US 12,437,401 B2
Systems and methods for determining anatomical deformations
Benjamin Planche, Briarwood, NY (US); Pierre Sibut-Bourde, Gex (FR); Ziyan Wu, Lexington, MA (US); Meng Zheng, Cambridge, MA (US); Zhongpai Gao, Rowley, MA (US); and Abhishek Sharma, Boston, MA (US)
Assigned to Shanghai United Imaging Intelligence Co., Ltd., Shanghai (CN)
Filed by Shanghai United Imaging Intelligence Co., Ltd., Shanghai (CN)
Filed on May 26, 2023, as Appl. No. 18/202,588.
Prior Publication US 2024/0394870 A1, Nov. 28, 2024
Int. Cl. G06T 7/00 (2017.01); G06T 7/10 (2017.01); G06T 7/11 (2017.01)
CPC G06T 7/0012 (2013.01) [G06T 7/10 (2017.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus, comprising:
at least one processor configured to:
obtain a first representation of one or more anatomical structures of a person, wherein the first representation indicates respective geometric characteristics of the one or more anatomical structures under a first set of conditions;
determine, based at least on the first representation of the one or more anatomical structures, a feature descriptor that encodes one or more intrinsic properties of the one or more anatomical structures independent of the first set of conditions;
obtain information regarding a second set of conditions; and
generate a second representation of the one or more anatomical structures that indicates the respective geometric characteristics of the one or more anatomical structures under the second set of conditions, wherein the second representation is generated based at least on the feature descriptor, a first machine-learning (ML) model trained for mapping a three-dimensional (3D) space associated with the first representation of the one or more anatomical structures to a template 3D space associated with the one or more anatomical structures, and a second ML model trained for predicting respective distances of a point in the template 3D space to the one or more anatomical structures.