Method and apparatus for fast detection of spiculated lesions in digital mammograms
First Claim
1. A method of detecting spiculations in an M×
- N pixel image using a cumulative array of size AM×
BN pixels, where A<
1 and B<
1, comprising the steps of;
determining a region of potential intersection for substantially every image pixel using line information and direction information related to that image pixel, said step of determining a region of potential intersection comprising the steps of;
determining, according to the line information related to the image pixel, whether the image pixel is located along a line, the image pixel having coordinates; and
if the image pixel is located along a line, selecting a region centered on a proportional center pixel corresponding to a predetermined pattern, said predetermined pattern being rotated by an amount related to the direction information related to the image pixel, wherein said proportional center pixel has coordinates equal to said coordinates of the image pixel scaled by A and B;
accumulating said regions of potential intersection to produce said cumulative array; and
processing information contained in said cumulative array for identifying the spiculations in the image.
11 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for the fast detection of spiculated lesions in a digital mammogram, the method for use in a computer aided diagnosis system for assisting a radiologist in identifying and recognizing the spiculations among a multiplicity of lines corresponding to standard fibrous breast tissue. A line and direction image is created from a digital mammogram, and a region of potential intersection for substantially every pixel in the digital mammogram image is determined. The region of potential intersection for each pixel is a predetermined pattern, such as a high aspect ratio rectangle or trapezoid, positioned around the pixel and rotated in a direction corresponding to direction information for that pixel. The regions of potential intersection are accumulated among the pixels to produce a cumulative array, and information in the cumulative array is processed for identifying spiculations in the digital mammogram.
-
Citations
21 Claims
-
1. A method of detecting spiculations in an M×
- N pixel image using a cumulative array of size AM×
BN pixels, where A<
1 and B<
1, comprising the steps of;determining a region of potential intersection for substantially every image pixel using line information and direction information related to that image pixel, said step of determining a region of potential intersection comprising the steps of; determining, according to the line information related to the image pixel, whether the image pixel is located along a line, the image pixel having coordinates; and if the image pixel is located along a line, selecting a region centered on a proportional center pixel corresponding to a predetermined pattern, said predetermined pattern being rotated by an amount related to the direction information related to the image pixel, wherein said proportional center pixel has coordinates equal to said coordinates of the image pixel scaled by A and B; accumulating said regions of potential intersection to produce said cumulative array; and processing information contained in said cumulative array for identifying the spiculations in the image. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12)
- N pixel image using a cumulative array of size AM×
-
13. A method of detecting spiculations in an image, the image having pixels, comprising the steps of:
-
determining a region of potential intersection for each of a plurality of image pixels using line information and direction information related to that image pixel; accumulating said regions of potential intersection to produce a cumulative array; processing information contained in said cumulative array for identifying the spiculations in the image; wherein said step of determining a region of potential intersection is performed for substantially every image pixel and comprises the steps of; determining, according to the line information related to the image pixel, whether the image pixel is located along a line; and if the image pixel is located along a line, selecting a region centered on the image pixel corresponding to a predetermined pattern, said predetermined pattern being rotated by an amount related to the direction information related to the image pixel; wherein said step of determining a region of potential intersection comprises the step of selecting a null region if the image pixel is not located alone a line; wherein said step of accumulating said regions of potential intersection comprises the step of, for each of said image pixels, incrementing all pixels in said cumulative array located within said region of potential intersection for the image pixel by a first amount; computing line information and direction information related to substantially every image pixel; and computing a weighting function based on statistical information related to the direction information for substantially every image pixel and being a function of direction information, said first amount being equal to the weighting function corresponding to the direction information related to the image pixel; wherein the image pixel has coordinates (i,j), said weighting function is a function WT(theta), said direction information related to the pixel is an angle THETA(i,j) and corresponds to a tangent of a line passing through the image pixel, wherein said first amount is equal to WT(THETA(i,j)). - View Dependent Claims (14)
-
-
15. A method of detecting spiculations in an image, said method being capable of locating noneccentric spiculations, the image having pixels, said method comprising the steps of:
-
determining a region of potential intersection for each of a plurality of image pixels using line information and direction information related to that image pixel; computing a plurality of weights corresponding to each of said plurality of image pixels; accumulating, for each of said plurality of image pixels, said plurality of weights into a plurality of accumulation planes for those pixels located within said region of potential intersection for that image pixel; and processing information contained in said plurality of accumulation planes for identifying the noneccentric spiculations in the image; wherein said plurality of image pixels comprises substantially every pixel in said image; wherein said step of determining a region of potential intersection comprises the steps of; determining, according to the line information related to the image pixel, whether the image pixel is located along a line; and if the image pixel is located along a line, selecting a region centered on the image pixel corresponding to a predetermined pattern, said predetermined pattern being rotated by an amount related to the direction information related to the image pixel; wherein said step of determining a region of potential intersection further comprises the step of selecting a null region if the image pixel is not located along a line; wherein said plurality of accumulation planes comprises a first accumulation plane ACC1, a second accumulation plane ACC2, and a third accumulation plane ACC3, wherein said first, second, and third accumulation planes ACC1, ACC2, and ACC3 are capable of being processed for producing a spiculation activity plane ACT and a spiculation eccentricity plane ECC for use in locating noneccentric spiculations; wherein said plurality of accumulation planes, said spiculation activity plane ACT and said spiculation eccentricity plane ECC each comprise pixels with coordinates (i,j), and wherein said step of processing information in said plurality of accumulation planes further comprises the steps of; computing said spiculation activity plane ACT and said spiculation eccentricity plane ECC using information in said plurality of accumulation planes such that said spiculation activity plane ACT comprises pixel values related to the presence of spiculations, and such that said spiculation eccentricity plane ECC comprises pixel values related to the presence of eccentric spiculations; forming a spiculation output plane SO, said spiculation output plane comprising pixels with coordinates (i,j), said spiculation output plane SO being a function of said spiculation activity plane ACT and said spiculation eccentricity plane ECC at each pixel; and using information in said spiculation output plane SO for identifying the noneccentric spiculations in the image. - View Dependent Claims (16, 17, 18, 19, 20, 21)
-
Specification