Patent · US Active

Method and system for convolutional neural network regression based 2D/3D image registration

US10235606B2 · kind B2 · utility

38Cited by

0References

21Claims

0Family size

Assignees

Siemens Healthcare Limited · DE
University of British Columbia · CA

Inventors

Shun Miao · Plainsboro, US
Rui Liao · Princeton Junction, US
Zhen Wang · Shanghai, CN

Key dates

Filing date	Jul 11, 2016
Grant date	Mar 19, 2019
Priority date	—
Expiry date	Jun 7, 2037

Classification

Technology area (CPC G)Physics
CPC primaryG06T2207/30052
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

A method and apparatus for convolutional neural network (CNN) regression based 2D/3D registration of medical images is disclosed. A parameter space zone is determined based on transformation parameters corresponding to a digitally reconstructed radiograph (DRR) generated from the 3D medical image. Local image residual (LIR) features are calculated from local patches of the DRR and the X-ray image based on a set of 3D points in the 3D medical image extracted for the determined parameter space zone. Updated transformation parameters are calculated based on the LIR features using a hierarchical series of regressors trained for the determined parameter space zone. The hierarchical series of regressors includes a plurality of regressors each of which calculates updates for a respective subset of the transformation parameters.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.