CPC A61B 6/502 (2013.01) [A61B 6/0414 (2013.01); A61B 6/0435 (2013.01); A61B 6/481 (2013.01); A61B 6/5205 (2013.01); G06T 7/0012 (2013.01); G06T 7/11 (2017.01); G06T 7/13 (2017.01); G06T 7/44 (2017.01); G06V 10/421 (2022.01); G06V 10/46 (2022.01); G06V 10/50 (2022.01); A61B 6/12 (2013.01); G06T 2207/10116 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30068 (2013.01); G06V 2201/03 (2022.01)] | 18 Claims |
1. A method of processing a given region of interest (ROI) of an X-ray image of a person's breast to determine presence of a malignancy, the X-ray image having X-ray pixels that indicate intensity of X-rays that passed through the breast to generate the image, the method comprising:
for each given X-ray pixel in the given ROI and each of a selection of J(r) X-ray pixels at respective pixel radii PR(r), 1≤r≤R, from the given x-ray pixel, determining a binary number that provides a measure X-ray intensity indicated by the selected X-ray pixel relative to X-ray intensity indicated by the given X-ray pixel;
using the determined binary numbers for the J(r) selected X-ray pixels at each pixel radius PR(r) to determine a decimal number for the pixel radius PR(r);
histogramming the frequency of occurrence of values of the determined decimal numbers as a function of pixel radius for the given X-ray pixels in the given ROI;
determining a texture feature vector, for the given ROI having components that are equal to the frequencies of occurrence for a selection of M histogrammed values; and
processing the histogrammed frequencies of occurrence for the M values to determine whether the given ROI is malignant.
|