Method for deriving amount of dense tissue from mammographic image
AssigneeAgfa Healthcare NV
InventorTom Bertens
Life of this patent
Estimating the term…
Abstract
A region of interest in the image is determined and for each pixel in each location (x,y) in that region of interest a value hint(x,y) is computed representing the amount of dense tissue below said pixel as h int ( x , y ) = 2 μ fat ( E ) - μ int ( E ) ( Log ( p ( x , y ) ) - Log ( p fat ) ) wherein p(x,y) represents the pixel value of a pixel at location (x,y) in said digital mammographic image, pfat is a reference fat pixel value, μfat,μint are the linear attenuation coefficients of fat tissue and dense tissue and E is the mono-energetic energy value of the x-ray source at image recording, whereby E and pfat are derived from said digital mammographic image representation.
Claims16
A method for deriving the amount of dense tissue from a digital mammographic image representation generated by a digital image acquisition system, the method comprising:
- (1) determining a region of interest, which is comprised of pixels, in said digital mammographic image representation from the digital image acquisition system by computational analysis,
- (2) computing for each pixel in said digital mammographic image representation in each location (x,y) in said region of interest a value hint(x,y) representing the amount of dense tissue below said pixel as
- h
- int
-
- (
- x
- ,
- y
- )
- =
- 2
- μ
- fat
-
- (
- E
- )
- -
- μ
- int
-
- (
- E
- )
-
- (
- Log
-
- (
- p
-
- (
- x
- ,
- y
- )
- )
- -
- Log
-
- (
- p
- fat
- )
- )
- wherein p(x,y) represents a pixel value of a pixel at location (x,y) in said digital mammographic image, pfat is a reference fat pixel value, μfat,μint are the linear attenuation coefficients of fat tissue and dense tissue as a function of E and E is the mono-energetic energy value of an x-ray source of the digital image acquisition system at image recording, in which E and pfat are derived from said digital mammographic image representation, and
- (3) from hint, deducing a breast classification value representative of the amount of dense tissue is deduced from said digital mammographic image representation from the digital image acquisition system by computational analysis.
Dependencies are inferred from the claim text.