SUREMap: Predicting Uncertainty in CNN-based Image Reconstruction Using Stein's Unbiased Risk Estimate

TL;DR

This paper introduces SUREMap, a method that uses Stein's unbiased risk estimate to generate pixel-wise confidence heatmaps for CNN-based image reconstructions, enhancing trust and interpretability in critical applications.

Contribution

It develops a novel approach to quantify uncertainty in CNN reconstructions using SURE, providing per-pixel confidence intervals in a computational imaging context.

Findings

Heatmaps effectively indicate reconstruction confidence levels.

Method improves trustworthiness of CNN reconstructions in medical imaging.

Provides a practical tool for uncertainty quantification in CNN-based imaging.

Abstract

Convolutional neural networks (CNN) have emerged as a powerful tool for solving computational imaging reconstruction problems. However, CNNs are generally difficult-to-understand black-boxes. Accordingly, it is challenging to know when they will work and, more importantly, when they will fail. This limitation is a major barrier to their use in safety-critical applications like medical imaging: Is that blob in the reconstruction an artifact or a tumor? In this work we use Stein's unbiased risk estimate (SURE) to develop per-pixel confidence intervals, in the form of heatmaps, for compressive sensing reconstruction using the approximate message passing (AMP) framework with CNN-based denoisers. These heatmaps tell end-users how much to trust an image formed by a CNN, which could greatly improve the utility of CNNs in various computational imaging applications.