CT imaging methods and systems
Disclosed is a CT imaging method and system. The method includes: CT scanning an object with a dual-energy CT system to obtain a first complete set of projection data in a first scan mode, and to obtain a second incomplete set of projection data in a second scan mode; reconstructing a first attenuation coefficient image of the object from the first set of projection data, and extracting, from the first attenuation coefficient image, prior structure information of the object indicating edge intensity; and reconstructing a second attenuation coefficient image of the object from the second incomplete set of projection data using the extracted prior structure information as a constraint. With the method using the prior structure information of the imaged object as a constraint in reconstruction, it is possible to dramatically reduce an amount of data required for reconstruction, and achieve satisfactory effects even with ill-conditioned problems of limited-angle and inner reconstruction.
1. A CT imaging method, comprising:
CT scanning an object with a dual-energy CT system to obtain a first complete set of projection data at a first energy level, and to obtain a second incomplete set of projection data at a second energy level different from the first energy level;
reconstructing a first image of the object from the first set of projection data, and extracting, from the first image, prior structure information of the object indicating edge intensity, wherein the extracted prior structure information describes boundaries and details having obvious difference in the first image; and
reconstructing a second image of the object from the second incomplete set of projection data using the extracted prior structure information as a constraint.
2. The method according to claim 1 , wherein
the dual-energy CT system comprises multiple rows of low-energy detectors, and high-energy detectors disposed behind part of the rows of low-energy detectors;
said obtaining a first complete set of projection data in a first scan mode comprises performing a 360-degree circular or helical CT scan on the object with rays at a first energy level, to obtain the first complete set of projection data;
said obtaining a second incomplete set of projection data in a second scan mode comprises performing a limited-angle CT scan on the object with rays at a second energy level different from the first energy level, to obtain the second incomplete set of projection data.
3. The method according to claim 1 , wherein
the dual-energy CT system comprises multiple low-energy detectors, and multiple high-energy detectors disposed behind the low-energy detectors;
said obtaining a first complete set of projection data in a first scan mode comprises performing a 360-degree circular or helical CT scan on the object with rays at a first energy level, to obtain the first complete set of projection data;
said obtaining a second incomplete set of projection data in a second scan mode comprises performing a sparse-angle sampling CT scan on the object with rays at a second energy level different from the first energy level, to obtain the second incomplete set of projection data.
4. The method according to claim 1 , wherein
the dual-energy CT system comprises at least one row of low-energy detectors, and a plurality of high-energy detectors which are uniformly provided behind part of the low-energy detectors;
said obtaining a first complete set of projection data in a first scan mode comprises performing a 360-degree circular or helical CT scan on the object with rays at a first energy level, to obtain the first complete set of projection data;
said obtaining a second incomplete set of projection data in a second scan mode comprises performing a detector undersampling CT scan on the object with rays at a second energy level different from the first energy level, to obtain the second incomplete set of projection data.
5. The method according to claim 1 , wherein
the dual-energy CT system comprises at least one row of low-energy detectors, and a plurality of high-energy detectors which are concentrated and provided behind part of the low-energy detectors;
said obtaining a first complete set of projection data in a first scan mode comprises performing a 360-degree circular or helical CT scan on the object with rays at a first energy level, to obtain the first complete set of projection data;
said obtaining a second incomplete set of projection data in a second scan mode comprises performing an inner reconstruction CT scan on the object with rays at a second energy level different from the first energy level, to obtain the second incomplete set of projection data.
6. The method according to claim 1 , wherein said extracting prior structure information of the object from the first image comprises performing edge extraction on the first image to obtain the prior structure information.
7. The method according to claim 6 , wherein said reconstructing a second image of the object from the second incomplete set of projection data comprises:
calculating the second image f according to an equation
min
G
∇
f
1
+
λ
∇
f
1
=
min
(
G
+
λ
I
)
∇
f
1
s
.
t
.
Hf
-
p
W
≤
ɛ
wherein ε is a quantity related to an overall noise level in the second set of projection data, and λ is used to balance a prior structure information constraint and a Total Variation (TV) constraint; the prior structure information constraint is dominant when λ≦1, while minimization of the TV is a dominant constraint when λ>1; the second image is denoted as f={f 1 , f 2 , . . . , f n }, the second set of projection data obtained by the CT scan is denoted as p={p 1 , p 2 , . . . , p m }, a line integral projection process is denoted as
H
{
h
i
j
}
m
×
n
,
the prior structure information is denoted as g={g 1 , g 2 , . . . , g n }, W is a weighted matrix reflecting noise characteristic of projection data, I is identity matrix, and a weighted matrix is generated from the prior structure information as G=diag(g)=diag{g 1 , g 2 , . . . , g n }; a larger g i represents a weaker edge, while a smaller g i represents a stronger edge.
8. The method according to claim 6 , wherein said edge extraction is performed using one of Laplace operator, Candy operator, discrete gradient transform, or adaptive thresholding.
9. A CT imaging method, comprising:
extracting prior structure information of an object from a first image reflecting an internal structure of the object, wherein the first image is a design drawing of the object, the prior structure information indicates edge intensity, and the extracted prior structure information describes boundaries and details having obvious difference in the first image;
CT scanning the object with a CT system to obtain a set of projection data; and
reconstructing a second image of the object from the set of projection data using the extracted prior structure information as a constraint.
10. The method according to claim 9 , wherein the CT system performs a limited-angle CT scan on the object to obtain the set of projection data.
11. The method according to claim 9 , wherein the CT system performs a sparse-angle sampling CT scan on the object to obtain the set of projection data.
12. The method according to claim 9 , wherein the CT system performs a detector undersampling CT scan on the object to obtain the set of projection data.
13. The method according to claim 9 , wherein the CT system performs an inner reconstruction CT scan on the object to obtain the set of projection data.
14. The method according to claim 9 , wherein said extracting prior structure information of the object from the first image comprises performing edge extraction on the first image to obtain the prior structure information.
15. The method according to claim 14 , wherein said reconstructing a second image of the object from the set of projection data comprises:
calculating the second image f according to an equation
min
G
∇
f
1
+
λ
∇
f
1
=
min
(
G
+
λ
I
)
∇
f
1
s
.
t
.
Hf
-
p
W
≤
ɛ
wherein εis a quantity related to an overall noise level in the second set of projection data, and λ is used to balance a prior structure information constraint and a Total Variation (TV) constraint; the prior structure information constraint is dominant when λ≦1, while minimization of the TV is a dominant constraint when λ>1; the second image is denoted as f={f 1 , f 2 , . . . , f n }, the second set of projection data obtained by the CT scan is denoted as p={p 1 , p 2 , . . . , p m }, a line integral projection process is denoted as
H
{
h
i
j
}
m
×
n
,
the prior structure information is denoted as g={g 1 , g 2 , . . . , g n }, W is a weighted martrix reflecting noise characteristic of projection data, I is identity matrix, and a weighted matrix is generated from the prior structure information as G=diag(g)=diag{g 1 , g 2 , . . . , g n }; a larger g i represents a weaker edge, while a smaller g i represents a stronger edge.
16. The method according to claim 14 , wherein said edge extraction is performed using one of Laplace operator, Candy operator, discrete gradient transform, or adaptive thresholding.
17. The method according to claim 9 , wherein the first image is an image reconstructed from a previous CT scan of the object, or a design drawing of the object.
18. A CT imaging system, comprising:
a ray source configured to generate dual-energy X rays;
a detection & collection device configured to receive dual-energy X rays penetrating an object;
a control device configured to control the ray source and the detection & collection device to CT scan the object to obtain a first complete set of projection data at a first energy level, and to obtain a second incomplete set of projection data at a second energy level different from the first energy level; and
a reconstruction device configured to reconstruct a first image of the object from the first set of projection data, extract, from the first image, prior structure information of the object indicating edge intensity, and reconstruct a second image of the object from the second incomplete set of projection data using the extracted prior structure information as a constraint, wherein the extracted prior structure information describes boundaries and details having obvious difference in the first image.
19. A CT imaging system, comprising:
a ray source configured to generate dual-energy X rays;
a detection & collection device configured to receive dual-energy X rays penetrating an object; a control device configured to control the ray source and the detection & collection device to CT scan the object to obtain a set of projection data; and
a reconstruction device configured to extract prior structure information of the object from a first image reflecting an internal structure of the object, wherein the first image is a design drawing of the object, the prior structure information indicates edge intensity, and reconstruct a second image of the object from the set of projection data using the extracted prior structure information as a constraint, wherein the extracted prior structure information describes boundaries and details having obvious difference in the first image.
20. The system according to claim 19 , wherein the first image is an image reconstructed from a previous CT scan of the object, or a design drawing of the object.