System and method for characterizing anatomical features

Patent metadata

Field	Value
Publication number	US-10448915-B2
Application number	US-201715633819-A
Country	US
Kind code	B2
Filing date	Jun 27, 2017
Priority date	Jun 27, 2017
Publication date	Oct 22, 2019
Grant date	Oct 22, 2019

Abstract

A method for characterizing anatomical features includes receiving scanned data and image data corresponding to a subject. The scanned data comprises sinogram data. The method further includes identifying a first region in an image of the image data corresponding to a region of interest. The method also includes determining a second region in the scanned data. The second region corresponds to the first region. The method further includes identifying a sinogram trace corresponding to the region of interest. The sinogram trace comprises sinogram data present within the second region. The method includes determining a data feature of the subject based on the sinogram trace and a deep learning network. The method also includes determining a diagnostic condition corresponding to a medical condition of the subject based on the data feature.

First claim

The invention claimed is:

1. A method, comprising:
receiving scanned data and image data corresponding to a subject, wherein the scanned data comprises sinogram data;
identifying a first region in an image of the image data corresponding to a region of interest;
determining a second region in the scanned data, wherein the second region corresponds to the first region;
identifying a sinogram trace corresponding to the region of interest, wherein the sinogram trace comprises sinogram data present within the second region;
characterizing an anatomical feature corresponding to the region of interest of the subject based on the sinogram trace and a trained deep learning network; and
determining a diagnostic condition corresponding to a medical condition of the subject based on the anatomical feature, wherein identifying the sinogram comprises projecting the image data present within the first region using a Radon transformation.

2. The method of claim 1 , further comprising processing the image data present within the first region using the trained deep learning network.

3. The method of claim 1 , wherein the image data is generated by reconstructing the scanned data based on input from an operator on a display device.

4. The method of claim 1 , wherein identifying the sinogram trace comprises:
applying an image mask to the image data present within the first region to generate a background image, wherein the background image does not comprise the region of interest;
determining a second sinogram trace by projecting the background image; and
generating a third sinogram trace as a difference of a first sinogram trace and the second sinogram trace, wherein the third sinogram trace is representative of a sinogram of the region of interest.

5. The method of claim 1 , wherein the first region is determined by an automated segmentation technique.

6. The method of claim 1 , wherein the sinogram data further comprises photon count data obtained from a photon-counting CT detector.

7. The method of claim 1 , wherein the anatomical feature comprises at least one of a tumor, a stenosis condition, a plaque, a nodule, a lesion, a bleeding condition and a motion field.

8. The method of claim 7 , wherein the anatomical feature further comprises at least one of a quantitative measure, a descriptive measure, and a classification of the diagnostic condition.

9. The method of claim 1 , wherein determining the anatomical feature comprises training the deep learning network using a plurality of simulated sinogram traces having known anatomical features.

10. A system, comprising:
a data acquisition unit configured to acquire scanned data and image data corresponding to a subject, from a computed tomography detector, wherein the scanned data comprises sinogram data;
an image processor communicatively coupled to the data acquisition unit and configured to:
identify a first region in an image of the image data corresponding to a region of interest in the subject;
determine a second region in the scanned data, wherein the second region corresponds to the first region;
determine a sinogram trace corresponding to the region of interest, wherein the sinogram trace comprises sinogram data present within the second region, wherein identifying the sinogram comprises projecting the image data present within the first region using a Radon transformation;

a trained deep learning network communicatively coupled to the image processor and configured to:
characterize an anatomical feature corresponding to the region of interest of the subject based on the sinogram trace; and
determine a diagnostic condition corresponding to a medical condition of the subject based on the anatomical feature.

11. The system of claim 10 , wherein the trained deep learning network is further configured to process the image data present within the first region to characterize the anatomical feature of the subject.

12. The system of claim 10 , wherein the image processor is further configured to:
apply an image mask to the image data present within the first region to generate a background image, wherein the background image does not comprise the region of interest;
determine a second sinogram trace by projecting the background image; and
generate a third sinogram trace as a difference of a first sinogram trace and the second sinogram trace, wherein the third sinogram trace is representative of a sinogram of the region of interest.

13. The system of claim 10 , wherein the image processor is further configured to generate the image data by reconstructing the scanned data based on input from an operator on a display device.

14. The system of claim 13 , wherein the image processor is further configured to determine the image data present within the first region using an automated segmentation technique.

15. The system of claim 10 , wherein the trained deep learning network is configured to detect at least one of a tumor condition, a stenosis condition, a plaque, a nodule, a lesion, a bleeding condition, and a motion field.

16. The system of claim 15 , wherein the trained deep learning network is further configured to determine at least one of a quantitative measure, a descriptive measure, and a classification of the diagnostic condition.

17. A computed tomography imaging system, comprising:
an x-ray source configured to emit x-ray beam towards an organ during examination of a subject;
a computed tomography detector configured to receive the emitted x-ray beam attenuated by the region to generate scanned data;
a data acquisition unit configured to acquire the scanned data from the computed tomography detector, wherein the scanned data comprises sinogram data;
an image processor communicatively coupled to the data acquisition unit and configured to:
generate image data by reconstructing the scanned data based on input from an operator on a display device;
identify a first region in an image of the image data corresponding to a region of interest in the subject;
determine a second region in the scanned data, wherein the second region corresponds to the first region;
determine a sinogram trace corresponding to the region of interest, wherein the sinogram trace comprises sinogram data present within the second region, wherein identifying the sinogram comprises projecting the image data present within the first region using a Radon transformation;

a trained deep learning network communicatively coupled to the image processor and configured to:
characterize an anatomical feature of the region based on the sinogram trace; and
determine a diagnostic condition corresponding to a medical condition of the subject based on the anatomical feature.

18. A non-transitory computer readable medium having instructions to enable at least one processor module to:
receive scanned data and image data corresponding to a subject, wherein the scanned data comprises sinogram data and the image data is generated by reconstructing the scanned data based on input from an operator on a display device;
identify a first region in an image of the image data corresponding to a region of interest in the subject;
determine a second region in the scanned data, wherein the second region corresponds to the first region;
determine a sinogram trace corresponding to the region of interest, wherein the sinogram trace comprises sinogram data present within the second region;
characterize an anatomical feature corresponding to the region of interest of the subject based on the sinogram trace and a trained deep learning netw

Assignees

• Gen Electric

Inventors

• De Man Bruno Kristiaan Bernard
• Pack Jed Douglas
• Haneda Eri
• Ramani Sathish
• Hsieh Jiang
• Miller James Vradenburg
• Edic Peter Michael

Classifications

• G16H50/30

  for calculating health indices; for individual health risk assessment · CPC title

• G06T7/0012

  Biomedical image inspection · CPC title

• A61B6/5205

  involving processing of raw data to produce diagnostic data · CPC title

• G06T2207/10072

  Tomographic images · CPC title

• A61B5/055

  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging · CPC title