Quantifying Sources of Uncertainty in Deep Learning-Based Image Reconstruction

TL;DR

This paper introduces a scalable framework for quantifying both aleatoric and epistemic uncertainties in deep learning-based image reconstruction, enhancing reliability in computed tomography applications.

Contribution

It combines Bayesian deep gradient descent with heteroscedastic noise modeling to estimate uncertainties in learned iterative image reconstruction methods.

Findings

Competitive performance against conventional CT benchmarks.

Uncertainty estimates reflect variability due to limited measurement data.

Method effectively captures uncertainty from missing information.

Abstract

Image reconstruction methods based on deep neural networks have shown outstanding performance, equalling or exceeding the state-of-the-art results of conventional approaches, but often do not provide uncertainty information about the reconstruction. In this work we propose a scalable and efficient framework to simultaneously quantify aleatoric and epistemic uncertainties in learned iterative image reconstruction. We build on a Bayesian deep gradient descent method for quantifying epistemic uncertainty, and incorporate the heteroscedastic variance of the noise to account for the aleatoric uncertainty. We show that our method exhibits competitive performance against conventional benchmarks for computed tomography with both sparse view and limited angle data. The estimated uncertainty captures the variability in the reconstructions, caused by the restricted measurement model, and by missing information, due to the limited angle geometry.