Processing a computed tomography image to reduce windmill artifacts

Patent metadata

Field	Value
Publication number	US-10068332-B2
Application number	US-201615178233-A
Country	US
Kind code	B2
Filing date	Jun 9, 2016
Priority date	Jun 11, 2015
Publication date	Sep 4, 2018
Grant date	Sep 4, 2018

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing a CT (Computed Tomography) image are provided. An example method includes accessing an original CT image that is reconstructed from a first set of raw data and includes windmill artifacts, generating a high-frequency image by processing the original CT image, generating a low-frequency image by processing a plurality of thick images reconstructed from a second set of raw data and combining the plurality of processed thick images, the second set of raw data including the first set of raw data and each of the plurality of thick images including substantially no windmill artifacts, generating an intermediate image by synthesizing the high-frequency image and the low-frequency image, and obtaining a target CT image based on the generated intermediate image.

First claim

The invention claimed is:

1. A method of processing a Computed Tomography (CT) image, comprising:
accessing an original CT image, wherein the original CT image is reconstructed from a first set of raw data and includes distortion representative of windmill artifacts;
generating a high-frequency image by processing the original CT image with a first frequency division process;
generating a low-frequency image by processing a plurality of thick images with a second frequency division process and combining the plurality of processed thick images, wherein the plurality of thick images is reconstructed from a second set of raw data that includes the first set of raw data, and each of the plurality of thick images includes substantially no distortion representative of windmill artifacts;
generating an intermediate image by synthesizing the high-frequency image and the low-frequency image; and
obtaining a target CT image based on the generated intermediate image.

2. The method of claim 1 , wherein generating a high-frequency image by processing the original CT image with a first frequency division process comprises:
generating frequency domain data for the original CT image by performing a Fourier transform on the original CT image;
extracting a high-frequency component from the generated frequency domain data; and
generating the high-frequency image by performing an inverse Fourier transform on the extracted high-frequency component.

3. The method of claim 2 , wherein extracting a high-frequency component from the generated frequency domain data comprises:
calculating a low-frequency weighting coefficient for each of one or more frequency positions in the generated frequency domain data;
calculating a low-frequency value for each of the one or more frequency positions according to a value for the corresponding frequency position in the frequency domain data and the corresponding calculated low-frequency weighting coefficient;
generating a high-frequency value for each of the one or more frequency positions by calculating a difference between the value for the frequency position and the corresponding low-frequency value; and
assembling the generated high-frequency values for the one or more frequency positions to constitute the high-frequency component in the frequency domain data of the original CT image.

4. The method of claim 1 , wherein generating a low-frequency image comprises:
determining a thick image reconstruction parameter;
reconstructing, according to the determined thick image reconstruction parameter, the plurality of thick images from the second set of raw data;
generating a plurality of low-frequency thick images by processing each of the reconstructed thick images with the second frequency division process; and
generating the low-frequency image by performing a sharpening combination on the plurality of generated low-frequency thick images.

5. The method of claim 4 , wherein the thick image reconstruction parameter comprises a reconstruction interval, an image thickness, and an image number,
wherein reconstructing the plurality of thick images comprises reconstructing the plurality of thick images from the second set of raw data along a scanning bed direction based on the reconstruction interval, a thickness of each of the reconstructed thick images being the same as the image thickness and a number of the reconstructed thick images being consistent with the image number.

6. The method of claim 4 , wherein generating a plurality of low-frequency thick images by processing each of the reconstructed thick images with the second frequency division process comprises:
generating frequency domain data for the thick image by performing a Fourier transform on the thick image;
extracting a low-frequency component from the generated frequency domain data; and
generating a corresponding low-frequency thick image of the thick image by performing an inverse Fourier transform on the extracted low-frequency component.

7. The method of claim 6 , wherein extracting a low-frequency component from the generated frequency domain data comprises:
calculating a low-frequency weighting coefficient for each of one or more frequency positions in the generated frequency domain data;
calculating a low-frequency value for each of the one or more frequency positions according to a value for the corresponding frequency position in the frequency domain data and the corresponding low-frequency weighting coefficient; and
assembling the calculated low-frequency values for the one or more frequency positions to constitute the low-frequency component in the frequency domain data of the thick image.

8. The method of claim 4 , wherein generating the low-frequency image by performing a sharpening combination on the plurality of generated low-frequency thick images comprises:
determining a corresponding weighting for each of the plurality of low-frequency thick images to be combined;
relating, for each of the plurality of low-frequency thick images, a corresponding pixel value to the determined corresponding weighting to generate a corresponding weighted pixel value; and
accumulating the weighted pixel values corresponding to an identical pixel of the plurality of low-frequency thick images to generate an accumulated pixel value corresponding to the same pixel of the low-frequency image.

9. The method of claim 1 , wherein synthesizing the high-frequency image and the low-frequency image comprises one of:
adding pixel values corresponding to an identical pixel of the high-frequency image and the low-frequency image to generate a pixel value corresponding to the same pixel of the intermediate image, and generating pixel values for pixels of the intermediate image, and
adding values for an identical frequency position in frequency domain data of the high-frequency image and frequency domain data of the low-frequency image together to generate a value for the same frequency position in frequency domain data of the intermediate image, generating the frequency domain data of the intermediate image based on the generated values for the frequency positions in the frequency domain data, and generating the intermediate image by performing an inverse Fourier transform on the generated frequency domain data.

10. The method of claim 1 , wherein obtaining a target CT image based on the intermediate image comprises:
determining a confidence parameter according to a difference between the intermediate image and the original CT image; and
correcting the intermediate image according to the determined confidence parameter and the original CT image to generate a corrected intermediate image as the target CT image.

11. The method of claim 1 , wherein the second set of raw data including the first set of raw data is obtained by a detector of a CT scan device in a CT scan for a subject,
wherein the first set of raw data corresponds to a first scanned region of the subject, the second set of raw data corresponds to a second scanned region of the subject, and the second scanned region covers the first scanned region along a scanning bed direction, and
wherein a first anatomy thickness of the first scanned region along the scanning bed direction is smaller than a second anatomy thickness of the second scanned region along the scanning bed direction.

12. A CT image processing device comprising:
a processor which invokes machine readable instructions corresponding to a CT image processing logic stored on a storage medium and executes the machine readable instructions to:
access an original CT image, wherein the original CT image is reconstructed from a first set of raw data and includes dist

Assignees

• Shenyang Neusoft Medical Sys

Inventors

• Zheng Han
• Lou Shanshan

Classifications

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• A61B6/032 Primary

  Transmission computed tomography [CT] · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06T12/30

  Image post-processing, e.g. metal artefact correction · CPC title

• G06T12/00

  Tomographic reconstruction from projections · CPC title