System and method for reconstructing image by using straight-line trajectory scan

• US 20070116175A1
• Filed: 10/12/2006
• Published: 05/24/2007
• Est. Priority Date: 11/21/2005
• Status: Active Grant

• Current Assignee: Nuctech Company Limited (Tsinghua Tongfang Co., Ltd.), Tsinghua University
• Original Assignee: Nuctech Company Limited (Tsinghua Tongfang Co., Ltd.), Tsinghua University
• Inventors: Liu, Yinong, Li, Yuanjing, Xing, Yuxiang, Zhang, Li, Cheng, Jianping, Xiao, Yongshun, Chen, Zhiqiang, Zhao, Ziran, Kang, Kejun, Gao, Hewei
• Granted Patent: US 7,424,089 B2
• Time in Patent Office: Days
• US Class Current: 378/21
• CPC Class Codes: A61B 6/032 Transmission computed tomography [CT]
• CPC Class Codes: G01N 2223/3303 object fixed; source and detector move
• CPC Class Codes: G01N 2223/419 computed tomograph
• CPC Class Codes: G01N 23/046 using tomography, e.g. computed tomography [CT]
• CPC Class Codes: G06T 11/006 Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
• CPC Class Codes: G06T 2211/421 Filtered back projection [FBP]

Abstract

It is disclosed a system and a method for reconstructing an image by using a straight-line trajectory scan to avoid image spatial resolution reduction due to interpolations in angular direction and detector direction during data rebinning. This system comprises: a projection data conversion section for converting projection data from straight-line trajectory scan into projection data under quasi-parallel-beam scan; a filtration section for obtaining filtered projection data by convoluting the projection data under quasi-parallel-beam scan with a predetermined convolutional kernel; and a back-projection section for reconstructing an image by back-projecting the filtered projection data with a weighting factor. By using the inventive system and method, the spatial resolution in the reconstructed image is improved, and the influence of data truncation on the reconstructed image is reduced. The present invention applies the filtration and back-projection mode, and thus has general advantages of the filtration and back projection, such as simplicity and efficiency. And it is easy to be parallelized and accelerated.

Claims

1. A system for reconstructing an image by using a straight-line trajectory scan, comprising:

a projection data conversion section for converting projection data from straight-line trajectory scan into projection data under quasi-parallel-beam scan;

a filtration section for obtaining filtered projection data by convoluting the projection data under quasi-parallel-beam scan with a predetermined convolutional kernel; and

a back-projection section for reconstructing an image by back-projecting the filtered projection data with a weighting factor.

2. The system for reconstructing an image according to claim 1, further comprising:

a detector matrix comprising a plurality of detector elements for receiving transmitted signals caused by radiations which are emitted from a radiation source and penetrate through an object to be inspected, and for converting the transmitted radiations into the projection data.

3. The system for reconstructing an image according to claim 2, wherein the plurality of detector elements are arranged in accordance with one and the same spacing interval.

4. The system for reconstructing an image according to claim 3, wherein:

the projection data conversion section reverses and shifts the projection data p(l, t, z) to obtain the projection data q(l,t,z) under quasi-parallel-beam scan, wherein the projection data p(l,t,z) denotes a projection value at a coordinate of t in the z^th slice of the detector matrix when the object to be inspected relatively moves to a coordinate of l on the line;

the filtration section performs one-dimension convolution of the projection data q(l,t,z) under quasi-parallel-beam scan with the predetermined convolutional kernel in the l direction to obtain the filtered projection data Q(l′

,t,z); and

the back-projection section back-projects the filtered projection data Q(l′

,t, z) with a weighting factor along the radiation projection direction to obtain the reconstructed image.

5. The system for reconstructing an image according to claim 2, wherein the plurality of detector elements are arranged with regard to the radiation source in accordance with one and the same angular interval.

6. The system for reconstructing an image according to claim 5, wherein:

the projection data conversion section reverses and shifts the projection data p(l,γ

,z) to obtain the projection data q(l,γ

,z) under quasi-parallel-beam scan, wherein the projection data p(l,γ

, z) denotes a projection value at an angular position of γ

in the z^th slice of the detector matrix when the object to be inspected relatively moves to a coordinate of l on the line;

the filtration section performs one-dimension convolution of the projection data q(l,γ

,z) under quasi-parallel-beam scan with the predetermined convolutional kernel in the l direction to obtain the filtered projection data Q(l′

,γ

,z); and

the back-projection section back-projects the filtered projection data Q(l′

,γ

,z) with a weighting factor along the radiation projection direction to obtain the reconstructed image.

7. A method for reconstructing an image by using a straight-line trajectory scan, comprising:

a projection data conversion step for converting projection data from straight-line trajectory scan into projection data under quasi-parallel-beam scan;

a filtration step for obtaining filtered projection data by convoluting the projection data under quasi-parallel-beam scan with a predetermined convolutional kernel; and

a back-projection step for reconstructing an image by back-projecting the filtered projection data with a weighting factor.

8. The method for reconstructing an image according to claim 7, further comprising a step of:

receiving transmitted signals caused by radiations which are emitted from a radiation source and penetrate through an object to be inspected and converting the transmitted radiations into the projection data by a detector matrix including a plurality of detector elements.

9. The method for reconstructing an image according to claim 8, wherein the plurality of detector elements are arranged in accordance with one and the same spacing interval.

10. The method for reconstructing an image according to claim 9, wherein:

at the projection data conversion step, reversing and shifting the projection data p(l,t,z) to obtain the projection data q(l,t,z) under quasi-parallel-beam scan, wherein the projection data p(l,t,z) denotes a projection value at a coordinate of t in the z^th slice of the detector matrix when the object to be inspected relatively moves to a coordinate of l on the line;

at the filtration step, performing one-dimension convolution of the projection data q(l,t,z) under quasi-parallel-beam scan with the predetermined convolutional kernel in the l direction to obtain the filtered projection data Q(l′

,t,z); and

at the back-projection step, back-projecting the filtered projection data Q(l′

,t,z) with a weighting factor along the radiation projection direction to obtain the reconstructed image.

11. The method for reconstructing an image according to claim 8, wherein the plurality of detector elements are arranged with regard to the radiation source in accordance with one and the same angular interval.

12. The method for reconstructing an image according to claim 11, wherein:

at the projection data conversion step, reversing and shifting the projection data p(l,γ

,z) to obtain the projection data q(l,γ

,z) under quasi-parallel-beam scan, wherein the projection data p(l,γ

,z) denotes a projection value at an angular position of γ

in the z^th slice of the detector matrix when the object to be inspected relatively moves to a coordinate of l on the line;

at the filtration step, performing one-dimension convolution of the projection data q(l,γ

,z) under quasi-parallel-beam scan with the predetermined convolutional kernel in the l direction to obtain the filtered projection data Q(l′

,γ

,z); and

at the back-projection step, back-projecting the filtered projection data Q(l′

,γ

,z) with a weighting factor along the radiation projection direction to obtain the reconstructed image.