Surgical Planning Laboratory - Brigham & Women's Hospital - Boston, Massachusetts USA - a teaching affiliate of Harvard Medical School

Surgical Planning Laboratory

The Publication Database hosted by SPL

All Publications | Upload | Advanced Search | Gallery View | Download Statistics | Help | Import | Log in

Registration Uncertainty Quantification via Low-dimensional Characterization of Geometric Deformations

1Computer Science and Engineering, Washington University in St. Louis, MO, United States of America.
2Surgical Planning Laboratory, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA.
3Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, USA.
Elsevier Science
Publication Date:
Magn Reson Imaging
Volume Number:
Magn Reson Imaging. 2019 Dec;64:122-31.
PubMed ID:
Bandlimited space, Bayesian image registration, Laplace approximation, Uncertainty quantification
Appears in Collections:
P41 EB015898/EB/NIBIB NIH HHS/United States
P41 EB015902/EB/NIBIB NIH HHS/United States
Generated Citation:
Wang J., Wells III W.M., Golland P., Zhang M. Registration Uncertainty Quantification via Low-dimensional Characterization of Geometric Deformations. Magn Reson Imaging. 2019 Dec;64:122-31. PMID: 31181245.
Export citation:
Google Scholar: link

This paper presents an efficient approach to quantifying image registration uncertainty based on a low-dimensional representation of geometric deformations. In contrast to previous methods, we develop a Bayesian diffeomorphic registration framework in a bandlimited space, rather than a high-dimensional image space. We show that a dense posterior distribution on deformation fields can be fully characterized by much fewer parameters, which dramatically reduces the computational complexity of model inferences. To further avoid heavy computation loads introduced by random sampling algorithms, we approximate a marginal posterior by using Laplace's method at the optimal solution of log-posterior distribution. Experimental results on both 2D synthetic data and real 3D brain magnetic resonance imaging (MRI) scans demonstrate that our method is significantly faster than the state-of-the-art diffeomorphic registration uncertainty quantification algorithms, while producing comparable results.