TY - GEN
T1 - Image Registration via Stochastic Gradient Markov Chain Monte Carlo
AU - Grzech, Daniel
AU - Kainz, Bernhard
AU - Glocker, Ben
AU - le Folgoc, Loïc
N1 - Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020/1/1
Y1 - 2020/1/1
N2 - We develop a fully Bayesian framework for non-rigid registration of three-dimensional medical images, with a focus on uncertainty quantification. Probabilistic registration of large images along 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 backprop and the variational inference by backprop frameworks in order to efficiently draw thousands of samples from the posterior distribution. Regarding the modelling issues, we carefully design a Bayesian model for registration to overcome the existing barriers when using a dense, high-dimensional, and diffeomorphic parameterisation of the transformation. This results in improved calibration of uncertainty estimates.
AB - We develop a fully Bayesian framework for non-rigid registration of three-dimensional medical images, with a focus on uncertainty quantification. Probabilistic registration of large images along 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 backprop and the variational inference by backprop frameworks in order to efficiently draw thousands of samples from the posterior distribution. Regarding the modelling issues, we carefully design a Bayesian model for registration to overcome the existing barriers when using a dense, high-dimensional, and diffeomorphic parameterisation of the transformation. This results in improved calibration of uncertainty estimates.
UR - https://www.scopus.com/pages/publications/85093074342
U2 - 10.1007/978-3-030-60365-6_1
DO - 10.1007/978-3-030-60365-6_1
M3 - Conference contribution
AN - SCOPUS:85093074342
SN - 9783030603649
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 3
EP - 12
BT - Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, and Graphs in Biomedical Image Analysis - 2nd International Workshop, UNSURE 2020, and 3rd International Workshop, GRAIL 2020, Held in Conjunction with MICCAI 2020, Proceedings
A2 - Sudre, Carole H.
A2 - Fehri, Hamid
A2 - Arbel, Tal
A2 - Baumgartner, Christian F.
A2 - Dalca, Adrian
A2 - Tanno, Ryutaro
A2 - Van Leemput, Koen
A2 - Wells, William M.
A2 - Sotiras, Aristeidis
A2 - Papiez, Bartlomiej
A2 - Ferrante, Enzo
A2 - Parisot, Sarah
PB - Springer Science and Business Media Deutschland GmbH
T2 - 2nd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, UNSURE 2020, and the 3rd International Workshop on Graphs in Biomedical Image Analysis, GRAIL 2020, held in conjunction with the 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2020
Y2 - 8 October 2020 through 8 October 2020
ER -