US 12,475,552 B2
Methods and apparatuses for generating anatomical models using diagnostic images
Mert Karaoglu, Munich (DE); and Alexander Ladikos, Munich (DE)
Assigned to Boston Scientific Scimed, Inc., Maple Grove, MN (US)
Filed by Boston Scientific Scimed Inc., Maple Grove, MN (US)
Filed on Jan. 14, 2022, as Appl. No. 17/576,271.
Claims priority of provisional application 63/138,186, filed on Jan. 15, 2021.
Prior Publication US 2022/0230303 A1, Jul. 21, 2022
Int. Cl. G06T 7/00 (2017.01); A61B 34/10 (2016.01); A61B 34/20 (2016.01); G06V 10/25 (2022.01); G06V 10/44 (2022.01); G06V 10/774 (2022.01)
CPC G06T 7/0012 (2013.01) [A61B 34/10 (2016.02); A61B 34/20 (2016.02); G06V 10/25 (2022.01); G06V 10/454 (2022.01); G06V 10/774 (2022.01); A61B 2034/105 (2016.02); A61B 2034/107 (2016.02); A61B 2034/2065 (2016.02); G06T 2207/10068 (2013.01); G06T 2207/30061 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus, comprising:
at least one processor;
a memory coupled to the at least one processor, the memory comprising instructions that, when executed by the at least one processor, cause the at least one processor to:
access a plurality of endoscopic training images comprising a plurality of synthetic images and a plurality of real images, access a plurality of depth ground truths associated with the plurality of synthetic images, perform supervised training of at least one computational model using the plurality of synthetic images and the plurality of depth ground truths to generate a synthetic encoder and synthetic decoder in a first training process, and
perform domain adversarial training on the synthetic encoder using the plurality of real images to generate a real image encoder for the at least one computational model in a second training phase subsequent to the first training phase, wherein the real image encoder is generated as a separate encoder distinct from the synthetic encoder.