronics N.V. Computerized Tomography Method With Helical Relative Movement and Conical Beam
CN101404088A (zh) * 2008-11-05 2009-04-08 华中科技大学 Ct图像重建的方法及系统

Family Cites Families (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5406479A (en) * 1993-12-20 1995-04-11 Imatron, Inc. Method for rebinning and for correcting cone beam error in a fan beam computed tomographic scanner system
US7583777B2 (en) * 2004-07-21 2009-09-01 General Electric Company Method and apparatus for 3D reconstruction of images
WO2006073584A2 (en) * 2004-11-24 2006-07-13 Wisconsin Alumni Research Foundation Cone-beam filtered backprojection image reconstruction method for short trajectories
US8433033B2 (en) * 2005-10-21 2013-04-30 Axion Japan Co., Ltd. Panoramic imaging apparatus
JP4675753B2 (ja) * 2005-11-11 2011-04-27 ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー X線ct装置
US7590216B2 (en) * 2005-12-28 2009-09-15 University Of Central Florida Research Foundation, Inc. Cone beam local tomography
US7269244B2 (en) * 2006-01-25 2007-09-11 General Electric Company Methods and apparatus for generating thick images in cone beam volumetric CT
CN101594825B (zh) * 2006-02-27 2014-07-23 罗切斯特大学 锥束ct动态成像的方法和设备
US20090238412A1 (en) * 2006-06-28 2009-09-24 Michael Grass Local motion compensated reconstruction of stenosis
US8031829B2 (en) * 2007-10-26 2011-10-04 General Electric Company Method for analytic reconstruction of cone-beam projection data for multi-source inverse geometry CT systems
US8131042B2 (en) * 2008-02-08 2012-03-06 General Electric Company Methods and apparatus for hybrid cone beam image reconstruction
JP5159543B2 (ja) * 2008-09-30 2013-03-06 株式会社東芝 X線ct装置
JP5305821B2 (ja) * 2008-10-10 2013-10-02 株式会社東芝 医用画像処理装置及び医用画像診断装置
US8913805B2 (en) * 2010-08-30 2014-12-16 The Regents Of The University Of Michigan Three-dimensional forward and back projection methods

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030161444A1 (en) * 2001-08-16 2003-08-28 Alexander Katsevich Method of reconstructing images for spiral and non-spiral computer tomography
US20080013676A1 (en) * 2004-06-09 2008-01-17 Koninklijke Philips Electronics N.V. Computerized Tomography Method With Helical Relative Movement and Conical Beam
CN101404088A (zh) * 2008-11-05 2009-04-08 华中科技大学 Ct图像重建的方法及系统

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
FELDKAMP L.A., L.C.DAVIS, J.W.KRESS: "Practical cone-beam algorithm [J", JOURNAL OF THE OPTICAL SOCIETY OF AMERICA, 1984, pages 612 - 619
See also references of EP2469472A4
ZHANG JIAN ET AL.: "Review of recent development in FDK reconstruction algorithms for 3D cone beam CT.", CHINESE JOURNAL OF STEREOLOGY AND IMAGE ANALYSIS., vol. 10, no. 2, June 2005 (2005-06-01), pages 116 - 121, XP008169045 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9858690B2 (en) 2013-10-11 2018-01-02 National Yang-Ming University Computed tomography (CT) image reconstruction method

Also Published As

Publication number Publication date
EP2469472A4 (en) 2018-02-14
CN102456227A (zh) 2012-05-16
US8724889B2 (en) 2014-05-13
CN102456227B (zh) 2015-05-27
EP2469472A1 (en) 2012-06-27
US20120106832A1 (en) 2012-05-03

Similar Documents

Publication Publication Date Title
WO2012055147A1 (zh) 计算机断层成像(ct)图像重建方法及装置
JP5019193B2 (ja) スキャン対象に関するct画像内の再構成点における画像データ値を決定する再構成方法及びx線コンピュータ断層撮影装置
US7203272B2 (en) Cone-beam filtered backprojection image reconstruction method for short trajectories
US7251307B2 (en) Fan-beam and cone-beam image reconstruction using filtered backprojection of differentiated projection data
US6574297B2 (en) System and method for image reconstruction in a cone beam imaging system
WO2004019279A2 (en) Apparatus and method for reconstruction of volumetric images in a divergent scanning computed tomography system
EP1994504A1 (en) Method of reconstructing an image function from radon data
JP4557321B2 (ja) 画像再構成装置
JP4342164B2 (ja) コンピュータ断層撮影装置
JP3787374B2 (ja) 断層撮影画像内のアーチファクトを低減するための方法および装置
Bruder et al. Single-slice rebinning reconstruction in spiral cone-beam computed tomography
WO2011011684A2 (en) Selection of optimal views for computed tomography reconstruction
JP2003180676A (ja) コンピュータトモグラフィ用の像再構成方法
CN107233105B (zh) 一种用于ct图像重建的修正方法及修正系统
Park et al. A fully GPU-based ray-driven backprojector via a ray-culling scheme with voxel-level parallelization for cone-beam CT reconstruction
JP2004113785A (ja) コンピュータ断層撮影法における画像形成方法およびこの方法を実施するためのct装置
US20100232663A1 (en) Computed tomography reconstruction for two tilted circles
WO2007004196A2 (en) Exact fbp type algorithm for arbitrary trajectories
Kong et al. Analytic reconstruction approach for parallel translational computed tomography
US6999550B2 (en) Method and apparatus for obtaining data for reconstructing images of an object
Zou et al. Dual helical cone-beam CT for inspecting large object
JPWO2007096936A1 (ja) 断層撮影装置および演算処理プログラム
JP2005522249A (ja) コンピュータ断層撮影方法
Hoppe et al. Accurate image reconstruction using real C-arm data from a Circle-plus-arc trajectory
Zeng et al. BPF-type reconstruction for dual helical cone-beam CT

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 2011743411

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 13140761

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11743411

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE