@inproceedings{d1ad2fd8ea0944a4a5272275cebd18f1,
title = "Finding novelty with uncertainty",
abstract = "Medical images are often used to detect and characterize pathology and disease; however, automatically identifying and segmenting pathology in medical images is challenging because the appearance of pathology across diseases varies widely. To address this challenge, we propose a Bayesian deep learning method that learns to translate healthy computed tomography images to magnetic resonance images and simultaneously calculates voxel-wise uncertainty. Since high uncertainty occurs in pathological regions of the image, this uncertainty can be used for unsupervised anomaly segmentation. We show encouraging experimental results on an unsupervised anomaly segmentation task by combining two types of uncertainty into a novel quantity we call scibilic uncertainty.",
keywords = "Image translation, Uncertainty quantification, Unsupervised anomaly segmentation",
author = "Reinhold, {Jacob C.} and Yufan He and Shizhong Han and Yunqiang Chen and Dashan Gao and Junghoon Lee and Prince, {Jerry L.} and Aaron Carass",
note = "Funding Information: This work was supported by 12 Sigma Technologies. Publisher Copyright: {\textcopyright} 2020 SPIE. All rights reserved.; Medical Imaging 2020: Image Processing ; Conference date: 17-02-2020 Through 20-02-2020",
year = "2020",
doi = "10.1117/12.2549341",
language = "English (US)",
series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",
publisher = "SPIE",
editor = "Ivana Isgum and Landman, {Bennett A.}",
booktitle = "Medical Imaging 2020",
}