Granted Patent B2

US 8,315,446 · App. 12/223,550 · Granted Nov 20, 2012

Abstract

Claims (35)

1. A method of processing a mammogram image to derive a value for a parameter useful in detecting differences in breast tissue in subsequent images of the same breast or relative to a control group of such images, said derived parameter being a parameter that changes alongside changes in breast density, the method comprising the steps of processing an image of at least part of a breast by:

computing for pixels of the image a quotient value representative of the aspect ratio of tissue structures depicted in the image;

using a trained classifier to classify said pixels according to their respective said quotient values and assigning a score to the respective pixels representing their classification with respect to at least two classes;

deriving said parameter that changes alongside changes in breast density based on the aggregate pixel membership scores of said classes.

2. A method as claimed in claim 1 , wherein said trained classifier is trained by unsupervised learning.

3. A method as claimed in claim 1 , wherein said trained classifier is trained by supervised learning.

4. A method as claimed in claim 3 , wherein said classifier is trained by supervised learning based on a set of images associated with a higher breast density and a set of images associated with a lower breast density.

5. A method as claimed in claim 1 , wherein said pixels are scored as belonging to one of said classes.

6. A method as claimed in claim 1 , wherein said pixels are scored according to their probability of belonging to at least one of said classes.

7. A method as claimed in claim 1 , wherein said quotient value is determined for each said pixel at each of a plurality of scales.

8. A method as claimed in claim 1 , wherein said quotient values are determined as the normalised difference between eigenvalues of a Hessian matrix based on Gaussian derivatives at a predetermined scale of pixels of the image, which Gaussian derivatives relate the intensity of each pixel to the intensities of the neighbours of said pixel.

9. A method as claimed in claim 1 , comprising

plotting said quotient values on a pre-determined model of a cluster map, said cluster map having at least two pre-plotted points;

clustering the plotted values around said at least two pre-plotted points to result in two discrete clusters; and

deriving said parameter that changes alongside changes in breast density by computing the difference in area between said two clusters.

10. A method as claimed in claim 9 , wherein the pre-determined model is defined in 3-dimensional space in which the three dimensions respectively represent the quotient value when calculated at a respective one of three different scales.

11. A method as claimed in claim 9 , wherein the step of clustering further comprises:

calculating which of said at least two pre-plotted points each plotted value is closest to; and

assigning each value accordingly to said closest pre-plotted point to result in said two discrete clusters of plotted points.

12. A method as claimed in claim 9 , comprising a preliminary step of preparing said pre-determined model of the cluster map incorporating said at least two pre-plotted points.

13. A method as claimed in claim 12 , wherein the step of preparing the model of the cluster map comprises:

a)computing a quotient value for pixels in a test image of a breast based on the results of Hessian matrices compiled for said pixels in the image;

b)plotting each quotient value in 3-dimensional space;

c)randomly selecting at least two points amongst the quotient values as starting points;

d)calculating the distance between each quotient value and each starting point and assigning each quotient value to the starting point to which it is nearest;

e)calculating a mean of each group of points and redefining the starting point as the calculated mean point;

iteratively repeating steps d) and e)until the calculated mean point is equal to the starting point, whereby the final calculated mean points are arranged to be the pre-plotted points of the model of the cluster map.

14. A method as claimed in claim 9 , wherein the number of pre-plotted points determines the number of resulting clusters.

15. A method as claimed in claim 9 , wherein the pre-determined model of the cluster map has four pre-plotted points.

16. A method as claimed in claim 15 , wherein four points are randomly selected as starting points to result in said four pre-plotted points for the model of the cluster map.

17. A method as claimed in claim 14 , wherein the Hessian matrices are derived from Gaussian derivatives of the pixels in the image.

18. A method as claimed in claim 17 , further comprising deriving Gaussian derivates at three different scales of the image to result in three different quotient values for each of said pixels, wherein the three quotient values correspond to the three dimensions of the pre-determined model.

19. A method as claimed in claim 9 , wherein said quotient values define characteristics representative of the shape of objects present in the image.

20. A method as claimed in claim 9 , wherein a quotient value of relatively large magnitude represents a substantially elongate object located in the image.

21. A method as claimed in claim 1 , further comprising deriving a parameter of the same breast at a subsequent period of time and computing the difference of the first and subsequent parameter, wherein the difference is representative of changes in the breast tissue of the breast.

Assignments (22)