Method and system of generating a 3D visualization from 2D images

1. A method for generating a three dimensional visualization from a plurality of two dimensional images obtained by a transverse scan, the method comprising the steps of:

generating a plurality of transverse scans of a target area with an image generation device;

segmenting a region of interest in each image by:

projecting a plurality of radial profiles at angularly equi-spaced radii from a lumen centroid to an edge of the lumen, wherein end points of the plurality of radial profiles form a closed contour;

modeling the lumen using a multi-level thresholding algorithm to identify a plurality of pixels of the image which are within the region of interest;

obtaining an ellipse to encapsulate the region of interest; and

temporally smoothing the region of interest in each image;

tracking an anatomical structure within the region of interest between each image; and

rendering a three-dimensional visualization using said segmented and tracked region of interest.

2. The method of claim 1 , wherein said transverse scan includes a cinema graphic moving picture.

3. The method of claim 1 , wherein said region of interest is defined by a seed point defined in at least two of said images.

4. The method of claim 3 , wherein said region of interest is further defined by an ellipse surrounding said seed point.

5. The method of claim 3 , further comprising tracking said region of interest by said seed points.

6. The method of claim 1 , further comprising stabilizing said region of interest.

7. The method of claim 1 , wherein said tracking comprises generating two independent tracks.

8. The method of claim 7 , wherein one of said independent tracks is generated from a first of said images to a second of said images and another of said independent tracks is generated from said second of said images to said first of said images.

9. The method of claim 7 , wherein said generating comprises utilizing a block-matching with mutual information process.

10. The method of claim 9 , wherein said block-matching with mutual information process further comprises a constraint based on the angle generated by two of said seed points in an adjacent set of two of said images.

11. The method of claim 1 , wherein the multi-level thresholding algorithm further comprises generating a grid pattern surrounding a center point of the region of interest.

12. The method of claim 1 , wherein the multi-level thresholding algorithm further comprises disambiguating closely-located features within the image by applying a connected-component filtering process.

13. The method of claim 1 , wherein temporally smoothing the region of interest in each image further comprises weighting an outer wall of the lumen based on a size of a manually marked region of interest.

14. The method of claim 13 , wherein weighting further comprises:

filtering each image; and

generating a histogram of a median filtered image.

15. The method of claim 13 , further comprising the steps of:

generating a plaque threshold value;

generating a radial profile of the lumen centroid to the outer wall of the lumen for each image;

filtering each image using a Kalman filtering process; and

generating a plaque accumulation value based on a space between the lumen and the outer wall of the lumen.

16. The method of claim 1 , wherein tracking the anatomical structure within the region of interest between each image further comprises using a cross-correlation process.

17. The method of claim 1 , further comprising the steps of:

automatically detecting at least one of a measurement of the region of interest, the at least one measurement comprising at least one of: a minimum diameter of the region of interest, a maximum diameter of the region of interest, and a cross-sectional area of the diameter of the region of interest; and

calculating a percentage of the region of interest affected by an abnormality based on the at least one detected measurement of the region of interest.

18. The method of claim 1 wherein the lumen further comprises a blood vessel.

19. A method for generating a three dimensional visualization from a plurality of two dimensional images obtained by a transverse scan, the method comprising the steps of:

generating a plurality of transverse scans of a target area with an image generation device;

segmenting a region of interest in each image by:

projecting a plurality of profiles at the region of interest by identifying boundary points of the region of interest, wherein the boundary points form a closed contour about the region of interest;

modeling the region of interest by using a multi-level thresholding algorithm to identify a plurality of pixels of the image which are within the closed contour of the region of interest;

obtaining an ellipse to encapsulate the region of interest; and

temporally smoothing the region of interest in each image;

tracking an anatomical structure within the region of interest between each image; and

rendering a three-dimensional visualization using said segmented and tracked region of interest.