US 12,229,917 B2
System and methods for filtering medical images
Erik Normann Steen, Horten (NO)
Assigned to GE PRECISION HEALTHCARE LLC, Wauwatosa, WI (US)
Filed by GE Precision Healthcare LLC, Wauwatosa, WI (US)
Filed on May 20, 2022, as Appl. No. 17/664,389.
Prior Publication US 2023/0377104 A1, Nov. 23, 2023
Int. Cl. G06T 5/20 (2006.01); G06T 5/50 (2006.01); G06T 5/70 (2024.01); G06T 5/73 (2024.01)
CPC G06T 5/20 (2013.01) [G06T 5/50 (2013.01); G06T 5/70 (2024.01); G06T 5/73 (2024.01); G06T 2207/10016 (2013.01); G06T 2207/20021 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/20104 (2013.01); G06T 2207/20192 (2013.01); G06T 2207/20221 (2013.01); G06T 2207/30004 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method comprising:
receiving a medical image;
applying a filter to the medical image to produce a filtered medical image, wherein a subregion of the filtered medical image includes a filtering artifact;
displaying the filtered medical image via a display device;
receiving a selection of the subregion of the filtered medical image from a user input device;
blending, in the subregion, pixel intensity values from the filtered medical image with pixel intensity values from the medical image, to produce a blended image, wherein a visibility of the filtering artifact is attenuated in the blended image;
displaying the blended image via the display device;
storing the medical image and the blended image as a training data pair;
training a machine learning model using the training data pair;
dividing the medical image into a first plurality of patches;
dividing the blended image into a second plurality of patches, wherein a number of the first plurality of patches is equal to a number of the second plurality of patches;
pairing each patch of the first plurality of patches with a spatially corresponding patch from the second plurality of patches, to form a plurality of training data pairs; and
storing the plurality of training data pairs in non-transitory memory.