Hierarchical Uncertainty Estimation for Learning-based Registration in Neuroimaging

TL;DR

This paper introduces a hierarchical uncertainty estimation framework for deep learning-based neuroimaging registration, improving uncertainty correlation and registration accuracy by propagating spatial uncertainties through transformation models.

Contribution

It proposes a novel hierarchical uncertainty propagation method tailored for neuroimaging registration, enhancing uncertainty estimation and registration performance.

Findings

Uncertainty estimates correlate better with registration error than Monte Carlo dropout.

Uncertainty-aware fitting improves brain MRI registration accuracy.

Sampling from the posterior enables uncertainty propagation to downstream tasks.

Abstract

Over recent years, deep learning based image registration has achieved impressive accuracy in many domains, including medical imaging and, specifically, human neuroimaging with magnetic resonance imaging (MRI). However, the uncertainty estimation associated with these methods has been largely limited to the application of generic techniques (e.g., Monte Carlo dropout) that do not exploit the peculiarities of the problem domain, particularly spatial modeling. Here, we propose a principled way to propagate uncertainties (epistemic or aleatoric) estimated at the level of spatial location by these methods, to the level of global transformation models, and further to downstream tasks. Specifically, we justify the choice of a Gaussian distribution for the local uncertainty modeling, and then propose a framework where uncertainties spread across hierarchical levels, depending on the choice of transformation model. Experiments on publicly available data sets show that Monte Carlo dropout correlates very poorly with the reference registration error, whereas our uncertainty estimates correlate much better. Crucially, the results also show that uncertainty-aware fitting of transformations improves the registration accuracy of brain MRI scans. Finally, we illustrate how sampling from the posterior distribution of the transformations can be used to propagate uncertainties to downstream neuroimaging tasks. Code is available at: https://github.com/HuXiaoling/Regre4Regis.