US 11,880,962 B2
System and method for synthesizing magnetic resonance images
Suchandrima Banerjee, Menlo Park, CA (US); Enhao Gong, Palo Alto, CA (US); Greg Zaharchuk, Palo Alto, CA (US); and John M. Pauly, Palo Alto, CA (US)
Assigned to GENERAL ELECTRIC COMPANY, Schenectady, NY (US); and THE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY, Stanford, CA (US)
Appl. No. 16/966,125
Filed by General Electric Company, Schenectady, NY (US); and THE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY, Stanford, CA (US)
PCT Filed Feb. 14, 2019, PCT No. PCT/US2019/017993
§ 371(c)(1), (2) Date Jul. 30, 2020,
PCT Pub. No. WO2019/161043, PCT Pub. Date Aug. 22, 2019.
Claims priority of provisional application 62/631,102, filed on Feb. 15, 2018.
Prior Publication US 2021/0027436 A1, Jan. 28, 2021
Int. Cl. G06T 5/00 (2006.01); A61B 5/055 (2006.01); G06T 5/50 (2006.01); G06T 7/00 (2017.01)
CPC G06T 5/009 (2013.01) [A61B 5/055 (2013.01); G06T 5/50 (2013.01); G06T 7/0012 (2013.01); G06T 2207/10088 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method for synthesizing a magnetic resonance (MR) contrast image, the method comprising:
performing a quantification scan, wherein the quantification scan is the measurement of MR signals reflecting absolute values of physical parameters of tissues being scanned;
using a trained deep neural network to synthesize the MR contrast image from a quantitative acquisition obtained by the quantification scan;
outputting the MR contrast image synthesized by the trained deep neural network; and
wherein the quantification scan is the input to the trained deep neural network.