Method of microcalcification detection in mammography
View Patent ↗A method of microcalcification detection in a digital mammographic image identifies one or more potential microcalcification sites in the mammographic image according to spot clustering. Each of the one or more potential microcalcification sites is assigned either as a member of a positive candidate set or as a member of a rejected candidate set. Optionally at least one subsequent classifier process that selectively assigns zero or more members of the positive candidate set to the rejected candidate set is executed, according to results from the at least one subsequent classifier process. One or more members of the rejected candidate set are selected as a reclamation candidate set according to results from the initial and any subsequent classifier process. One or more members of the reclamation candidate set are assigned either back to the rejected candidate set or to the positive candidate set according to results from a reclamation classifier process.
1. A method of microcalcification detection in a digital mammographic image, comprising:
identifying one or more potential microcalcification sites in the mammographic image according to spot clustering criteria;
assigning each of the one or more potential microcalcification sites either as a member of a positive candidate set or as a member of a rejected candidate set according to results from an initial classifier process;
optionally executing at least one subsequent classifier process that selectively assigns zero or more members of the positive candidate set to the rejected candidate set according to results from the at least one subsequent classifier process;
selecting one or more members of the rejected candidate set as a reclamation candidate set according to results from the initial and any subsequent classifier process; and
assigning one or more members of the reclamation candidate set either back to the rejected candidate set or to the positive candidate set according to results from a reclamation classifier process.
2. The method of claim 1 wherein the initial classifier process employs one or more pattern classification algorithms as classifiers.
3. The method of claim 1 wherein identifying one or more potential microcalcification sites in the mammographic image according to spot clustering criteria comprises:
applying one or more image enhancement algorithms to the mammographic image;
applying an image segmentation algorithm to the enhanced mammographic image, based on the pixel intensity values; and
grouping segmented image pixels into spots.
4. The method of claim 1 wherein identifying one or more potential microcalcification sites in the mammographic image according to spot clustering criteria comprises:
forming spots from connected pixels having brightness of at least a predetermined threshold; and
identifying each cluster having at least a predetermined number of spots wherein any one spot is within a predefined distance from at least one of the other spots within the cluster.
5. The method of claim 1 wherein the reclamation classifier process employs one or more pattern classification algorithms as classifiers.
6. The method of claim 1 further comprising displaying microcalcification sites corresponding to the positive candidate set on a display screen with markers of a first color.
7. The method of claim 6 further comprising displaying microcalcification sites that were added to the positive candidate set by the reclamation classifier process with markers of a second color.
8. The method of claim 1 further comprising displaying microcalcification sites corresponding to the positive candidate set on a display screen with markers of a first shape.
9. The method of claim 8 further comprising displaying microcalcification sites that were added to the positive candidate set by the reclamation classifier process with markers of a second shape.
10. The method of claim 1 wherein at least one of the initial classifier, subsequent classifier, or reclamation classifier processes uses a trained classifier algorithm.
11. The method of claim 1 wherein assigning one or more members of the reclamation candidate set comprises:
selecting a subset of pixels associated with the reclamation candidate set according to pixel intensity;
computing gradient magnitude data for the selected subset of pixels;
applying one or more ring filters to the gradient magnitude data for each pixel in the selected subset and storing ring filter size and pixel location and filter response data in a spot candidate list; and
sorting the spot candidate list according to filter response data.
12. A method of microcalcification detection in a digital mammographic image, comprising:
a) identifying one or more potential microcalcification sites in the mammographic image according to spot clustering criteria;
b) obtaining parameters for one or more classifiers from a human operator;
c) executing the one or more classifiers, with the obtained parameters, to assign each of the one or more potential microcalcification sites to either a positive candidate set or a rejected candidate set, wherein the positive candidate set and rejected candidate sets are mutually exclusive; and
d) responding to an operator instruction to execute at least one additional classifier on a portion of the rejected candidate set and re-assigning at least one member of the portion of the rejected candidate set to the positive candidate set.
13. The method of claim 12 further comprising displaying calcification sites assigned to the positive candidate set in step c) with markers of a first color, size, or shape and displaying re-assigned calcification sites from step d) with markers of a second color, size, or shape.
14. The method of claim 12 wherein the portion of the rejected candidate set in step d) is selected according to ranking from a previous classifier execution in step c).