Method and apparatus for spectral computed tomography (CT) with multi-material decomposition into three or more material components
View Patent ↗A method and apparatus uses multi-material decomposition of three or more material components to generate material-component images from spectral images reconstructed from spectral computed tomography data. In three-component material decomposition e.g., the Mendonça method is used for multi-material decomposition when the attenuation values satisfy an assumed volume fraction condition (i.e., for a given voxel, the attenuation values are within a triangle having vertices given by unit volume fractions of three respective material components). However, when the volume fraction condition fails (e.g., the attenuation values are outside the triangle), either a shortest-Hausdorff-distance method or a closest-edge method is used for multi-material decomposition. For example, the attenuation values of the voxel are projected onto a lower-dimensional space (e.g., the space of a closest edge) and decomposed into a pair/single material component(s) of the lower-dimensional space.
1. An apparatus, comprising:
circuitry configured to
obtain images corresponding to respective energy components, and
perform a material decomposition on the images to decompose attenuation values for the respective energy components of a voxel of the images into material components, wherein,
to perform the material decomposition, the circuitry is further configured to
determine whether the attenuation values of the voxel are consistent with a volume-fraction condition,
when the attenuation values of the voxel are consistent with the volume-fraction condition, decompose the attenuation values into n material components, wherein n is a natural number of three or greater, and,
when the attenuation values of the voxel are not consistent with the volume-fraction condition, project the attenuation values of the voxel onto a closest (n-2) dimensional space of n-1 material components to the attenuation values of the voxel to generate a projected point, and decompose attenuation values of the projected point into the n-1 material components.
2. The apparatus according to claim 1 , wherein
the volume-fraction condition is that (i) volume fractions of the material components in the voxel are non-negative and (ii) a sum of the volume fractions for the voxel is one, and,
to determine whether the attenuation values of the voxel are consistent with the volume-fraction condition, the circuitry is further configured to
determine a point in an attenuation-energy space having coordinates given by the attenuation values of the voxel, the attenuation-energy space representing, along respective axes, attenuation values corresponding to respective components of the plurality of energy components,
select an n-tuple by comparing the determined point with respective (n-1)-polytopes of n-tuples of a plurality of tuples, wherein an (n-1)-polytope of a given n-tuple has vertices corresponding respectively to n material components in the given n-tuple, each vertex of the (n-1)-polytope having coordinates of attenuation values of a unit volume fraction of a corresponding material component of the given n-tuple,
when the determined point is within the (n-1)-polytope of the selected n-tuple, apply a Mendonça method to determine volume fractions of the n material components in the selected n-tuple as the material decomposition of the voxel, and,
when the determined point is not within the selected (n-1)-polytope, determine a closest (n-2)-polytope of the (n-1)-polytope of the selected n-tuple that is closest to the determined point, and decompose the attenuation values of the voxel into volume fractions of material components of an (n-1)-tuple corresponding to the closest (n-2)-polytope.
3. The apparatus according to claim 2 , wherein selecting the n-tuple by comparing the determined point with the respective (n-1)-polytopes is performed by the circuitry being further configured to select the selected n-tuple by comparing Hausdorff distances between the determined point and the respective (n-1)-polytopes and selecting, as the selected n-tuple, an n-tuple corresponding to a smallest Hausdorff distance between the determined point and the corresponding (n-1)-polytope.
4. The apparatus according to claim 2 , wherein selecting the n-tuple by comparing the determined point with the respective (n-1)-polytopes is performed by the circuitry being further configured to select the selected n-tuple by determining a (n-1)-polytope that either (i) includes the determined point or, if none of the respective (n-1)-polytopes includes the determined point, (ii) includes a point within the respective (n-1)-polytopes that is closest to the determined point.
5. The apparatus according to claim 4 , wherein, if multiple (n-1)-polytopes of the respective (n-1)-polytopes either includes the determined point or includes the point that is closest to the determined point, the selected n-tuple corresponds to a smallest (n-1)-polytopes of the multiple (n-1)-polytopes.
6. The apparatus according to claim 1 , wherein
the volume-fraction condition is that (i) volume fractions of the material components in the voxel are non-negative and (ii) a sum of the volume fractions for the voxel is one, and,
to determine whether the attenuation values of the voxel are consistent with the volume-fraction condition, the circuitry is further configured to
determine a point in an attenuation-energy space having coordinates given by the attenuation values of the voxel, the attenuation-energy space representing, along respective axes, attenuation values corresponding to respective components of the plurality of energy components,
determine whether the determined point is within an enclosed region defined by a union of respective (n-1)-polytopes in the attenuation-energy space corresponding to respective n-tuples of material components, wherein an (n-1)-polytope of a given n-tuple has vertices corresponding to respective material components in the given n-tuple, a respective vertex of the (n-1)-polytope having coordinates of attenuation values of a unit volume fraction of a corresponding material component of the given n-tuple,
when the determined point is within the enclosed region, apply a Mendonça method to determine volume fractions of the n material components in a selected n-tuple that corresponds to an (n-1)-polytope of the respective (n-1)-polytopes that is closest to the determined point, and,
when the determined point is not within the enclosed region, determine a closest (n-2)-polytope within the respective (n-1)-polytopes that is closest to the determined point, and decompose the attenuation values of the voxel into volume fractions of material components of an (n-1)-tuple corresponding to the closest (n-2)-polytope.
7. The apparatus according to claim 2 , wherein
the projected point is an orthogonal projection of the determined point onto a (n-2)-dimensional space of the closest (n-2)-polytope, and,
the circuitry is further configured to, when 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,
determine a closest (n-3)-polytope to the determined point, the closest (n-3)-polytope being a (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point, and
decompose the attenuation values of the voxel into volume fractions of (n-2) material components corresponding to the closest (n-3)-polytope.
8. The apparatus according to claim 6 , wherein
the projected point is an orthogonal projection of the determined point onto a (n-2)-dimensional space of the closest (n-2)-polytope, and,
when the determined point is not within the enclosed region and the projected point is not within the closest (n-2)-polytope, the circuitry is further configured to
determine a closest (n-3)-polytope to the determined point, the closest (n-3)-polytope being a (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point, and
decompose the attenuation values of the voxel into volume fractions of (n-2) material components corresponding to the closest (n-3)-polytope.
9. The apparatus according to claim 2 , wherein
n is three or four,
when n is four, the (n-1)-polytopes are triangles, the (n-2)-polytope is an edge, the selected n-tuple is a triad of the material components, and the determined (n-1)-tuple is a pair of the material components, and,
when n is four, the (n-1)-polytopes are triangular pyramids, the (n-2)-polytope is a triangle, the selected n-tuple is a tetrad of the material components, and the determined (n-1)-tuple is a triad of the material components.
10. The apparatus according to claim 1 , wherein, to obtain the images, the circuitry is configured to
acquire projection data corresponding to the energy components for each detector element of a plurality of detector elements, the energy components representing radiation detected at the plurality of detector elements, and
perform computed tomography on the projection data, reconstructing the images, each of the images corresponding to one of the respective energy components.
11. An apparatus, comprising:
an X-ray source configured to transmit X-rays;
a plurality of detector elements, wherein the plurality of detector elements is configured to
detect a plurality of energy components of the X-rays, which are transmitted from the X-ray source through an object and are detected at the plurality of detector elements, and
generate projection data having a plurality of energy components for each detector element of the plurality of detector elements, the energy components representing radiation of a respective energy profile; and
circuitry configured to
reconstruct images for the projection data, each of the images corresponding to a respective component of the plurality of energy components,
perform a material decomposition on the images to decompose attenuation values for the respective energy components of a voxel of the images into material components, wherein,
to perform the material decomposition, the circuitry is further configured to
determine whether the attenuation values of the voxel are consistent with a volume-fraction condition,
when the attenuation values of the voxel are consistent with the volume-fraction condition, decompose the attenuation values into n material components, wherein n is a natural number of three or greater, and,
when the attenuation values of the voxel are not consistent with the volume-fraction condition, project the attenuation values of the voxel onto a closest (n-2) dimensional space of n-1 material components to the attenuation values of the voxel to generate a projected point, and decompose attenuation values of the projected point into the n-1 material components.
12. A method, comprising:
obtaining images corresponding to respective energy components, and
performing a material decomposition on the images to decompose attenuation values for the respective energy components of a voxel of the images into material components, wherein the material decomposition includes
determining whether the attenuation values of the voxel are consistent with a volume-fraction condition,
when the attenuation values of the voxel are consistent with the volume-fraction condition, decomposing the attenuation values into n material components, wherein n is a natural number of three or greater, and,
when the attenuation values of the voxel are not consistent with the volume-fraction condition, projecting the attenuation values of the voxel onto a closest (n-2) dimensional space of n-1 material components to the attenuation values of the voxel to generate a projected point, and decomposing attenuation values of the projected point into the n-1 material components.
13. The method according to claim 12 , wherein
the volume-fraction condition is that (i) volume fractions of the material components in the voxel are non-negative and (ii) a sum of the volume fractions for the voxel is one, and
the step of determining whether the attenuation values of the voxel are consistent with the volume-fraction condition further comprises
determining a point in an attenuation-energy space having coordinates given by the attenuation values of the voxel, the attenuation-energy space representing, along respective axes, attenuation values corresponding to respective components of the plurality of energy components,
selecting an n-tuple by comparing the determined point with respective (n-1)-polytopes of n-tuples of a plurality of tuples, wherein an (n-1)-polytope of a given n-tuple has vertices corresponding respectively to n material components in the given n-tuple, each vertex of the (n-1)-polytope having coordinates of attenuation values of a unit volume fraction of a corresponding material component of the given n-tuple,
when the determined point is within the selected (n-1)-polytope, applying a Mendonça method to determine volume fractions of the n material components in the selected n-tuple as the material decomposition of the voxel, and,
when the determined point is not within the selected (n-1)-polytope, determining a closest (n-2)-polytope of the (n-1)-polytope of the selected n-tuple that is closest to the determined point, and decomposing the attenuation values of the voxel into volume fractions of material components of an (n-1)-tuple corresponding to the closest (n-2)-polytope.
14. The method according to claim 13 , wherein selecting the n-tuple further comprises comparing Hausdorff distances between the determined point and the respective (n-1)-polytopes and selecting, as the selected n-tuple, an n-tuple corresponding to a smallest Hausdorff distance between the determined point and the corresponding (n-1)-polytope.
15. The method according to claim 12 , wherein selecting the n-tuple further comprises comparing the determined point with the respective (n-1)-polytopes is performed by the circuitry being further configured to select the selected n-tuple by determining a (n-1)-polytope that either (i) includes the determined point or, if none of the respective (n-1)-polytopes includes the determined point, (ii) includes a point within the respective (n-1)-polytopes that is closest to the determined point.
16. The method according to claim 12 , wherein
the volume-fraction condition is that (i) volume fractions of the material components in the voxel are non-negative and (ii) a sum of the volume fractions for the voxel is one, and
the step of determining whether the attenuation values of the voxel are consistent with the volume-fraction condition further comprises
determining a point in an attenuation-energy space having coordinates given by the attenuation values of the voxel, the attenuation-energy space representing, along respective axes, attenuation values corresponding to respective components of the plurality of energy components,
determining whether the determined point is within an enclosed region defined by a union of respective (n-1)-polytopes in the attenuation-energy space corresponding to respective n-tuples of material components, wherein an (n-1)-polytope of a given n-tuple has vertices corresponding to respective material components in the given n-tuple, a respective vertex of the (n-1)-polytope having coordinates of attenuation values of a unit volume fraction of a corresponding material component of the given n-tuple,
when the determined point is within the enclosed region, applying a Mendonça method to determine volume fractions of the n material components in a selected n-tuple that corresponds to an (n-1)-polytope of the respective (n-1)-polytopes that is closest to the determined point, and,
when the determined point is not within the enclosed region, determining a closest (n-2)-polytope within the respective (n-1)-polytopes that is closest to the determined point, and decomposing the attenuation values of the voxel into volume fractions of material components of an (n-1)-tuple corresponding to the closest (n-2)-polytope.
17. The method according to claim 16 , wherein
the projected point is an orthogonal projection of the determined point onto a (n-2)-dimensional space of the closest (n-2)-polytope, and
the method further comprises, when the determined point is not within the enclosed region and the projected point is not within the closest (n-2)-polytope,
determining a closest (n-3)-polytope to the determined point, the closest (n-3)-polytope being a (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point, and
decomposing the attenuation values of the voxel into volume fractions of (n-2) material components corresponding to the closest (n-3)-polytope.
18. The method according to claim 13 , wherein
the projected point is an orthogonal projection of the determined point onto a (n-2)-dimensional space of the closest (n-2)-polytope, and
the method further comprises, when 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,
determining a closest (n-3)-polytope to the determined point, the closest (n-3)-polytope being a (n-3)-polytope within the closest (n-2)-polytope that is closest to the determined point, and
decomposing the attenuation values of the voxel into volume fractions of (n-2) material components corresponding to the closest (n-3)-polytope.
19. The method according to claim 12 , wherein the step of obtaining the images further comprises
acquiring projection data corresponding to the energy components for each detector element of a plurality of detector elements, the energy components representing radiation detected at the plurality of detector elements, and
performing computed tomography on the projection data, reconstructing the images, each of the images corresponding to one of the respective energy components.
20. A non-transitory computer readable storage medium including executable instruction, wherein the instructions, when executed by circuitry, cause the circuitry to perform the method according to claim 12 .