Abstract:Automated medical image segmentation inherently involves a certain degree of uncertainty. One key factor contributing to this uncertainty is the ambiguity that can arise in determining the boundaries of a target region of interest, primarily due to variations in image appearance. On top of this, even among experts in the field, different opinions can emerge regarding the precise definition of specific anatomical structures. This work specifically addresses the modeling of segmentation uncertainty, known as inter-rater uncertainty. Its primary objective is to explore and analyze the variability in segmentation outcomes that can occur when multiple experts in medical imaging interpret and annotate the same images. We introduce a novel Bayesian neural network-based architecture to estimate inter-rater uncertainty in medical image segmentation. Our approach has three key advancements. Firstly, we introduce a one-encoder-multi-decoder architecture specifically tailored for uncertainty estimation, enabling us to capture the rater-specific representation of each expert involved. Secondly, we propose Bayesian modeling for the new architecture, allowing efficient capture of the inter-rater distribution, particularly in scenarios with limited annotations. Lastly, we enhance the rater-specific representation by integrating an attention module into each decoder. This module facilitates focused and refined segmentation results for each rater. We conduct extensive evaluations using synthetic and real-world datasets to validate our technical innovations rigorously. Our method surpasses existing baseline methods in five out of seven diverse tasks on the publicly available \emph{QUBIQ} dataset, considering two evaluation metrics encompassing different uncertainty aspects. Our codes, models, and the new dataset are available through our GitHub repository: <a class="link-external link-https" href="https://github.com/HaoWang420/bOEMD-net" rel="external noopener nofollow">this https URL</a> .

A Decoupled Uncertainty Model for MRI Segmentation Quality Estimation

A Heteroscedastic Uncertainty Model for Decoupling Sources of MRI Image Quality

Estimating MRI Image Quality via Image Reconstruction Uncertainty

Quantifying model uncertainty for semantic segmentation of Fluorine-19 MRI using stochastic gradient MCMC

Uncertainty-based quality assurance of carotid artery wall segmentation in black-blood MRI

Estimating Uncertainty in Neural Networks for Cardiac MRI Segmentation: A Benchmark Study

Structural-Based Uncertainty in Deep Learning Across Anatomical Scales: Analysis in White Matter Lesion Segmentation

Uncertainty Quantification in Machine Learning Based Segmentation: A Post-Hoc Approach for Left Ventricle Volume Estimation in MRI

Give me (un)certainty -- An exploration of parameters that affect segmentation uncertainty

Aleatoric Uncertainty Estimation with Test-Time Augmentation for Medical Image Segmentation with Convolutional Neural Networks.

Test-time Augmentation with Uncertainty Estimation for Deep Learning-Based Medical Image Segmentation

Inter-Rater Uncertainty Quantification in Medical Image Segmentation via Rater-Specific Bayesian Neural Networks

QCResUNet: Joint Subject-level and Voxel-level Segmentation Quality Prediction

Temporal Uncertainty Localization to Enable Human-in-the-loop Analysis of Dynamic Contrast-enhanced Cardiac MRI Datasets

Quantifying U-Net Uncertainty in Multi-Parametric MRI-based Glioma Segmentation by Spherical Image Projection

Uncertainty Quantification in Medical Image Segmentation with Multi-decoder U-Net

Real-time Prediction of Segmentation Quality

Calibrating Ensembles for Scalable Uncertainty Quantification in Deep Learning-based Medical Segmentation

A Novel Quality Control Algorithm for Medical Image Segmentation Based on Fuzzy Uncertainty

Uncertainty Estimation and Propagation in Accelerated MRI Reconstruction

Learning to segment subcortical structures from noisy annotations with a novel uncertainty-reliability aware learning framework