Uncertainty quantification using Bayesian neural networks in classification: Application to biomedical image segmentation

• Published in: Computational statistics & data analysis Vol. 142; p. 106816
• Main Authors: Kwon, Yongchan, Won, Joong-Ho, Kim, Beom Joon, Paik, Myunghee Cho
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.02.2020
• Subjects: Aleatoric and epistemic uncertainty; algorithms; Bayesian neural network; Bayesian theory; computer vision; data collection; Ischemic stroke lesion segmentation; neural networks; prediction; Retinal blood vessel segmentation; stroke; uncertainty; Uncertainty quantification; Uncertainty quantification; Retinal blood vessel segmentation; Ischemic stroke lesion segmentation; Bayesian neural network; Aleatoric and epistemic uncertainty
• ISSN: 0167-9473; 1872-7352
• DOI: 10.1016/j.csda.2019.106816

Most recent research of deep neural networks in the field of computer vision has focused on improving performances of point predictions by developing network architectures or learning algorithms. Reliable uncertainty quantification accompanied by point estimation can lead to a more informed decision, and the quality of prediction can be improved. In this paper, we invoke a Bayesian neural network and propose a natural way of quantifying uncertainties in classification problems by decomposing the moment-based predictive uncertainty into two parts: aleatoric and epistemic uncertainty. The proposed method takes into account the discrete nature of the outcome, yielding the correct interpretation of each uncertainty. We demonstrate that the proposed uncertainty quantification method provides additional insights into the point prediction using two Ischemic Stroke Lesion Segmentation Challenge datasets and the Digital Retinal Images for Vessel Extraction dataset.