System and method for automatically segmenting bones in computed tomography angiography data

1. A method for segmenting bone regions from Computed Tomography Angiography (CTA) volume data comprising the steps of:

receiving grayscale CTA volume data, said CTA volume data containing bone regions and contrast agent filled vessels;

automatically detecting and distinguishing between bone regions and contrast agent filled vessels in the CTA volume data;

filtering noise from detected bone regions; and

dilating the filtered detected bone regions to adjacent bones,

wherein bone regions are detected by computing the following formula:

γ=|∇f(x+r,y,z)|+|∇f(x−r,y,z)|−|∇f(x+r,y,z)+∇f(x−r,y,z)|, where f(x,y,z) is an intensity at spatial coordinate (x,y,z) and r is a resolution at which the bone regions are to be detected, and

wherein a value of γ>τ is identified as an initial bone region prior to filtering and dilating, where τ is a targeted threshold value.

2. The method of claim 1 further comprising the step of:

removing disconnected regions from the CTA volume data.

3. The method of claim 1 wherein the detecting step is performed by a locally operated bone detector that utilizes gradient vectors around a voxel for distinguishing bone regions from vessels with similar CTA volume data value intensities.

4. The method of claim 1 further comprising the steps of:

applying the dilated detected bone region to the received grayscale CTA volume data; and

displaying the applied CTA volume data.

5. The method of claim 3 wherein the locally operated bone detector uses symmetric processing techniques on multiple threads and central processing units.

6. The method of claim 1 wherein the filtering step is performed using checkerboard structure elements for matching noise grain patterns to determine a type of filtering action.

7. A system for automatically segmenting bone regions from Computed Tomography Angiography (CTA) volume data comprising:

a locally operated bone detector for distinguishing between bone regions and contrast agent filled vessels;

a filtering operator for removing noise from the detected bone regions;

a dilator for expanding the filtered detected bone regions to adjacent trabecular bones;

a processor for applying the dilated detected bone regions to the CTA volume data; and

a display for displaying the applied CTA volume data,

wherein the bone regions are detected by computing the following formula:

γ=|∇f(x+r,y,z)|+|∇f(x−r,y,z)|−|∇f(x+r,y,z)+∇f(x−r,y,z)|, where f(x,y,z) is an intensity at spatial coordinate (x,y,z) and r is a resolution at which the bone regions are to be detected, and

wherein a value of γ>τ is identified as an initial bone region prior to filtering and dilating, where τ is a targeted threshold value.

8. The system of claim 7 wherein the locally operated bone detector utilizes adjacent gradient vectors around a voxel for distinguishing bone regions from vessels with similar intensities.

9. The system of claim 7 wherein the locally operated bone detector utilizes symmetric processing techniques on multiple threads and central processing units.

10. The system of claim 7 wherein the filtering operator utilizes checkerboard structure elements for matching noise grain patterns to determine a type of filtering action.

11. The method of claim 1 , wherein the filtering step further comprises simultaneously removing noise grains from a background and a foreground of a detected bone region.

12. The system of claim 7 , wherein the filtering operator simultaneously removes noise grains from a background and a foreground of a detected bone region.