Systems and methods for computer-aided fold detection
View Patent ↗Described herein is a computer-aided technology for facilitating detection of folds of an object surface. In one implementation, image data is processed to determine curvature characteristics at one or more points of the surface of a region of interest. If the curvature characteristics of one of the one or more points correspond to a hyperbolic curvature, the point is identified as a seed point. A region of a predetermined thickness is grown around the seed point so as to detect the presence of any fold feature within the grown region.
1. A method of detecting folds in a digitized medical image, comprising:
receiving image data including a region of interest;
calculating curvature characteristics at one or more points corresponding to a surface of the region of interest;
identifying at least one of the one or more points as a seed point if its curvature characteristics correspond to a hyperbolic curvature; and
detecting a fold by growing a region around the seed point to a predetermined thickness and identifying at least one fold feature within the region.
2. The method of claim 1 wherein the region of interest comprises a portion of a colon.
3. The method of claim 1 wherein the curvature characteristics comprise a minimum curvature, a maximum curvature and a principal direction.
4. The method of claim 1 wherein the identifying the at least one of the one or more points comprises identifying the at least one of the one or more points as a seed point if its minimum curvature is less than a first threshold value and its maximum curvature more than a second threshold value.
5. The method of claim 1 wherein the growing comprises finding voxels adjacent to the seed point having intensity values within a predetermined range identified for a material of interest.
6. The method of claim 5 wherein the material of interest comprises tissue.
7. The method of claim 1 wherein the growing comprises growing the region to different predetermined thicknesses to detect multiple fold features.
8. The method of claim 7 wherein the fold features comprise a ridge, a face, a base or an extended base of the fold.
9. The method of claim 1 wherein the fold feature comprises a ridge of the fold.
10. The method of claim 1 wherein the fold feature comprises a face of the fold.
11. The method of claim 1 wherein the fold feature comprises a base or an extended base of the fold.
12. The method of claim 1 wherein the identifying the at least one fold feature comprises detecting the fold feature within one or more local neighborhoods within the region, wherein the one or more local neighborhoods are associated with a material composition that satisfies a pre-defined criterion.
13. The method of claim 12 wherein the one or more local neighborhoods comprises a sphere centered at the seed point.
14. The method of claim 12 wherein the material composition comprises a ratio of air and tissue.
15. The method of claim 14 wherein the pre-defined criterion comprises the ratio being less than a third threshold value.
16. The method of claim 12 wherein the one or more local neighborhoods are defined by different radii to detect multiple fold features.
17. The method of claim 1 wherein the identifying the at least one fold feature comprises detecting the fold feature within one or more local neighborhoods within the region, wherein the one or more local neighborhoods are associated with a minimum tissue thickness that is within a predetermined range.
18. The method of claim 1 wherein the identifying the at least one fold feature comprises detecting the fold feature within one or more local neighborhoods within the region, wherein the one or more local neighborhoods are associated with a distribution of tissue thicknesses, wherein a difference between the distribution of tissue thicknesses and a predetermined distribution is less than a fourth threshold value.
19. A non-transitory computer readable medium embodying a program of instructions executable by a machine to perform steps for detecting folds within a digitized medical image, the steps comprising:
receiving image data including a region of interest;
calculating curvature characteristics at one or more points corresponding to a surface of the region of interest;
identifying at least one of the one or more points as a seed point if its curvature characteristics correspond to a hyperbolic curvature; and
detecting a fold by growing a region around the seed point to a predetermined thickness and identifying at least one fold feature within the region.
20. An image analysis system, comprising:
a memory device for storing computer readable program code; and
a processor in communication with the memory device, the processor being operative with the computer readable program code to:
receive image data including a region of interest;
calculate curvature characteristics at one or more points corresponding to a surface of the region of interest;
identify at least one of the one or more points as a seed point if its curvature characteristics correspond to a hyperbolic curvature; and
detect a fold by growing a region around the seed point to a predetermined thickness and identifying at least one fold feature within the region.