Feature validation using motion registration

• US 20150235364A1
• Filed: 02/18/2014
• Published: 08/20/2015
• Est. Priority Date: 02/18/2014
• Status: Active Grant

• Current Assignee: EchoPixel, Inc.
• Original Assignee: EchoPixel, Inc.
• Inventors: Aguirre-Valencia, Sergio
• Granted Patent: US 9,361,684 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06T 2207/10072 Tomographic images
• CPC Class Codes: G06T 2207/30032 Colon polyp
• CPC Class Codes: G06T 7/0014 using an image reference approach
• CPC Class Codes: G06T 7/33 using feature-based methods

Abstract

During an analysis technique, the locations of a polyp candidate (and, more generally, a feature) in images of the colon acquired with different orientations of an individual are compared to determine if the polyp candidate is a true positive (an actual polyp) or a false positive (an artifact). In particular, the locations are compared for images when the individual is supine and prone. During this rotation by approximately 180° (and, more generally, a symmetry operation), an actual polyp is expected to also be rotated by approximately 180°. Consequently, the location change during the symmetry operation can be used to validate whether the polyp candidate is an actual polyp. Because the colon is a non-rigid object (with degrees of freedom including rotation, compression and expansion), reference markers (such as haustral folds) are used to assist in determining the location of the polyp candidate in the images acquired in the different orientations.

Claims

1. A computer-implemented method for validating a feature, the method comprising:

receiving an image of a portion of an individual and a second image of the portion of the individual, wherein the image corresponds to an orientation of the individual and the second image corresponds to a second orientation of the individual;

determining reference markers in a non-rigid object in the image and second image;

using the computer, identifying, relative to the reference markers, a location of a feature in the non-rigid object in the image and a second location of the feature in the non-rigid object in the second image;

calculating a difference vector based on the location and the second location;

determining if the difference vector matches an expected location change based on the orientation and the second orientation;

if there is a match, identifying the object as a true positive; and

otherwise, annotating the object as a false positive.

2. The method of claim 1, wherein the orientation and the second orientation correspond to a symmetry operation on the individual.

3. The method of claim 2, wherein the symmetry operation includes one of:

rotation by an angle about an axis, translation and scaling.

4. The method of claim 3, wherein the angle includes approximately 180°

; and

wherein the orientation and the second orientation include one of;

supine and prone; and

prone and supine.

5. The method of claim 1, wherein the non-rigid object includes one of:

a portion of a small intestine, a portion of a large intestine, a portion of a non-rigid organ, a portion of the vascular system, and a portion of an organ that includes a cavity with a fluid.

6. The method of claim 1, wherein the non-rigid object includes the colon;

and

wherein the reference markers include haustral folds.

7. The method of claim 1, wherein the feature includes a polyp candidate.

8. The method of claim 1, wherein the non-rigid object has rotation, compression and extension degrees of freedom.

9. The method of claim 1, wherein a match corresponds to a sign of the difference vector.

10. A computer-program product for use in conjunction with a computer system, the computer-program product comprising a non-transitory computer-readable storage medium and a computer-program mechanism embedded therein to validate a feature, the computer-program mechanism including:

instructions for receiving an image of a portion of an individual and a second image of the portion of the individual, wherein the image corresponds to an orientation of the individual and the second image corresponds to a second orientation of the individual;

instructions for determining reference markers in a non-rigid object in the image and second image;

instructions for identifying, relative to the reference markers, a location of a feature in the non-rigid object in the image and a second location of the feature in the non-rigid object in the second image;

instructions for calculating a difference vector based on the location and the second location;

instructions for determining if the difference vector matches an expected location change based on the orientation and the second orientation;

if there is a match, instructions for identifying the object as a true positive; and

otherwise, instructions for identifying the object as a false positive.

11. The computer-program product of claim 10, wherein the orientation and the second orientation correspond to a symmetry operation on the individual.

12. The computer-program product of claim 11, wherein the symmetry operation includes one of:

rotation by an angle about an axis, translation and scaling.

13. The computer-program product of claim 12, wherein the angle includes approximately 180°

; and

wherein the orientation and the second orientation include one of;

supine and prone; and

prone and supine.

14. The computer-program product of claim 10, wherein the non-rigid object includes one of a portion of a small intestine, a portion of a large intestine, a portion of a non-rigid organ, a portion of the vascular system, and a portion of an organ that includes a cavity with a fluid.

15. The computer-program product of claim 10, wherein the non-rigid object includes the colon;

and

wherein the reference markers include haustral folds.

16. The computer-program product of claim 10, wherein the feature includes:

a polyp candidate.

17. The computer-program product of claim 10, wherein the non-rigid object has rotation, compression and extension degrees of freedom.

18. The computer-program product of claim 10, wherein a match corresponds to a sign of the difference vector.

19. A computer system, comprising:

a processor;

memory; and

a program module, wherein the program module is stored in the memory and configurable to be executed by the processor to validate a feature, the program module including;

instructions for receiving an image of a portion of an individual and a second image of the portion of the individual, wherein the image corresponds to an orientation of the individual and the second image corresponds to a second orientation of the individual;

instructions for determining reference markers in a non-rigid object in the image and second image;

instructions for identifying, relative to the reference markers, a location of a feature in the non-rigid object in the image and a second location of the feature in the non-rigid object in the second image;

instructions for calculating a difference vector based on the location and the second location;

instructions for determining if the difference vector matches an expected location change based on the orientation and the second orientation;

if there is a match, instructions for identifying the object as a true positive; and

otherwise, instructions for identifying the object as a false positive.

20. The computer system of claim 19, wherein the orientation and the second orientation correspond to a symmetry operation on the individual;

and

wherein the symmetry operation includes one of;

rotation by an angle about an axis, translation and scaling.