Method and system for detection of bone fractures

1. A method for detection of bone fractures using image processing of a digitized x-ray image, the method comprising:

extracting a contour of a bone in the digitized x-ray image;

identifying a bounding box around an area of interest based on the extracted contour of the bone; and

applying an adaptive sampling scheme in which the bounding box is divided into a predetermined number of sampling points of a normalized sampling grid such that sampling locations in different images correspond to consistent locations within the area of interest.

2. The method as claimed in claim 1 , wherein the extracting of the contour of the bone in the digitized x-ray image comprises applying a Canny edge detector to the digitized x-ray image.

3. The method as claimed in claims 1 , wherein the extracting of the contour of the bone in the digitized x-ray image comprises applying a snake algorithm to the digitized x-ray image.

4. The method as claimed in claim 3 , wherein applying the snake algorithm to the digitized x-ray image comprises creating a Gradient Vector Flow (GVF).

5. The method as claimed in claim 1 , wherein a sampling region around the sampling points is chosen to cover image pixel points between the sampling points.

6. The method as claimed in claim 1 , wherein the method further comprises calculating one or more texture maps of the digitized x-ray image and detecting a bone fracture based on respective reference texture maps.

7. The method as claimed in claim 6 , wherein the texture maps comprise a Gabor texture orientation map.

8. The method as claimed in claim 6 , wherein the texture maps comprise an Intensity gradient direction map.

9. The method as claimed in claim 6 , wherein the texture maps comprise a Markov Random Field texture map.

10. The method as claimed in claim 6 , wherein the method further comprises calculating one or more difference maps between the respective texture maps calculated for the digitized x-ray image and the respective reference texture maps.

11. The method as claimed in claim 10 , wherein the difference maps are classified using one or more classifiers.

12. The method as claimed in claim 11 , wherein the difference maps are classified using Bayesian classifiers.

13. The method as claimed in claim 11 , wherein the difference maps are classified using Support Vector Machine classifiers.

14. The method as claimed in claim 1 , wherein the processing method further comprises:

determining a femoral shaft axis in the digitized x-ray image;

determining a femoral neck axis in the digitized x-ray image;

measuring an obtuse angle between the femoral neck axis and the femoral shaft axis; and

detecting the bone fracture based on the measured obtuse angle.

15. The method as claimed in claim 14 , further comprising calculating level lines from respective points on the contour of the bone in the digitized x-ray image and extending normally to the contour to respective other points on the extracted contour.

16. The method as claimed in claim 15 , wherein determining the femoral shaft axis is based on midpoints of the level lines in a shaft portion of the contour of the bone.

17. The method as claimed in claim 15 , wherein determining the femoral neck axis is based on the level lines in femoral head and neck portion of the contour of the bone.

18. A system for detection of bone fractures comprising:

means for receiving a digitized x-ray image; and

means for processing the digitized x-ray image for detection of bone fractures;

wherein the means for processing the digitized x-ray image comprises

means for extracting a contour of a bone in the digitized x-ray image,

means for identifying a bounding box around an area of interest based on the extracted contour of the bone, and

means for applying an adaptive sampling scheme in which the bounding box is divided into a predetermined number of sampling points of a normalized sampling grid such that sampling locations in different images correspond to consistent locations within the area of interest.

19. A system for detection of bone fractures comprising:

a database for receiving and storing a digitized x-ray image; and

a processor for processing the digitized x-ray image for detection of bone fractures, wherein the processor is configured to:

extract a contour of a bone in the digitized x-ray image;

identify a bounding box around an area of interest based on the extracted contour of the bone; and

apply an adaptive sampling scheme in which the bounding box is divided into a predetermined number of sampling points of a normalized sampling grid such that sampling locations in different images correspond to consistent locations within the area of interest.