US 12,254,677 B2
System and methods for augmenting X-ray images for training of deep neural networks
Grzegorz Andrzej Toporek, Cambridge, MA (US); Ashish Sattyavrat Panse, Burlington, MA (US); Sean Kyne, Brookline, MA (US); Molly Lara Flexman, Melrose, MA (US); and Jochen Kruecker, Andover, MA (US)
Assigned to KONINKLIJKE PHILIPS N.V., Eindhoven (NL)
Filed by KONINKLIJKE PHILIPS N.V., Eindhoven (NL)
Filed on Dec. 22, 2021, as Appl. No. 17/559,472.
Claims priority of provisional application 63/129,670, filed on Dec. 23, 2020.
Claims priority of application No. 21154325 (EP), filed on Jan. 29, 2021.
Prior Publication US 2022/0198784 A1, Jun. 23, 2022
Int. Cl. G06V 10/77 (2022.01); G06T 7/00 (2017.01); G06V 10/774 (2022.01)
CPC G06V 10/7747 (2022.01) [G06T 7/0012 (2013.01); G06T 2207/10116 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for providing medical training data for a machine learning model, the system comprising:
a processor in communication with memory, the processor configured to:
obtain the medical training data including medical training imagery of an anatomy of one or more patients,
compute a projection footprint of a model representative of a medical object associated with a medical procedure,
access a set of rules corresponding to the model, wherein the set of rules defines determining a location to position the model within the anatomy with respect to one or more structures that represents a realistic positioning of the medical object with respect to the one or more structures for the medical procedure, and
modify the medical training imagery by positioning the projection footprint in the medical training imagery based on the set of rules.