System and method for detecting intervertebral disc alignment using vertebrae segmentation

1. A computer-implemented method for intervertebral disc alignment comprising:

providing a spine image and a click point on the spine image;

applying an adaptive thresholding technique to segment vertebrae regions at the click point from the spine image;

filtering segmented vertebrae regions with a morphological operation;

estimating a multi-scale orientation field from filtered segmented vertebrae regions;

extracting an intervertebral disc region by applying region growing for each scale;

integrating the multi-scale orientation field in a locally segmented intervertebral disc region based on the intervertebral disc region;

estimating a disc centerline traversing a spinal axis from the filtered segmented vertebrae regions; and

fusing an integrated multi-scale orientation field with the disc centerline based on associated confidence measures to provide an intervertebral disc alignment.

2. The compute implemented method of claim 1 , wherein the confidence measures of the orientation field are determined by a gradient magnitude.

3. The computer-implemented method of claim 1 , wherein fusing further comprises an iterative re-weighted least square estimation applied to provide the intervertebral disc alignment.

4. The computer-implemented method of claim 3 , wherein the iterative re-weighted least square estimation is applied for an estimated multi-scale orientation field at different scales, the multi-scale orientation field estimates are integrated based on weighted averaging with a weight for each scale.

5. The computer-implemented method of claim 1 , wherein the disc center is determined for each scale to be an average of all points in a segmented vertebrae region and a weighted average of these disc centers for each scale with the same weights is determined for the multi-scale orientation field.

6. The computer-implemented method of claim 1 , wherein a estimating the disc centerline further comprises determining the disc centerline by minimizing distances from points on the disc centerline to respective nearest points on adjacent ones of the filtered segmented vertebrae regions along a normal direction to the disc orientation.

7. A non-transitory computer-readable medium embodying instructions executed by a processor to perform method steps for intervertebral disc alignment, the method steps comprising:

providing a spine image and a click point on the spine image;

applying an adaptive thresholding technique to segment vertebrae regions at the click point from the spine image;

filtering segmented vertebrae regions with a morphological operation;

estimating a multi-scale orientation field from filtered segmented vertebrae regions;

extracting an intervertebral disc region by applying region growing for each scale;

integrating the multi-scale orientation field in a locally segmented intervertebral disc region based on the intervertebral disc region;

estimating a disc centerline traversing a spinal axis from the filtered segmented vertebrae regions; and

fusing an integrated multi-scale orientation field with the disc centerline based on associated confidence measures to provide an intervertebral disc alignment.

8. The method of claim 7 , wherein the confidence measures of the orientation field are determined by a gradient magnitude.

9. The method of claim 7 , wherein fusing further comprises an iterative re-weighted least square estimation applied to provide the intervertebral disc alignment.

10. The method of claim 9 , wherein the iterative re-weighted least square estimation is applied for an estimated multi-scale orientation field at different scales, the multi-scale orientation field estimates are integrated based on weighted averaging with a weight for each scale.

11. The method of claim 7 , wherein the disc center is determined for each scale to be an average of all points in a segmented vertebrae region and a weighted average of these disc centers for each scale with the same weights is determined for the multi-scale orientation field.

12. The method of claim 7 , wherein estimating the disc centerline further comprises determining the disc centerline by minimizing distances from points on the disc centerline to respective nearest points on adjacent ones of the filtered segmented vertebrae regions along a normal direction to the disc orientation.

13. A non-transitory computer-readable medium embodying instructions executed by a processor to perform method steps for intervertebral disc alignment, the method steps comprising:

providing an integrated multi-scale orientation field in a segmented intervertebral disc region between adjacent vertebrae;

providing an estimated disc centerline traversing a spinal axis between the adjacent vertebrae; and

fusing an integrated multi-scale orientation field with the disc centerline based on associated confidence measures to provide an intervertebral disc alignment.

14. The method of claim 13 , wherein the confidence measures of the orientation field are determined by a gradient magnitude.

15. The method of claim 13 , wherein fusing further comprises an iterative re-weighted least square estimation applied to provide the intervertebral disc alignment.

16. The method of claim 15 , wherein the iterative re-weighted least square estimation is applied for an estimated multi-scale orientation field at different scales, the multi-scale orientation field estimates are integrated based on weighted averaging with a weight for each scale.