Reducing Risk and Uncertainty of Deep Neural Networks on Diagnosing COVID-19 Infection

TL;DR

This paper introduces uncertainty estimation techniques to improve the reliability of deep neural networks in COVID-19 diagnosis, enabling better detection of confusing cases for expert review and clinical validation.

Contribution

It is the first to apply and evaluate uncertainty estimation methods specifically for COVID-19 detection with deep neural networks.

Findings

Uncertainty estimation improves detection reliability.

Best method validated with medical professionals.

Provides a framework for clinical application.

Abstract

Effective and reliable screening of patients via Computer-Aided Diagnosis can play a crucial part in the battle against COVID-19. Most of the existing works focus on developing sophisticated methods yielding high detection performance, yet not addressing the issue of predictive uncertainty. In this work, we introduce uncertainty estimation to detect confusing cases for expert referral to address the unreliability of state-of-the-art (SOTA) DNNs on COVID-19 detection. To the best of our knowledge, we are the first to address this issue on the COVID-19 detection problem. In this work, we investigate a number of SOTA uncertainty estimation methods on publicly available COVID dataset and present our experimental findings. In collaboration with medical professionals, we further validate the results to ensure the viability of the best performing method in clinical practice.