Uncertainty quantification in non-rigid image registration via stochastic gradient Markov chain Monte Carlo
Daniel Grzech1, Mohammad Farid Azampour1,2,3, Huaqi Qiu1, Ben Glocker1, Bernhard Kainz1,4, Loïc Le Folgoc1
1: Imperial College London, UK, 2: Technische Universität München, Germany, 3: Sharif University of Technology, Tehran, Iran, 4: Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
UNSURE2020 special issue
Publication date: 2021/10/27
Abstract
We develop a new Bayesian model for non-rigid registration of three-dimensional medical images, with a focus on uncertainty quantification. Probabilistic registration of large images with calibrated uncertainty estimates is difficult for both computational and modelling reasons. To address the computational issues, we explore connections between the Markov chain Monte Carlo by backpropagation and the variational inference by backpropagation frameworks, in order to efficiently draw samples from the posterior distribution of transformation parameters. To address the modelling issues, we formulate a Bayesian model for image registration that overcomes the existing barriers when using a dense, high-dimensional, and diffeomorphic transformation parametrisation. This results in improved calibration of uncertainty estimates. We compare the model in terms of both image registration accuracy and uncertainty quantification to VoxelMorph, a state-of-the-art image registration model based on deep learning.
Keywords
deformable image registration · uncertainty quantification · SG-MCMC · SGLD
Bibtex
@article{melba:2021:016:grzech,
title = "Uncertainty quantification in non-rigid image registration via stochastic gradient Markov chain Monte Carlo",
author = "Grzech, Daniel and Azampour, Mohammad Farid and Qiu, Huaqi and Glocker, Ben and Kainz, Bernhard and Le Folgoc, Loïc",
journal = "Machine Learning for Biomedical Imaging",
volume = "1",
issue = "UNSURE2020 special issue",
year = "2021",
pages = "1--25",
issn = "2766-905X",
url = "https://melba-journal.org/2021:016"
}
RIS
TY - JOUR
AU - Grzech, Daniel
AU - Azampour, Mohammad Farid
AU - Qiu, Huaqi
AU - Glocker, Ben
AU - Kainz, Bernhard
AU - Le Folgoc, Loïc
PY - 2021
TI - Uncertainty quantification in non-rigid image registration via stochastic gradient Markov chain Monte Carlo
T2 - Machine Learning for Biomedical Imaging
VL - 1
IS - UNSURE2020 special issue
SP - 1
EP - 25
SN - 2766-905X
UR - https://melba-journal.org/2021:016
ER -
