Uncertainty Estimation in Unsupervised MR-CT Synthesis of Scoliotic Spines
Open Access
- 29 March 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Engineering in Medicine and Biology
- Vol. 5, 421-427
Abstract
Goal: Uncertainty estimations through approximate Bayesian inference provide interesting insights to deep neural networks' behavior. In unsupervised learning tasks, where expert labels are unavailable, it becomes ever more important to critique the model through uncertainties. Methods: This paper presents a proof-of-concept for generalizing the aleatoric and epistemic uncertainties in unsupervised MR-CT synthesis of scoliotic spines. A novel adaptation of the cycle-consistency constraint in CycleGAN is proposed such that the model predicts the aleatoric uncertainty maps in addition to the standard volume-to-volume translation between Magnetic Resonance (MR) and Computed Tomography (CT) data. Results: Ablation experiments were performed to understand uncertainty estimation as an implicit regularizer and a measure of the model's confidence. Conclusion: The aleatoric uncertainty helps in distinguishing between the bone and soft-tissue regions in CT and MR data during translation, while the epistemic uncertainty provides interpretable information to the user for downstream tasks.Keywords
Funding Information
- Natural Sciences and Engineering Research Council of Canada
This publication has 17 references indexed in Scilit:
- A review of uncertainty quantification in deep learning: Techniques, applications and challengesInformation Fusion, 2021
- Deep Generative Model for Synthetic-CT Generation with Uncertainty Predictions2020
- Exploring uncertainty measures in deep networks for Multiple sclerosis lesion detection and segmentationMedical Image Analysis, 2020
- Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networksNeurocomputing, 2019
- Cross-Modality Image Synthesis from Unpaired Data Using CycleGAN2018
- Translating and Segmenting Multimodal Medical Volumes with Cycle- and Shape-Consistency Generative Adversarial Network2018
- Methods for interpreting and understanding deep neural networksDigital Signal Processing, 2018
- Deep MR to CT Synthesis Using Unpaired Data2017
- 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation2016
- Texture Analysis for Automatic Segmentation of Intervertebral Disks of Scoliotic Spines From MR ImagesIEEE Transactions on Information Technology in Biomedicine, 2009