Method and apparatus for detecting spiculated masses in mammography

1. A method of detecting one or more spiculated masses in an image using a processor comprising the steps of:

receiving the image in the processor;

filtering the received image using one or more Gaussian filters to detect one or more central mass regions;

filtering the received image using one or more spiculated lesion filters to detect where the one or more spiculated masses converge;

filtering the received image using one or more Difference-of-Gaussian filters to suppress one or more linear structures;

creating an enhanced image showing the detected spiculated masses by combining an output from all of the filtering steps; and

providing the enhanced image to an output of the processor.

2. The method as recited in claim 1 , further comprising the step of applying a weight to each of the outputs from the filtering steps.

3. The method as recited in claim 1 , further comprising the step of determining one or more modeling parameters based on spicule widths, spicule lengths, central mass diameter, major and minor axes of central mass regions, or number of measured spicules.

4. The method as recited in claim 1 , wherein:

the one or more spiculated lesion filters comprise at least a cosine spiculation filter and a sine spiculation filter; and

wherein the cosine speculation filter and the sine speculation filter have the same or different σ, f and r 0 values.

5. The method as recited in claim 1 , wherein the filtering steps are performed in parallel or sequentially.

6. The method as recited in claim 1 , wherein:

the one or more Gaussian filters comprise a set of twenty-five Gaussian filters; and

the one or more Difference-of-Gaussian filters comprise a subset of six-teen Gaussian filters.

7. The method as recited in claim 1 , wherein the step of filtering the received image using a spiculated lesion filter to detect where the one or more spiculated masses converge comprises the steps of:

calculating a Radon Domain transform of the received image;

filtering the calculated Radon Domain transform to detect one or more peaks that represent the one or more spiculated masses;

calculating an inverse Radon Domain Transform of the filtered Radon Domain transform; and

providing an output from the calculated inverse Radon Domain transform that shows the detected spiculated masses.

8. The method as recited in claim 7 , further comprising the step of thresholding the Radon Domain transform.

9. The method as recited in claim 1 , wherein:

the processor comprises a computer, a personal data assistant, a hand-held device, a digital processor, an image processing card, an integrated circuit, scientific device/instrument or computer aided detection device;

the image is received from a memory, a portable storage medium, a personal data assistant, a scientific device/instrument, a hard disk, a floppy disk, an optical drive, a flash memory, a compact disc, a digital video disc, a remote location via a communication link, an image scanner or a mammogram; or

the output of the processor comprises a communications interface, a display, a printer, a memory, a portable storage medium, a personal data assistant, a scientific device/instrument, a hard disk, a floppy disk, an optical drive, a flash memory, a compact disc, a digital video disc, a remote location via a communication link, an image scanner or a mammogram.

10. A computer program embodied on a non-transitory computer readable medium that is executable by a processor for detecting one or more spiculated masses in an image comprising instructions for performing the functions of:

receiving the image in the processor;

filtering the received image using one or more Gaussian filters to detect one or more central mass regions;

filtering the received image using one or more spiculated lesion filters to detect where the one or more spiculated masses converge;

filtering the received image using one or more Difference-of-Gaussian filters to suppress one or more linear structures;

for creating an enhanced image showing the detected spiculated masses by combining an output from all of the filtering steps; and

for providing the enhanced image to an output of the processor.

11. The computer program as recited in claim 10 , further comprising instructions for performing the function of applying a weight to each of the outputs from the filtering steps.

12. The computer program as recited in claim 10 , further comprising instructions for performing the function of determining one or more modeling parameters based on spicule widths, spicule lengths, central mass diameter, major and minor axes of central mass regions, or number of measured spicules.

13. The computer program as recited in claim 10 , wherein the instructions for performing the function of filtering the received image using a spiculated lesion filter to detect where the one or more spiculated masses converge comprises instructions for performing the functions of:

calculating a Radon Domain transform of the received image;

filtering the calculated Radon Domain transform to detect one or more peaks that represent the one or more spiculated masses;

calculating an inverse Radon Domain Transform of the filtered Radon Domain transform; and

providing an output from the calculated inverse Radon Domain transform that shows the detected spiculated masses.

14. An apparatus comprising:

a processor;

a memory communicably coupled to the processor;

one or more communication interfaces communicably coupled to the processor; and

wherein the processor receiving an image from the one or more communication interfaces or the memory, filters the received image using one or more Gaussian filters to detect one or more central mass regions, filters the received image using one or more spiculated lesion filters to detect where the one or more spiculated masses converge, filters the received image using one or more Difference-of-Gaussian filters to suppress one or more linear structures, creates an enhanced image showing the detected spiculated masses by combining an output from all of the filtering steps, and provides the enhanced image to the memory or the one or more communication interfaces.

15. The apparatus as recited in claim 14 , wherein the processor further applies a weight to each of the outputs from the filtering steps.

16. The apparatus as recited in claim 14 , wherein the processor further determines one or more modeling parameters based on spicule widths, spicule lengths, central mass diameter, major and minor axes of central mass regions, or number of measured spicules.

17. The apparatus as recited in claim 14 , wherein the processor filters the received image using a spiculated lesion filter to detect where the one or more spiculated masses converge by calculating a Radon Domain transform of the received image, filtering the calculated Radon Domain transform to detect one or more peaks that represent the one or more spiculated masses, calculating an inverse Radon Domain Transform of the filtered Radon Domain transform, and providing an output from the calculated inverse Radon Domain transform that shows the detected spiculated masses.

18. The apparatus as recited in claim 14 , wherein the processor further thresholds the Radon Domain transform.

19. The apparatus as recited in claim 14 , wherein:

the apparatus comprises a computer, a personal data assistant, a hand-held device, a digital processor, an image processing card, an integrated circuit, scientific device/instrument or computer aided detection device;

the communications interface is communicably coupled to an output device, an input device, a local or wide area network or a scientific device/instrument; or

the output device or the input device comprise a display, a printer, a memory, a portable storage medium, a personal data assistant, a hard disk, a floppy disk, an optical drive, a flash memory, a compact disc, a digital video disc, an image scanner or a mammogram.

20. The apparatus as recited in claim 14 , wherein:

the one or more spiculated lesion filters comprise at least a cosine spiculation filter and a sine spiculation filter; and

wherein the cosine speculation filter and the sine speculation filter have the same or different σ, f and r 0 values.