System and method for filtering and automatic detection of candidate anatomical structures in medical images
View Patent ↗A system and method for determining the presence or absence of candidate objects in a digital image by computing a gradient field of the digital image and then applying predetermined filters to the gradient field to obtain a response image which can be further processed to detect the objects. The application of these filters to the digital image can be performed by convolution and the filters can be adapted based on the shape of the object to be searched.
1. A method for detecting an anatomical abnormality within a three-dimensional digital image, the method comprising:
computing a first gradient vector field of the digital image in x, y, and z directions;
computing a plurality of second gradient vector fields for each of a plurality of predetermined masks of different sizes;
convoluting the first gradient vector field and the plurality of second gradient vector fields to generate a corresponding set of response images;
integrating the set of response images to generate a single response volume; and
detecting the anatomical abnormality based on the single response volume, wherein each of the plurality of predetermined masks of different sizes is selected according to a shape of the anatomical abnormality being detected.
2. The method of claim 1 wherein the shape of the anatomical abnormality being detected is spherical or semi-spherical.
3. The method of claim 1 , wherein the three-dimensional digital image is a digital medical image.
4. The method of claim 3 , wherein the digital medical image is a CT scan, an MRI or a PET scan.
5. The method of claim 1 , wherein the set of response images are Divergent Gradient Field Response (DGFR) images.
6. The method of claim 5 , wherein the DGFR response images are final response images and are first represented as intermediate response images: DGFRx, DGFRy, and DGFRz, for each corresponding axis, wherein the intermediate response images are combined to form the final response images.
7. The method of claim 6 , wherein the intermediate response images are further processed or combined using either a single function or a combination of functions.
8. The method of claim 1 , wherein detecting the anatomical abnormality based on the single response volume includes analyzing a shape of the single response volume.
9. A computer system comprising:
a processor; and a program storage device readable by the computer system, embodying a program of instructions executable by the processor to perform method steps for detecting an anatomical abnormality within a three-dimensional digital image, the method comprising:
computing a first gradient vector field of the digital image in x, y, and z directions;
computing a plurality of second gradient vector fields for each of a plurality of predetermined masks of different sizes;
convoluting the first gradient vector field and the plurality of second gradient vector fields to generate a corresponding set of response images;
integrating the set of response images to generate a single response volume; and
detecting the anatomical abnormality based on the single response volume, wherein each of the plurality of predetermined masks of different sizes is selected according to a shape of the anatomical abnormality being detected.
10. The computer system of claim 9 , wherein the shape of the anatomical abnormality being detected is spherical or semi-spherical.
11. The computer system of claim 9 , wherein the three-dimensional digital image is a digital medical image.
12. The computer system of claim 11 , wherein the digital medical image is a CT scan, an MRI or a PET scan.
13. The computer system of claim 9 , wherein the set of response images are Divergent Gradient Field Response (DGFR) images.
14. The computer system of claim 13 , wherein the DGFR response images are final response images and are first represented as intermediate response images: DGFRx, DGFRy, and DGFRz, for each corresponding axis, wherein the intermediate response images are combined to form the final response images.
15. The computer system of claim 14 , wherein the intermediate response images are further processed or combined using either a single function or a combination of functions.
16. The method of claim 9 , wherein detecting the anatomical abnormality based on the single response volume includes analyzing a shape of the single response volume.