Medical data processing device, X-ray CT device, medical data processing method and program

The medical data processing device improves X-ray CT image quality by employing a processing circuit that adapts to volume fraction conditions, using the Mendonça method when valid and alternative methods when not, allowing for robust multi-material discrimination.

JP7578510B2Active Publication Date: 2024-11-06UNIVERSITY OF CHICAGO +1
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
JP2021031529
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-04-03
Filing Date
2021-03-01
Publication Date
2024-11-06
Estimated Expiration
2041-03-01

AI Technical Summary

Technical Problem

Existing material discrimination methods in X-ray CT devices, such as the Mendonça method, are limited to decomposing images into two material components and fail to provide sufficient image quality when volume fraction conditions are not met.

Method used

A medical data processing device and method that includes a processing circuit to perform multi-material discrimination by decomposing attenuation values of voxels into multiple material components, using the Mendonça method when volume fraction conditions are valid, and switching to the shortest Hausdorff or closest edge method when they are not, ensuring robust image decomposition.

Benefits of technology

Enhances image quality by enabling decomposition into three or more material components, maintaining image quality even when volume fraction conditions are violated, and providing accurate material discrimination in X-ray CT devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007578510000012
    Figure 0007578510000012
  • Figure 0007578510000013
    Figure 0007578510000013
  • Figure 0007578510000014
    Figure 0007578510000014
Patent Text Reader

Abstract

To improve image quality.SOLUTION: A medical data processing device according to an embodiment includes a processing circuit. The processing circuit acquires data corresponding to a plurality of energy components, performs substance discrimination of the data, and decomposes attenuation values of the respective plurality of energy components of voxel of the data into a plurality of substance components. The processing circuit (1) determines whether the attenuation values of the voxel match a volume fraction condition, (2) decomposes the attenuation values into n substance components with an n as a natural number 3 or larger when the attenuation values of the voxel match the volume fraction condition, (3) generates a point projected by projecting the attenuation values of the voxel to an (n-2) dimensional space of (n-1) substance components nearest the attenuation values of the voxel when the attenuation values of the voxel do not match the volume fraction condition, and decomposes the respective attenuation values into the plurality of substance components by decomposing the attenuation value of the projected point into the (n-1) substance components.SELECTED DRAWING: Figure 5
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The embodiments disclosed in this specification and the drawings relate to a medical data processing device, an X-ray CT device, a medical data processing method, and a program. [Background technology]

[0002] In an X-ray CT apparatus that acquires data corresponding to a plurality of energy components, such as a spectral CT, there is a material decomposition process that decomposes projection data or images into images for each material component.

[0003] However, material decomposition processes are usually limited to decomposing the projection data or images into two material components. For example, by using the Mendonca method, it is possible to decompose the projection data or images into three or more material components. However, the Mendonca method requires additional constraints / assumptions, such as volume fraction constraints, on the projection data, and if the projection data violates these additional constraints / assumptions, sufficient image quality may not be obtained. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] US Patent Application Publication No. 2019 / 0069865 [Patent Document 2] US Patent Application Publication No. 2011 / 0216951 [Patent Document 3] US Patent Application Publication No. 2009 / 0129673 Summary of the Invention [Problem to be solved by the invention]

[0005] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to improve image quality. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems. [Means for solving the problem]

[0006] A medical data processing device according to an embodiment includes a processing circuit. The processing circuit acquires data corresponding to a plurality of energy components, performs material decomposition on the data, and decomposes attenuation values ​​of each of the plurality of energy components of a voxel of the data into a plurality of material components. The processing circuit (1) determines whether the attenuation value of the voxel matches a volume fraction condition, (2) if the attenuation value of the voxel matches the volume fraction condition, decomposes the attenuation value into n material components, where n is a natural number equal to or greater than 3, and (3) if the attenuation value of the voxel does not match the volume fraction condition, projects the attenuation value of the voxel into an (n-2)-dimensional space of (n-1) material components that is the (n-2)-dimensional space closest to the attenuation value of the voxel to generate a projected point, and decomposes the attenuation value of the projected point into the (n-1) material components, thereby decomposing each attenuation value into the plurality of material components. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram showing an example of an X-ray source and an X-ray detector of an X-ray CT apparatus according to an embodiment. [Figure 2A] FIG. 2A is a diagram illustrating the processing performed by the medical data processing device according to the embodiment. [Figure 2B] FIG. 2B is a diagram illustrating the processing performed by the medical data processing device according to the embodiment. [Figure 3A] FIG. 3A is a flowchart illustrating a procedure of processing performed by the medical data processing device according to the embodiment. [Figure 3B]FIG. 3B is a flowchart illustrating the procedure of the processing performed by the medical data processing apparatus according to the embodiment. [Figure 3C] FIG. 3C is a flowchart illustrating the procedure of the processing performed by the medical data processing device according to the embodiment. [Figure 3D] FIG. 3D is a flowchart illustrating the procedure of the processing performed by the medical data processing device according to the embodiment. [Figure 3E] FIG. 3E is a flowchart illustrating the procedure of the processing performed by the medical data processing device according to the embodiment. [Figure 4A] FIG. 4A is a diagram illustrating the material decomposition process performed by the medical data processing device according to the embodiment. [Figure 4B] FIG. 4B is a diagram illustrating the material decomposition process performed by the medical data processing device according to the embodiment. [Figure 4C] FIG. 4C is a diagram illustrating the material decomposition process performed by the medical data processing device according to the embodiment. [Figure 4D] FIG. 4D is a diagram illustrating the material decomposition process performed by the medical data processing device according to the embodiment. [Diagram 5] FIG. 5 is a diagram illustrating an example of an X-ray CT apparatus according to an embodiment. [Figure 6] FIG. 6 is a diagram illustrating an example of an X-ray CT apparatus according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, embodiments of a medical data processing device, an X-ray CT device, a medical data processing method, and a program will be described in detail with reference to the drawings.

[0009] (Embodiment) The Mendonca method of multi-material decomposition is described in P. Mendonca et al., "A Flexible Method for Multi-Material Decomposition of Dual Energy CT Images," IEEE Trans. Med. Imag., 33, pp. 99-116 (2014) and U.S. Patent No. 9,036,879, which are incorporated herein by reference in their entirety. The Mendonca method imposes volume fraction conditions / constraints to uniquely decompose voxel attenuation values ​​from a Dual Energy (DE) computed tomography (CT) or spectral CT image into three material components. However, if the volume fraction conditions are not valid, the Mendonca method is unsuccessful. Therefore, a more robust method of multi-material decomposition is desirable.

[0010] The methods and apparatus described herein provide a more robust multi-material decomposition. For example, if the volume fraction condition is valid, the Mendonca method is used for multi-material decomposition. However, if the volume fraction condition is not valid, the shortest Hausdroff method or the closest edge method is used to determine the material decomposition. In the shortest Hausdroff method, a tuple of material components is selected based on a shortest Hausdroff distance criterion, and material decomposition is performed using the material components with the closest edge from the selected tuple. In the closest edge method, a tuple of material components is selected based on which edge is closest to the attenuation energy space, and material decomposition is performed using the selected tuple. Further details of both methods are provided below.

[0011] Returning to the drawings, where the same reference numerals denote the same or corresponding parts in the various views, FIG. 1 shows the source and detector portion of a computed tomography (CT) scanner having both an energy-integrated detector arranged in a third generation geometry and a photon-counting detector arranged in a fourth generation geometry. This is only a non-limiting example of an X-ray CT device configuration that can be used to acquire spectral CT data and reconstruct respective energy component images, and thus for the multi-material decomposition method described herein. FIG. 1 shows an implementation for placing a photon-counting detector (PCD) of a predetermined fourth generation geometry in combination with a detector 103 of a predetermined third generation geometry in an X-ray CT device system. FIG. 1 shows the relative positions of the object OBJ to be scanned, which is placed on a stage 116, the X-ray source 112, the collimator / filter 114, the X-ray detector 103, and the photon-counting detectors PCD1 to PCDN.

[0012] The configuration of the X-ray CT apparatus according to the embodiment, including the configuration other than the radiation source and the detector, will be described later with reference to Fig. 5 and Fig. 6. The processing circuit 1010 described later with reference to Fig. 5 or the processor 1470 in Fig. 6 is an example of the processing circuit.

[0013] In one embodiment, the X-ray source 112 and the collimator / filter 114 are fixedly connected to a rotating component 110 that is rotatably connected to a frame, and the PCD is fixedly connected to a circular component 120 that is fixedly connected to the frame. The frame of the X-ray CT apparatus also includes an opening 115 that allows the object OBJ to be placed in the projection plane of the X-rays from the X-ray source. The X-ray detector 103 is fixedly connected to a rotating component 130 that is rotatably connected to the frame. The rotating components 120 and 130 can rotate simultaneously, holding the X-ray source 112 and the X-ray detector 103 diametrically opposed to each other to acquire projection data of the object OBJ at progressive projection angles. A sinusogram is created by arranging projection data in which the projection angle is arranged along one axis and the spatial position of the projection data is arranged along the other axis.

[0014] This non-limiting example is one option for generating spectral CT projection data. Other options may include using a PCD in Generation 3 geometry without a detector in Generation 4 geometry. Additionally, an energy integrating detector arranged in Generation 3 geometry may be used with a dual energy source (e.g., a fast kVp switching source) to generate a spectral CT without a PCD. Several other variations of spectral CT gantry configurations may be used, as would be understood by one of ordinary skill in the art.

[0015] Spectral CT uses radiation with multiple energy components to perform projection measurements of the object OBJ. These projection measurements are made at a range of angles, allowing for conventional CT image reconstruction similar to non-spectral CT. However, unlike non-spectral CT, spectral CT produces additional information (i.e., spectral attenuation information) that allows for the decomposition of the reconstructed spectral image into its material components, usually two material components. The result of material decomposition is a two-component material, since there are two main interaction mechanisms that cause the attenuation of the X-ray beam traversing the imaged object OBJ. These interaction mechanisms are Compton scattering and photoelectric absorption. The mapping of projection data from the spectral domain to the material domain can be performed either before or after the image reconstruction process.

[0016] The attenuation of X-rays in biological materials is governed by two physical processes, namely photoelectric absorption and Compton scattering. Therefore, the attenuation value corresponding to a given voxel can be expressed as a function of energy and approximated by the decomposition shown in Equation (1) below.

[0017]

number

[0018] Here, μPE(E,x,y) is the photoelectric attenuation and μC(E,x,y) is the Compton attenuation.

[0019] Alternatively, the attenuation coefficients can be rearranged into a decomposition of a high-Z material (i.e., material 1) and a low-Z material (i.e., material 2) as shown in equation (2) below.

[0020]

number

[0021] where c1(x,y) and c2(x,y) correspond to the first and second material components, respectively.

[0022] 2A and 2B show the absorption coefficients μ1(E) and μ2(E) for muscle (water) and bone, respectively.

[0023] The detected spectrum is given by the following equation (3).

[0024]

number

[0025] Here, the attenuation coefficients μ1 and μ2 are known functions of X-ray energy, and the spectrum Sair corresponding to X-rays propagating in air without an absorbing object OBJ is also known, for example based on the calibration described above. The detected spectrum can be roughly decomposed into X-ray energy bins (e.g., five energy bins may be used, each covering a respective energy subband, such that the combined energy bins span the energy spectrum from approximately 20 keV to approximately 160 keV). The count value Nm of the mth energy bin can be given by the following equation (4):

[0026]

number

[0027] Here, wm(E) is the window setting function corresponding to the energy subband of the mth energy bin.

[0028] For each energy bin / component, a respective CT image can be reconstructed from the projection data corresponding to the energy bin / component. Any known method of image reconstruction can be used. For example, the image reconstruction process can be performed using any of the following: filtered backprojection, iterative image reconstruction (e.g., using a total variance minimization regularization term), Fourier-based reconstruction, or probabilistic image reconstruction.

[0029] In other words, the processing circuit of the medical data processing device according to the embodiment acquires projection data corresponding to multiple energy components of the X-rays emitted from the X-ray source 112 and passing through the subject OBJ for each of the multiple detection elements included in the detection unit 103, etc., reconstructs the projection data to generate images corresponding to each energy component, and performs material decomposition on the images. Note that the embodiment is not limited to this, and the processing circuit may acquire projection data corresponding to energy components as data for each detection element of the multiple detection elements, perform material decomposition on the acquired projection data, and reconstruct the data obtained by material decomposition to generate an image.

[0030] When reconstructing images of at least two different energy components, the attenuation values ​​of a voxel can be decomposed into material components. The term "voxel" refers to a volume pixel. For example, μi,j=μ(rj(vector quantity), Ei) is the attenuation of the i-th energy component and the j-th voxel at location rj(vector quantity). The attenuation values ​​of the first and second material components are given by μi(2)=μ(2)(Ei), respectively. These attenuation values ​​are scaled to correspond to a unit volume fraction of the respective material component (i.e., the amount of attenuation if the entire volume of the voxel is filled with the material component), and if the energy components are distributed over a range of energy values, the unit volume fraction is integrated over the range of energy values. For example, the following equation (5) holds:

[0031]

number

[0032] (where wi is a normalized window function indicating the relative contribution of the i-th energy component to the range of energy values). Using the above nomenclature, material decomposition can be performed by solving the matrix equation shown in Equation (6) below for the volume fraction αi.

[0033]

number

[0034] Here, αi=Vi / ΣjVj is the fraction of the i-th material component, and ΣjVj is the total volume of the voxel.

[0035] Additional constraints may be imposed on the decomposition of the voxel into three material components, for example the volume fraction of the three materials may be exactly 1, i.e. given by equation (7) below.

[0036]

number

[0037] That is, within the volume of a voxel, each part of the volume is occupied by one of the three material components. The above equation can then be modified to include the three materials as in equation (8) below:

[0038]

number

[0039] Furthermore, the volume fraction may be non-negative and less than or equal to 1. To increase the number of material components beyond 3, additional constraints may be imposed.

[0040] In general, the volume fraction is taken to be 1 since no part of the volume is occupied by vacuum (i.e., material occupies every part of the volume), so the volume fraction condition holds. For example, in a voxel corresponding to a region of mixed bone and lung, the three material components might be air, water, and bone. In a voxel corresponding to a region of blood vessels containing contrast agent (e.g., iodine, gadolinium), different volumes within the voxel are occupied by water, contrast agent, or some third component (e.g., fat or bone). By choosing the correct triad of material components, in general, a linear combination of the volume fractions can be found that solves the matrix equation above (i.e., the linear combination is 1 and resolves the attenuation value μ1,i for the voxel).

[0041] However, there may be cases where a linear combination of the volume fractions does not have a triad to solve the matrix equation above, and the volume fraction condition is not satisfied. If the volume fraction condition is not satisfied, another method (e.g., the shortest Hausdroff distance method or the closest edge method) is used instead.

[0042] When the multiple material decomposition results in an n-tuple (i.e., a 3-tuple) of three material components, the shortest Hausdroff distance method involves choosing the triangle of the 3-tuple (i.e., triad) of material components that has the shortest Hausdroff distance in attenuation energy space to the point corresponding to the attenuation value of the voxel. Within the chosen triangle, the pair of material components corresponding to the side of the triangle that is closest to the point of voxel attenuation is then used for the material decomposition. If the vertex of the triangle is closer than any other side, then the material decomposition is such that the voxel is entirely the material component corresponding to the closest vertex.

[0043] The closest edge method is similar to the shortest Hausdroff distance method, but instead of using the triangle with the shortest Hausdroff distance, the closest edge is selected from all of the triangles corresponding to the triad of each material component.

[0044] 3A and 3B show a flow diagram of a method 300 for performing multiple material decomposition. That is, a medical data processing device (X-ray CT device) according to an embodiment includes a processing circuit. The processing circuit acquires data corresponding to multiple energy components, performs material decomposition on the data, and resolves the attenuation values ​​of each of the multiple energy components of a voxel of the data into multiple material components.

[0045] In FIG. 3A, the method is generalized to decompose the spectral image into n material components.

[0046] For example, if n is 3, then the (n-1)-polytope is a triangle, the (n-2)-polytope is an edge, the n-tuple selected in step 320 is a triplet of material components, and the (n-1)-tuple determined in step 330 is a pair of material components.

[0047] Also, if n is 4, then the (n-1)-polytope is a triangular pyramid, the (n-2)-polytope is a triangle, the n-tuple selected in step 320 is a quartet of material components, and the (n-1)-tuple determined in step 330 is a triad of material components.

[0048] 3B, the n material components are set to 3 for ease of explanation. The method 300 is described for a given voxel in the spectral image 303. That is, performing the steps of the method 300 on a given voxel results in the decomposition of the energy component attenuation values ​​of the given voxel into material component attenuation values. To decompose the spectral image 303 into material component images, the illustrated steps are repeated for each voxel in the spectral image 303.

[0049] In step 310 of the method 300, for a given voxel of the spectral image 303, the attenuation value of each energy component is used to determine a point in the attenuation energy space. That is, the processing circuit determines a point in the attenuation energy space where the attenuation value of the voxel is located, along each axis of which is shown the attenuation value corresponding to each of the plurality of energy components. Figures 4A to 4D show an attenuation energy space where the attenuation value μ(E1) of a first energy component is along the horizontal axis and the attenuation value μ(E2) of a second energy component is along the vertical axis. In the non-limiting example shown in Figures 4A to 4D, the first axis corresponds to the measured attenuation value μ(E1) of a first material component and the second axis corresponds to the measured attenuation value μ(E2) of a second material component.

[0050] For example, when kVp switching is used to acquire the spectral CT data, the first energy component may be the X-ray spectrum when a low kVp setting is applied to the X-ray tube, and the second energy component may be the X-ray spectrum when a high kVp setting is applied to the X-ray tube. When a PCD is used to acquire the spectral CT data, the first energy component may correspond to a first energy bin of the PCD, and the second energy component may correspond to a second energy bin of the PCD. Furthermore, the acquired spectral CT may be used to generate monoenergetic spectral images at two distinct energies E1 and E2. For example, FIG. 4D shows simulated attenuation coefficients for a monoenergetic spectral image with a first energy component E1=75 keV and a second energy component E2=135 keV.

[0051] When only two dominant attenuation mechanisms contribute to the X-ray attenuation, which is typically the case for low-Z (atomic number) atoms found in biological materials, only two energy components can carry all the non-redundant information from the material's spectral attenuation signature, as described above. However, additional spectral information can be obtained for atoms with high-Z atoms due to K-edge effects. This additional information beyond the two energy components is used for material decomposition, where the attenuation energy space can have three or more dimensions. However, in Figures 4A to 4D, the non-limiting case of an attenuation energy space with only two energy components is used to illustrate a robust multi-material decomposition method.

[0052] In Fig. 4A, for example, the attenuation value of a given voxel defines a point δ, and the unit volume fractions of the three material components define the vertices of a triangle. That is, if the volume corresponding to a voxel is completely filled with a first material, the attenuation corresponding to the first and second material components has a value indicated by vertex α. Similarly, vertex β corresponds to the attenuation value of the unit volume fraction of the second material, and vertex γ corresponds to the attenuation value of the unit volume fraction of the third material. Fig. 4A shows the case where point δ is inside triangle αβγ, while Figs. 4B and 4C show two cases where point δ is outside triangle αβγ.

[0053] For simplicity, Figs. 4A-4C show only three material components, but typically more than three material components may be used in multiple material decomposition. When more than three material components may be used in multiple material decomposition, each set of three material components forms a triangle. Fig. 4D shows a non-limiting example of five material components, for example. When the spatial attenuation value of a given voxel is decomposed into triads of material components (i.e., an n-tuple with n=3) and there are four or more material components from which the triads may be selected, the number of possible triads in the decomposition may be as large as m-2 factorial, where m is the total number of material components from which triads may be selected. Each possible triad of material components corresponds to a respective triangle in the attenuation energy space. In some embodiments (e.g., the Hausdroff method), the triad of the material component whose corresponding triangle is closest to point δ is selected for material decomposition.

[0054] At step 320 of method 300, a triad (n-tuple) of material components is selected for material decomposition based on the spatial relationship between point δ and the triad's corresponding triangle ((n-1)-polytope). That is, at step 320, the processing circuitry selects an n-tuple by comparing point δ determined at step 310 to each of the (n-1)-polytopes of the n-tuple to determine whether the voxel's attenuation value meets a volume fraction condition. Each of the (n-1)-polytopes of the n-tuple has vertices corresponding to each of the n material components and having coordinates of unit volume fraction attenuation values ​​for each of the n material components. In one embodiment, the triad corresponding to the triangle with the shortest Hausdroff distance is selected as the triad for material decomposition. That is, in step 320, the processing circuitry compares the Hausdroff distances between the determined point δ and each of the (n-1)-polytopes and selects the n-tuple with the smallest Hausdroff distance between the determined point δ and each of the (n-1)-polytopes.

[0055] In another embodiment, the triad containing point δ that corresponds to the triangle with the smallest area is selected as the triad for material decomposition, or if point δ is not inside such a triangle, the triangle containing the point with the smallest area and closest to point δ is selected as the triad for material decomposition.

[0056] That is, the processing circuitry selects the n-tuple by determining a (n-1)-polytopes that contains either (1) the determined point δ or (2) a point in each (n-1)-polytopes that is closest to the determined point δ (if each (n-1)-polytopes does not contain the determined point δ). Note that the processing circuitry may select the n-tuple corresponding to the smallest (n-1)-polytopes among the multiple (n-1)-polytopes if multiple (n-1)-polytopes among each (n-1)-polytopes contain the determined point δ or contain a point closest to the determined point δ.

[0057] In step 325 of method 300 , a query is performed to determine whether the point δ is inside or outside the triangle of the triad selected in step 320 .

[0058] That is, in step 325, the processing circuitry determines whether the attenuation value of the voxel meets a volume fraction condition, where (1) the volume fraction of each of the voxel's multiple material components is nonnegative, and (2) the sum of the volume fractions of the voxel is 1. The attenuation value of the voxel meeting the volume fraction condition corresponds to the determined point δ being inside the (n-1)-polytope of the selected n-tuple.

[0059] If the point δ is inside the selected triangle ((n-1)-polytope), the method 300 proceeds to step 360. That is, if the attenuation value of the voxel meets the volume fraction condition, the process proceeds to step 360, where the processing circuitry decomposes the attenuation value into n material components, where n is a natural number equal to or greater than 3. For example, if the determined point δ is within the (n-1)-polytope of the selected n-tuple, the processing circuitry determines the volume fractions of the n material components in the selected n-tuple as the material decomposition of the voxel using the Mendonca algorithm.

[0060] On the other hand, if the point δ is not inside the selected triangle, the method 300 proceeds to step 330 .

[0061] FIG. 4A shows the case where the point δ is inside the triangle, and FIGS. 4B and 4C show the case where the point δ is outside the triangle.

[0062] In step 360 of method 300, a multi-material decomposition is performed to decompose the attenuation values ​​of a voxel into triads (n-tuples) of material components. For example, the material decomposition may be performed by solving for volume fractions αi that satisfy the matrix equation shown in Equation (8) below, subject to the condition that the volume fractions αi are non-negative.

[0063]

number

[0064] Other methods of performing the material decomposition of the material components into the selected triads may also be used, as would be understood by one of ordinary skill in the art.

[0065] In step 330 of method 300, a pair ((n-1)-tuple) of material components is selected from the triad (n-tuple) of material components selected in step 320. This pair ((n-1)-tuple) of material components is selected to correspond to the edge ((n-2)-polytope) of the triangle ((n-1)-polytope) that is closest to the point δ.

[0066] That is, the processing circuitry determines the (n-2)-polytope of the (n-1)-polytope that is closest to the determined point if the determined point is not within the (n-1)-polytope of the selected n-tuple.

[0067] In Figure 4B, the edge αγ is closest to the point δ. The closest edge ((n-2)-polytope) can be determined, for example, by determining the line segment (i.e., n-1 dimensional space) defined by each edge ((n-2)-polytope) of the triangle ((n-1)-polytope) and orthogonally projecting the point δ to the projected point ε in the line segment (n-1 dimensional space). The edge ((n-2)-polytope) whose projected point ε is closest to the point δ is the closest edge ((n-2)-polytope), as shown in Figures 4B and 4C.

[0068] That is, the processing circuitry projects the attenuation value of the voxel into the (n-2) dimensional space of the (n-1) material components that is closest to the attenuation value of the voxel if the attenuation value of the voxel does not meet the volume fraction condition to generate a projected point, the projected point being an orthogonal projection of the determined point into the (n-2) dimensional space of the closest (n-2)-polytope.

[0069] At step 340 of method 300, multiple material decomposition is performed to decompose the attenuation values ​​of the voxel into a pair ((n-1)-tuple) of material components from step 330. For example, in FIG. 4B, the volume fraction of the second material component is determined to be zero (i.e., α2=0). Thus, material decomposition may be performed by solving for the volume fraction αi that satisfies the matrix equation given in equation (9) below:

[0070]

number

[0071] Here, μi,j(ε) is the attenuation value of the i-th energy component at the projected point ε. The above matrix equation has the condition that the volume fraction αi is non-negative.

[0072] That is, the processing circuitry decomposes each attenuation value into a plurality of material components by causing the attenuation value of the projected point to be decomposed into (n-1) material components, or in other words, the processing circuitry decomposes the attenuation value of a voxel into a volume fraction of an (n-1)-tuple of material components corresponding to the nearest (n-2)-polytope.

[0073] In Figure 4C, the projected point ε is not yet within the triangle αβγ. So a second projection can be performed to determine the closest point ((n-2)-polytope) to the projected point ε and the point δ. This point corresponds to a single ((n-1)-tuple) material component. For example, in Figure 4C, vertex γ is closest to the projected point ε and the point δ. Therefore, the material decomposition will be α3=1.

[0074] That is, the processing circuitry determines the closest (n-3)-polytope for the determined point in step 310 if the determined point is not within the (n-1)-polytope of the selected n-tuple and the projected point is not within the closest (n-2)-polytope, where the closest (n-3)-polytope is the (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point. In this manner, the processing circuitry decomposes the attenuation value of the voxel into volume fractions of (n-2) material components that correspond to the closest (n-3)-polytope.

[0075] In one embodiment, the space is divided into the following regions (n-1 dimensional regions): (i) triangle αβγ, (ii) region A (bounded by lines αAD and αAF), (iii) region B (bounded by lines βBF and βBE), (iv) region C (bounded by lines γCE and γCD), (v) region D (bounded by lines αγ, αAD, and γCD), (vi) region E (bounded by lines βγ, βBE, and γCE), (vii) region F (bounded by lines αβ, αAF, and βBF). Lines αAF and βBF are perpendicular to line segment αβ. Lines βBE and γCE are perpendicular to line segment βγ. These lines, lines αAD and γCD, are perpendicular to line segment αγ.

[0076] Material decomposition may depend on which of these seven regions point δ lies in. If point δ lies in triangle αβγ, material decomposition is performed by decomposing the attenuation values ​​into all three material components. If point δ lies in regions A, B, and C, the points in the triangle that are closest to point δ are the vertices. Thus, material decomposition is performed by decomposing the attenuation values ​​into the material components that correspond to the closest vertices.

[0077] If point δ is in regions E, F, and G, then the point in the triangle that is closest to point δ is along one of the three edges. Therefore, material decomposition is performed by decomposing the attenuation values ​​into the two material components that correspond to the closest edges. This can be performed, for example, by orthogonally projecting point δ onto the edges and solving the matrix equation given by Equation (10) below, as previously described.

[0078]

number

[0079] After the attenuation values ​​are decomposed into multiple material components in step 360 or step 340 and material decomposition is performed in step 393, the processing circuitry then generates an image in step 395 based on the results of the material decomposition performed in step 393. Examples of the image generated in step 395 include an image showing the volume fraction of each material component and an image showing which material component has the largest volume fraction for each voxel. The image generated in step 395 is, for example, a color map image. The processing circuitry also causes the generated images to be displayed on the display device 1016.

[0080] Figure 3C illustrates another embodiment of method 300, where step 320 is performed immediately before step 360, rather than immediately before step 325. Furthermore, instead of asking whether point δ is inside a particular triangle, step 325 asks whether point δ is inside the entire closed region of the union of all triangles.

[0081] That is, the processing circuitry determines in step 310 whether the determined point is within a closed region defined by the union of respective (n-1)-polytopes in attenuation energy space corresponding to each n-tuple of a plurality of material components, where each (n-1)-polytope of the n-tuple has vertices corresponding to each of the n material components and having coordinates of unit volume fraction attenuation values ​​for each of the n material components.

[0082] If the point determined in step 310 is within a closed region, the processing circuitry performs steps 320 and 360 to determine, for example using the Mendonca method, the volume fractions of the n material components in the selected n-tuple corresponding to the (n-1)-polytope closest to the determined point δ.

[0083] On the other hand, if a point δ is not in any triangular closed region, then it is not in any triangular region either, so it is decomposed into the pair of materials corresponding to the nearest edge, or the material corresponding to the nearest vertex if that vertex is the closest point to point δ in the closed region.

[0084] That is, if the attenuation value of the voxel does not meet the volume fraction condition, the processing circuitry decomposes each attenuation value into a plurality of material components by projecting the attenuation value of the voxel into a (n-2)-dimensional space of (n-1) material components that is the (n-2)-dimensional space that is closest to the attenuation value of the voxel to generate a projected point, and decomposing the attenuation value of the projected point into said (n-1) material components. In other words, if the determined point δ is not within a closed region, the processing circuitry determines a (n-2)-polytope within the (n-1)-polytope that is the closest (n-2)-polytope to the determined point δ, and decomposes the attenuation value of the voxel into a volume fraction of the (n-1)-tuple of material components corresponding to the closest (n-2)-polytope.

[0085] Here, the projected point is the orthogonal projection of the determined point δ into the (n-2)-dimensional space of the nearest (n-2)-polytope.

[0086] Additionally, the processing circuitry determines a closest (n-3)-polytope for the determined point δ if the determined point δ is not within the closed region and the projected point is not within the closest (n-2)-polytope, where the closest (n-3)-polytope is the (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point δ. The processing circuitry decomposes the attenuation value of the voxel into a volume fraction of (n-2) material components corresponding to the closest (n-3)-polytope.

[0087] In this embodiment, the choice of finding the closest triangle or edge / vertex may be performed early in the method, before determining the closest triangle to point δ. As will be appreciated by one of ordinary skill in the art, this embodiment of method 300 may be generalized to higher dimensions, just as Figure 3A generalizes the embodiment of Figure 3B to higher dimensions. Figure 3D shows a generalization of the embodiment of method 300 of Figure 3C to any n-dimension greater than 2.

[0088] FIG. 3E shows a flow diagram of step 340. As previously described with reference to FIGS. 4B and 4C, if point δ is not within the triangle, a series of orthogonal projections into the reduced dimensional space of the nearest polytope will eventually result in the projected point being within the triangle, and once this is achieved, material decomposition can be performed. For example, in FIG. 4B, point ε will be within the triangle if only the projection of point δ into the one-dimensional space of line segment αγ is sufficient. However, in FIG. 4C, two projections were required to obtain the result that the projected point ζ is within the triangle (i.e., a first projection into the one-dimensional space of line segment αγ, and a second projection of the projected point ε into the zero-dimensional space of vertex γ).

[0089] This process allows generalization to the case of n=4, and instead of starting by evaluating whether point δ is in the nearest triangle, the initial question is whether point δ is in the nearest triangular pyramid (i.e., the 3-polytope). If point δ is not in the nearest triangular pyramid, it is projected into the 2-dimensional space of the nearest triangle. If the point projected into 2-dimensional space is not in the nearest triangle, the process continues as in the n=3 case above, up to a reduced dimension.

[0090] 3E, this process is generalized to any value of n. In step 341, a loop variable i is initialized to a value of 2, and the point in the (n-1) dimensional space of the n-polytope is set to the point δ, determined in step 310.

[0091] In step 342, points from the (n+1-i)-dimensional space are projected into the (ni)-dimensional space of the closest (ni)-polytope.

[0092] In step 344, a query is performed as to whether the projected point onto the (ni)-dimensional space is within the closest (ni)-polytope. If not, the loop continues to step 346. If so, the loop proceeds to step 348.

[0093] In step 346, the loop variable i is decremented by 1 (i.e., i=i-1) and the closest (ni)-polytope projected into the (ni)-dimensional space is determined. After step 346, the loop continues to step 342.

[0094] In step 348, multi-material decomposition is performed to decompose the projected points into material components that correspond to the nearest (ni)-polytope.

[0095] Figure 4D illustrates the case assuming five material components: if the attenuation value of the first voxel is mapped to point B shown in Figure 4D, that point is not within any triangle of the respective triad of the material components.

[0096] According to the first method of determining the closest triangle using the Hausdroff distance (e.g., the shortest Hausdroff distance method), the triad of water, diluent OMNIPAQUE 300 (contrast agent), and bone corresponds to the closest triangle. Within this triangle, the closest edge corresponds to a water-bone pair. Thus, point B would be resolved into a superposition of the volume fractions of water and bone.

[0097] According to the second method, which determines the closest edge without first determining the closest triangle (e.g., the closest edge method), the closest edge corresponds to a water-bone pair, so the material decomposition will be similar to the first method.

[0098] If the attenuation value of the second voxel is mapped to point A in FIG. 4D, the point is not within any triangle of the material component's respective triad.

[0099] According to the first method of determining the closest triangle using the Hausdroff distance, air, fat, and diluted OMNIPAQUE 300 (contrast agent) correspond to the closest triangle. Within this triangle, the closest edge corresponds to a pair of air and diluted OMNIPAQUE 300. Thus, point A will be resolved into a superposition of volume fractions of air and diluted OMNIPAQUE 300.

[0100] According to the second method, which determines the closest edge without first determining the closest triangle, the closest edge corresponds to an air-bone pair. Thus, point A will be resolved into a superposition of air and bone volume fractions. This is a different result than the first method.

[0101] 5 shows the implementation of a radiological imaging gantry included in a CT device or scanner. As shown in FIG. 5, the radiological imaging gantry 1000 is shown in a side view and further includes an X-ray tube 1001, an annular frame 1002, and a multi-row or two-dimensional array type X-ray detection unit 1003. The X-ray tube 1001 and the X-ray detection unit 1003 are installed to face each other across an object OBJ on the annular frame 1002 that is rotatably held around a rotation axis RA. A rotating unit 1007 rotates the annular frame 1002 at a high speed, such as 0.4 seconds / rotation, while the object OBJ moves along the axis RA into or out of the illustrated page.

[0102] A first embodiment of an X-ray computed tomography (CT) apparatus according to the present invention will be described below with reference to the accompanying drawings. Note that X-ray CT apparatuses include various types of apparatus, such as a rotating / rotating type apparatus in which an X-ray tube and an X-ray detector rotate together around an object to be inspected, and a stationary / rotating type apparatus in which many detector elements are arranged in a ring or plane and only the X-ray tube rotates around an object to be inspected. The present invention can be applied to either type. Here, a rotating / rotating type apparatus, which is currently mainstream, is exemplified.

[0103] The multi-slice X-ray CT apparatus further includes a high voltage device 1009 that generates a tube voltage applied to the X-ray tube 1001 by a slip ring 1008 so that the X-ray tube 1001 generates X-rays. The X-rays are emitted toward the object OBJ, and the cross-sectional area is indicated by a circle. For example, the X-ray tube 1001 has an average X-ray energy in a first scan that is smaller than the average X-ray energy in a second scan. Therefore, two or more scans corresponding to different X-ray energies can be acquired. The X-ray detection unit 1003 is installed on the opposite side of the X-ray tube 1001 across the object OBJ, and detects the emitted X-rays that have passed through the object OBJ. The X-ray detection unit 1003 further includes individual detection elements or detection units.

[0104] The X-ray CT apparatus further includes other devices for processing detected signals from the X-ray detection unit 1003. A data acquisition circuit or data acquisition mechanism (Data Acquisition System: DAS) 1004 converts the signal output from the X-ray detection unit 1003 of each channel into a voltage signal, amplifies the signal, and further converts the signal into a digital signal. The X-ray detection unit 1003 and the DAS 1004 are configured to process a predetermined total number of projections per rotation (TPPR).

[0105] The aforementioned data is transmitted by a non-contact data transmission device 1005 to a pre-processing circuit 1006 housed in a console outside the radiography gantry 1000. The pre-processing circuit 1006 performs predetermined corrections, such as sensitivity corrections for the raw data. A storage device 1012 stores the result data, also called projection data at a stage immediately before reconstruction processing. The storage device 1012 is connected to the processing circuit 1010 by a data / control bus 1011 together with a reconstruction device 1014, an input interface 1015, and a display device 1016. The pre-processing circuit 1010 controls a current regulator 1013 that limits the current to a level sufficient to drive the CT device.

[0106] The detectors are rotated and / or fixed relative to the patient in various generations of X-ray CT scanner systems. In one embodiment, the aforementioned CT system may be an example of a combination of third generation geometry and fourth generation geometry systems. In a third generation system, the X-ray tube 1001 and the X-ray detector 1003 are mounted on an annular frame 1002 in an opposing manner and rotate around the object OBJ as the annular frame 1002 rotates about the rotation axis RA. In a fourth generation geometry system, the detectors are fixedly mounted around the patient and the X-ray tube rotates around the patient. In another embodiment, in the radiography gantry 1000, multiple detectors are arranged on an annular frame 1002 that is held by a C-arm and a stand.

[0107] The storage device 1012 may store a measurement value indicating the amount of X-ray irradiation at the X-ray detection unit 1003. Furthermore, the storage device 1012 may store a dedicated program for executing the method 300.

[0108] The reconstruction circuitry 1014 may perform various steps of the method 300. Additionally, the reconstruction circuitry 1014 may perform pre-reconstruction image processing, such as volume rendering and image differencing, as necessary.

[0109] Such pre-reconstruction processing of the projection data performed in pre-processing circuitry 1006 may include, for example, detector calibration, detector non-linearity, and correction for polar effects.

[0110] Post-reconstruction processing performed by the reconstruction circuitry 1014 may include image filtering and smoothing, volume rendering, and image subtraction, as appropriate. The image reconstruction processing may perform CT reconstruction to generate spectral images and to perform various steps of method 300 on the spectral images. The reconstruction circuitry 1014 may use memory to store, for example, projection data, reconstructed images, calibration data and parameters, and computer programs.

[0111] The reconfiguration circuit 1014 may include a CPU (processing circuitry) that may be implemented as discrete logic gates, as an Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA), or other Complex Programmable Logic Device (CPLD). The implementation of an FPGA or CPLD may be coded in VHDL, Verilog, or other hardware description language, and the code may be stored directly in electronic memory within the FPGA or CPLD or as other separate electronic memory. Furthermore, the storage 1012 may be non-volatile, such as ROM, EPROM, EEPROM, or FLASH memory. The storage 1012 may also be volatile, such as static or dynamic RAM, and a processor, such as a microcontroller or microprocessor, may be provided to manage the electronic memory as well as the interaction between the FPGA or CPLD and the memory.

[0112] Alternatively, the CPU of the reconstruction circuit 1014 may execute a computer program including a set of computer readable instructions to perform the functions described herein, the program being stored in any of the aforementioned non-transitory electronic memories and / or hard disk drives, CDs, DVDs, FLASH® drives, or any other known storage media. Furthermore, the computer readable instructions may be provided as a utility application, background daemon, or operating system component, or a combination thereof, that executes with a processor, such as Intel®'s Xenon® processor, AMD®'s Opteron® processor, and an operating system, such as Microsoft®'s VISTA®, UNIX®, Solaris®, LINUX®, Apple®'s MAC-OS®, known to those skilled in the art. Furthermore, the CPU may be implemented as a multi-processor that operates cooperatively in parallel to execute the instructions.

[0113] In one embodiment, the reconstructed image may be displayed on a display device 1016. The display device 1016 may be an LCD display, a CRT display, a plasma display, an OLED, an LED, or any other display device known to one of ordinary skill in the art.

[0114] The storage device 1012 may be a hard disk drive, a CD-ROM drive, a DVD drive, a FLASH drive, RAM, ROM, or any other electronic storage medium known to those of skill in the art.

[0115] Another embodiment of a computed tomography (CT) scanner having both an energy-integrating detector arranged in a third generation geometry and a PCD arranged in a fourth generation geometry is shown in Figure 6. Figure 6 shows an implementation for placing a PCD of a predetermined fourth generation geometry in combination with a detector section 103 of a predetermined third generation geometry in an x-ray CT device system. The relative positions of the x-ray source 112, collimator / filter 114, x-ray detector section 103, and photon counting detectors PCD1 to PCDN are shown in this figure.

[0116] In addition to the configuration of the X-ray source 112, detector including the detector 103, and PCD shown in FIG. 6, other types of combinations of X-ray detectors and X-ray sources can be used to acquire projection data. For example, either the detector 103 or the PCD can be omitted from the scanner shown in FIG. 6, and the scanner can still acquire projection data, but different projection data than that acquired using the complete system shown in FIG. 6. Furthermore, kV switching can be used with an energy-integrated detector or a PCD. In an embodiment, the PCD can be a direct X-ray detector that uses semiconductors to directly convert X-rays into photoelectrons without first generating scintillation photons. Furthermore, in an embodiment, the broadband X-ray source can be used with a spectrally resolving X-ray detector. These spectrally resolving X-ray detectors can include any configuration of PCD (e.g., a predefined third generation geometry or a predefined fourth generation geometry) or energy-integrated detectors preceded by respective spectral filters. In an embodiment, the X-ray source can include multiple narrowband X-ray sources, such as a dual-source X-ray CT device. In general, any known combination of detector types and configurations may be used with any known type or combination of x-ray source to generate projection data.

[0117] Returning to Figure 6, Figure 6 also illustrates circuitry and hardware for acquiring, storing, processing, and distributing x-ray projection data, including a processor 1470, a network controller 1480, a memory 1476, and a data acquisition system 1476.

[0118] In another embodiment, the X-ray CT apparatus includes a PCD but does not include the energy integrating detector 103 .

[0119] The X-ray source 112 and the detector 103 are housed in a platform 1440, and as they rotate around the circular paths 110 and 130, respectively, the photon counting detectors PCD and the detector 103 detect the transmitted X-ray radiation during data acquisition. The photon counting detectors PCD1 to PCDN intermittently detect the transmitted X-ray radiation for each of the predetermined energy bins and output a count value indicating the number of photons. Meanwhile, the sensing elements of the detector 103 continuously detect the transmitted X-ray radiation and output a detected signal as the detector 103 rotates. In one embodiment, the detector 103 has energy integration detectors densely arranged on the detector surface in the direction of the predetermined channel and segment.

[0120] In one embodiment, the X-ray source 112, the PCD and the detector 103 collectively form three predetermined circular paths with different radii. At least one X-ray source 112 rotates along a first circular path 110 while the photon counting detectors are sparsely placed along a second circular path 120. Additionally, the detector 103 moves along a third circular path 130. The first circular path 110, the second circular path 120 and the third circular path 130 may be determined by an annular ring that is rotatably mounted on the gantry 1440.

[0121] Further, alternative embodiments may be used to position the photon-counting detector in a given fourth generation geometry in combination with the detector in a given third generation geometry in the x-ray CT apparatus.

[0122] In one embodiment, the X-ray source 112 is optionally a single energy source. In another embodiment, the X-ray source 112 is configured to perform a kV switching function to emit X-ray radiation at a predetermined high level energy and a predetermined low level energy. In yet another embodiment, the X-ray source 112 is a single source that emits a broad spectrum of X-ray energies. In yet another embodiment, the X-ray source 112 includes multiple X-ray emitters, where each emitter is spatially and spectrally distinct.

[0123] The detector 103 may use an energy-integrating detector, such as a scintillation element, in conjunction with a photomultiplier tube or an avalanche photodiode, to detect scintillation photons from scintillation events that occur upon x-ray radiation interacting with the scintillator element, which may be crystalline, organic liquid, plastic, or other known scintillators.

[0124] PCDs may use direct x-ray radiation detectors based on semiconductors such as cadmium telluride (CdTe), cadmium zinc telluride (CZT), silicon (Si), mercuric iodide (HgI2) and gallium arsenide (GaAs).

[0125] The X-ray CT apparatus also includes a data channel that transmits projection measurements from the photon counting detector and detector 103 to a data acquisition system 1476, a processor 1470, a memory 1478, and a network controller 1480. The data acquisition system 1476 controls the acquisition, digitization, and transmission of projection data from the detector. The data acquisition system 1476 also includes radiography control circuitry that controls the rotation of the annular rotating frames 110 and 130. In one embodiment, the data acquisition system 1476 also includes control of the movement of the couch 116, the operation of the X-ray source 112, and the operation of the X-ray detector 103. The data acquisition system 1476 can be a centralized or distributed system. In an implementation, the data acquisition system 1476 is integrated with the processor 1470. The processor 1470 performs functions including reconstruction of images from projection data, pre-reconstruction processing of the projection data, and post-reconstruction processing of image data. The processor 1470 also performs the functions and methods described herein.

[0126] Pre-reconstruction processing of the projection data may include correction for detector calibration, detector nonlinearity, polar effects, noise balancing, and material decomposition.

[0127] The image reconstruction process may be performed using a filtered backprojection method, an iterative image reconstruction method, or a stochastic image reconstruction method. Moreover, the reconstruction process may include various steps of the method 300.

[0128] The post-reconstruction processing may include image filtering and smoothing, volume rendering, and image differencing, as appropriate. Moreover, the post-reconstruction processing may include various steps of method 300.

[0129] Both the processor 1470 and the data acquisition system 1476 may use memory 1476 to store, for example, projection data, reconstructed images, calibration data and parameters, and computer programs.

[0130] The processor 1470 may include a CPU (processing circuitry) that may be implemented as discrete logic gates, as an application specific integrated circuit (ASIC), as a field programmable gate array (FPGA), or as other complex programmable logic device (CPLD). An FPGA or CPLD implementation may be coded in VHDL, Verilog, or other hardware description language, and the code may be stored directly in electronic memory within the FPGA or CPLD, or as other separate electronic memory. Furthermore, the memory may be non-volatile, such as ROM, EPROM, EEPROM, or FLASH memory. The memory may also be volatile, such as static or dynamic RAM, and a processor, such as a microcontroller or microprocessor, may be provided to manage the electronic memory as well as the interaction between the FPGA or CPLD and the memory.

[0131] Alternatively, the CPU of the reconfiguration processor may execute a computer program including a set of computer readable instructions to perform the functions described herein, the program being stored in any of the aforementioned non-transitory electronic memories and / or hard disk drives, CDs, DVDs, FLASH® drives, or any other known storage media. Furthermore, the computer readable instructions may be provided as a utility application, background daemon, or operating system component, or a combination thereof, to run with a processor, such as Intel®'s Xenon® processor, AMD®'s Opteron® processor, and an operating system known to those skilled in the art, such as Microsoft®'s VISTA®, UNIX®, Solaris®, LINUX®, Apple®'s MAC-OS®, etc. Furthermore, the CPU may be implemented as a multi-processor operating in parallel and cooperatively to execute the instructions.

[0132] In one embodiment, the reconstructed image may be displayed on a display device, which may be an LCD display, a CRT display, a plasma display, an OLED, an LED, or any other display device known to one of ordinary skill in the art.

[0133] The memory 1478 may be a hard disk drive, a CD-ROM drive, a DVD drive, a FLASH drive, RAM, ROM, or any other electronic storage medium known to those of skill in the art.

[0134] A network controller 1480, such as an Intel Ethernet PRO network interface card from Intel Corporation of the United States of America, may interface between various parts of the X-ray CT apparatus. Additionally, the network controller 1480 may interface with an external network. As will be appreciated, the external network may be a public network, such as the Internet, or a private network, such as a LAN or WAN network or a combination thereof, and may also include a PSTN or ISDN sub-network. The external network may be wired, such as an Ethernet network, or wireless, such as a cellular network, including EDGE, 3G and 4G wireless cellular systems. The wireless network may also be wireless, such as WiFi, Bluetooth, or other known forms of communication.

[0135] According to at least one of the embodiments described above, image quality can be improved.

[0136] Although some embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, modifications, and combinations of embodiments can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope of the invention and its equivalents as described in the claims, as well as in the scope and spirit of the invention. [Explanation of symbols]

[0137] 1003 X-ray detector 1010 Processing circuit 1012 Storage device 1016 Display device

Claims

1. Acquire data corresponding to multiple energy components, a processing circuit for performing material decomposition on the data to resolve attenuation values ​​of each of the plurality of energy components of a voxel of the data into a plurality of material components; the processing circuit (1) determines whether the attenuation value of the voxel matches a volume fraction condition; (2) if the attenuation value of the voxel matches the volume fraction condition, decomposes the attenuation value into n material components, where n is a natural number greater than or equal to 3; and if the attenuation value of the voxel does not match the volume fraction condition, projects the attenuation value of the voxel into an (n-2) dimensional space of (n-1) material components that is closest to the attenuation value of the voxel to generate a projected point, and decomposes the attenuation value of the projected point into the (n-1) material components, thereby decomposing each of the attenuation values ​​into the multiple material components.

2. the volume fraction condition is: (1) the volume fractions of the material components at the voxel are each non-negative; and (2) the sum of the volume fractions at the voxel is 1; The processing circuitry determines whether the attenuation value of the voxel conforms to the volume fraction condition by: (1) determining a point in an attenuation energy space along each axis at which an attenuation value of the voxel is located, the attenuation value corresponding to each of the plurality of energy components being plotted; (2) selecting an n-tuple by comparing the determined points to a respective (n-1)-polytope of the n-tuple, the respective (n-1)-polytope of the n-tuple having vertices corresponding to each of the n material components and having coordinates of unit volume fraction attenuation values ​​for each of the n material components; The processing circuitry includes: if the determined point is within the (n-1)-polytope of the selected n-tuple, determining a volume fraction of the n material components in the selected n-tuple as a material decomposition of the voxel using Mendonca's algorithm; 2. The medical data processing apparatus of claim 1, further comprising: determining a (n-2)-polytope of the (n-1)-polytope that is closest to the determined point if the determined point is not within the (n-1)-polytope of the selected n-tuple; and decomposing the attenuation value of the voxel into a volume fraction of material components of the (n-1)-tuple that corresponds to the closest (n-2)-polytope.

3. 3. The medical data processing device according to claim 2, wherein the processing circuitry compares the Hausdroff distances between the determined points and the respective (n-1)-polytopes, and selects the n-tuple having the smallest Hausdroff distance between the determined points and the respective (n-1)-polytopes.

4. 3. The medical data processing device of claim 2, wherein the processing circuitry selects the n-tuple by determining an (n-1)-polytope that contains either (1) the determined point or (2) a point in the respective (n-1)-polytope that is closest to the determined point if the respective (n-1)-polytope does not contain the determined point.

5. the processing circuitry, if a plurality of (n-1)-polytopes of the respective (n-1)-polytopes contain the determined point or contain the point closest to the determined point, The medical data processing apparatus according to claim 4 , further comprising: selecting the n-tuple corresponding to the smallest (n−1)-polytope from among the plurality of (n−1)-polytopes.

6. the volume fraction condition is: (1) the volume fraction of each of the plurality of material components of the voxel is non-negative; and (2) the sum of the volume fractions of the voxel is 1; The processing circuitry determines whether the attenuation value of the voxel conforms to the volume fraction condition by: (1) determining a point at which the attenuation value of the voxel is located in an attenuation energy space along each axis, the attenuation value corresponding to each of the plurality of energy components being displayed; (2) determining whether the determined point is within a closed region defined by a union of respective (n-1)-polytopes in the attenuation energy space corresponding to each n-tuple of the plurality of material components, each (n-1)-polytope of the n-tuple having vertices corresponding to each of the n material components and having coordinates of unit volume fraction attenuation values ​​for each of the n material components; The processing circuitry includes: if the determined point is within the closed region, determining volume fractions of the n material components in a selected n-tuple corresponding to the (n-1)-polytope closest to the determined point using Mendonca's algorithm; 2. The medical data processing apparatus of claim 1, further comprising: determining, if the determined point is not within the closed region, a (n-2)-polytope within the (n-1)-polytope that is closest to the determined point; and decomposing the attenuation value of the voxel into a volume fraction of material components of an (n-1)-tuple corresponding to the closest (n-2)-polytope.

7. the projected point is an orthogonal projection of the determined point into the (n-2)-dimensional space of the nearest (n-2)-polytope, the processing circuitry determines a closest (n-3)-polytope for the determined point if the determined point is not within the (n-1)-polytope of the selected n-tuple and the projected point is not within the closest (n-2)-polytope, the closest (n-3)-polytope being the (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point; the processing circuitry decomposing the attenuation values ​​of the voxels into volume fractions of (n-2) material components corresponding to the nearest (n-3)-polytope. The medical data processing device according to claim 2 .

8. the projected point is an orthogonal projection of the determined point into the (n-2)-dimensional space of the nearest (n-2)-polytope; the processing circuitry determines a closest (n-3)-polytope for the determined point if the determined point is not within the closed region and the projected point is not within the closest (n-2)-polytope, the closest (n-3)-polytope being the (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point; 7. The medical data processing apparatus of claim 6, wherein the attenuation values ​​of the voxels are decomposed into volume fractions of (n-2) material components corresponding to the nearest (n-3)-polytope.

9. n is 3 or 4; if n is 3, the (n-1)-polytope is a triangle, the (n-2)-polytope is an edge, the selected n-tuple is a triplet of the material components, and the determined (n-1)-tuple is a pair of the material components; 3. The medical data processing apparatus of claim 2, wherein, when n is 4, the (n-1)-polytope is a triangular pyramid, the (n-2)-polytope is a triangle, the selected n-tuple is a quadruple of the material components, and the determined (n-1)-tuple is a triad of the material components.

10. The processing circuitry includes:

2. The medical data processing device according to claim 1, wherein projection data corresponding to the energy components is acquired as the data for each detection element of a plurality of detection elements, the projection data is reconstructed to generate images corresponding to each of the energy components, and the material decomposition is performed on the images.

11. The processing circuitry includes:

2. The medical data processing device according to claim 1, further comprising: acquiring projection data corresponding to the energy components for each of a plurality of detection elements as the data; performing the material decomposition on the projection data; and reconstructing the data obtained by the material decomposition to generate an image.

12. An X-ray source that emits X-rays; a plurality of detector elements for detecting a plurality of energy components of the X-rays emitted from the X-ray source and passing through a subject; acquiring data corresponding to the plurality of energy components; a processing circuit for performing material decomposition on the data to resolve attenuation values ​​of each of the plurality of energy components of a voxel of the data into a plurality of material components; the processing circuit (1) determines whether the attenuation value of the voxel matches a volume fraction condition; (2) if the attenuation value of the voxel matches the volume fraction condition, decomposes the attenuation value into n material components, where n is a natural number greater than or equal to 3; and if the attenuation value of the voxel does not match the volume fraction condition, projects the attenuation value of the voxel into an (n-2) dimensional space of (n-1) material components that is closest to the attenuation value of the voxel to generate a projected point, and decomposes the attenuation value of the projected point into the (n-1) material components, thereby decomposing each of the attenuation values ​​into the multiple material components.

13. A medical data processing method performed by a medical data processing device, comprising: Acquire data corresponding to multiple energy components, performing material decomposition on the data to decompose the attenuation values ​​of each of the plurality of energy components of a voxel of the data into a plurality of material components; The decomposition of each attenuation value into the plurality of material components includes:

1. A medical data processing method, comprising: (1) determining whether the attenuation value of the voxel matches a volume fraction condition; (2) if the attenuation value of the voxel matches the volume fraction condition, decomposing the attenuation value into n material components, where n is a natural number greater than or equal to 3; and (3) if the attenuation value of the voxel does not match the volume fraction condition, projecting the attenuation value of the voxel into an (n-2) dimensional space of n-1 material components that is closest to the attenuation value of the voxel to generate a projected point, and decomposing the attenuation value of the projected point into the (n-1) material components.

14. A program for causing a computer to execute the medical data processing method according to claim 13.

Citation Information

Patent Citations

  • Method and apparatus for determination of concentration of substance in body constituting elements

    JP2007229464A

  • Multi-material decomposition using dual energy computed tomography method

    JP2010125334A

  • X-ray CT apparatus, image processor, and program

    JP2013081765A

  • Apparatus for decomposition of multiple substances

    JP2019532691A

  • Method for segmentation of lesions

    US20090129673A1