US 12,277,627 B1
Correction of ultrasound probe-induced metal artifacts in X-ray based imaging systems
Sathyathas Puvanasunthararajah, Logan Central (AU); Saskia Camps, Prangins (CH); Davide Fontanarosa, Bulimba (AU); Adriano Garonna, Geneva (CH); and Marie-Luise Wille, Jindalee (AU)
Assigned to Queensland University of Technology, Brisbane (AU)
Filed by Queensland University of Technology, Brisbane (AU)
Filed on Aug. 29, 2022, as Appl. No. 17/897,771.
Claims priority of provisional application 63/237,856, filed on Aug. 27, 2021.
Int. Cl. G06T 11/00 (2006.01); A61B 6/00 (2024.01); A61B 6/03 (2006.01); A61B 34/20 (2016.01); A61N 5/10 (2006.01)
CPC G06T 11/008 (2013.01) [A61B 6/032 (2013.01); A61B 6/4417 (2013.01); A61B 6/5258 (2013.01); A61B 34/20 (2016.02); A61N 5/103 (2013.01); G06T 2211/421 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for correcting metal artifacts in an X-ray image of a subject, said system comprising:
a processor configured to:
obtain an original digital object corresponding to an original X-ray image with an X-ray based imaging system;
identify, within said original digital object, metal data associated with a device externally mounted to said subject to define a metal-only digital object;
generate a first tissue classified digital object from said original digital object;
incorporate a spatial relationship among adjacent pixels of said original digital object;
generate a metal-only sinogram from said metal-only digital object;
generate a first tissue classified sinogram from said first tissue classified digital object;
generate an original sinogram from said original digital object;
dilate and smooth said metal-only sinogram;
combine said metal-only sinogram, said first tissue classified sinogram, and said original sinogram to create an initial metal artifact reduction (MAR) digital object;
calculate differences of represented weighted linear attenuation coefficients between corresponding pixels of said original digital object and said initial MAR digital object;
identify pixels having said differences that are within a predetermined range;
designate said pixels identified in the step of identifying pixels as soft tissue on said initial MAR digital object;
generate a second tissue classified digital object from said MAR digital object;
generate a combined tissue classified digital object from said first tissue classified digital object and said second tissue classified digital object;
calculate mean absolute differences between corresponding pixels of said first tissue classified digital object and said second tissue classified digital object;
add said mean absolute differences with said first tissue classified digital object; and
generate a combined sinogram from said combined tissue classified digital object.