Efficient Bayesian Uncertainty Estimation for nnU-Net

by   Yidong Zhao, et al.

The self-configuring nnU-Net has achieved leading performance in a large range of medical image segmentation challenges. It is widely considered as the model of choice and a strong baseline for medical image segmentation. However, despite its extraordinary performance, nnU-Net does not supply a measure of uncertainty to indicate its possible failure. This can be problematic for large-scale image segmentation applications, where data are heterogeneous and nnU-Net may fail without notice. In this work, we introduce a novel method to estimate nnU-Net uncertainty for medical image segmentation. We propose a highly effective scheme for posterior sampling of weight space for Bayesian uncertainty estimation. Different from previous baseline methods such as Monte Carlo Dropout and mean-field Bayesian Neural Networks, our proposed method does not require a variational architecture and keeps the original nnU-Net architecture intact, thereby preserving its excellent performance and ease of use. Additionally, we boost the segmentation performance over the original nnU-Net via marginalizing multi-modal posterior models. We applied our method on the public ACDC and M M datasets of cardiac MRI and demonstrated improved uncertainty estimation over a range of baseline methods. The proposed method further strengthens nnU-Net for medical image segmentation in terms of both segmentation accuracy and quality control.


page 2

page 8


DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation

Semantic image segmentation is the process of labeling each pixel of an ...

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

Automated medical image segmentation inherently involves a certain degre...

SAUNet: Shape Attentive U-Net for Interpretable Medical Image Segmentation

Medical image segmentation is a difficult but important task for many cl...

Image Segmentation Using Hybrid Representations

This work explores a hybrid approach to segmentation as an alternative t...

MCU-Net: A framework towards uncertainty representations for decision support system patient referrals in healthcare contexts

Incorporating a human-in-the-loop system when deploying automated decisi...

nnU-Net: Breaking the Spell on Successful Medical Image Segmentation

Fueled by the diversity of datasets, semantic segmentation is a popular ...

DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation

Image segmentation is a key step for medical image analysis. Approaches ...

Please sign up or login with your details

Forgot password? Click here to reset