Generative Adversarial Registration for Improved Conditional Deformable Templates

Deformable templates are essential to large-scale medical image registration, segmentation, and population analysis. Current conventional and deep network-based methods for template construction use only regularized registration objectives and often yield templates with blurry and/or anatomically im...

Full description

Saved in:
Bibliographic Details
Published in2021 IEEE/CVF International Conference on Computer Vision (ICCV) pp. 3909 - 3921
Main Authors Dey, Neel, Ren, Mengwei, Dalca, Adrian V., Gerig, Guido
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2021
Subjects
and cell microscopy
biological
Estimation
Magnetic resonance imaging
Medical
Neuroimaging
Pediatrics
Sociology
Stochastic processes
Training
Online AccessGet full text

Cover

More Information
Summary:Deformable templates are essential to large-scale medical image registration, segmentation, and population analysis. Current conventional and deep network-based methods for template construction use only regularized registration objectives and often yield templates with blurry and/or anatomically implausible appearance, confounding downstream biomedical interpretation. We reformulate deformable registration and conditional template estimation as an adversarial game wherein we encourage realism in the moved templates with a generative adversarial registration framework conditioned on flexible image covariates. The resulting templates exhibit significant gain in specificity to attributes such as age and disease, better fit underlying group-wise spatiotemporal trends, and achieve improved sharpness and centrality. These improvements enable more accurate population modeling with diverse covariates for standardized downstream analyses and easier anatomical delineation for structures of interest.
ISSN:2380-7504
DOI:10.1109/ICCV48922.2021.00390