EDUE: Expert Disagreement-Guided One-Pass Uncertainty Estimation for Medical Image Segmentation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Abutalip, Kudaibergen, Saeed, Numan, Sobirov, Ikboljon, Andrearczyk, Vincent, Depeursinge, Adrien, Yaqub, Mohammad
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916175221358592
author Abutalip, Kudaibergen
Saeed, Numan
Sobirov, Ikboljon
Andrearczyk, Vincent
Depeursinge, Adrien
Yaqub, Mohammad
author_facet Abutalip, Kudaibergen
Saeed, Numan
Sobirov, Ikboljon
Andrearczyk, Vincent
Depeursinge, Adrien
Yaqub, Mohammad
contents Deploying deep learning (DL) models in medical applications relies on predictive performance and other critical factors, such as conveying trustworthy predictive uncertainty. Uncertainty estimation (UE) methods provide potential solutions for evaluating prediction reliability and improving the model confidence calibration. Despite increasing interest in UE, challenges persist, such as the need for explicit methods to capture aleatoric uncertainty and align uncertainty estimates with real-life disagreements among domain experts. This paper proposes an Expert Disagreement-Guided Uncertainty Estimation (EDUE) for medical image segmentation. By leveraging variability in ground-truth annotations from multiple raters, we guide the model during training and incorporate random sampling-based strategies to enhance calibration confidence. Our method achieves 55% and 23% improvement in correlation on average with expert disagreements at the image and pixel levels, respectively, better calibration, and competitive segmentation performance compared to the state-of-the-art deep ensembles, requiring only a single forward pass.
format Preprint
id arxiv_https___arxiv_org_abs_2403_16594
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle EDUE: Expert Disagreement-Guided One-Pass Uncertainty Estimation for Medical Image Segmentation
Abutalip, Kudaibergen
Saeed, Numan
Sobirov, Ikboljon
Andrearczyk, Vincent
Depeursinge, Adrien
Yaqub, Mohammad
Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
Deploying deep learning (DL) models in medical applications relies on predictive performance and other critical factors, such as conveying trustworthy predictive uncertainty. Uncertainty estimation (UE) methods provide potential solutions for evaluating prediction reliability and improving the model confidence calibration. Despite increasing interest in UE, challenges persist, such as the need for explicit methods to capture aleatoric uncertainty and align uncertainty estimates with real-life disagreements among domain experts. This paper proposes an Expert Disagreement-Guided Uncertainty Estimation (EDUE) for medical image segmentation. By leveraging variability in ground-truth annotations from multiple raters, we guide the model during training and incorporate random sampling-based strategies to enhance calibration confidence. Our method achieves 55% and 23% improvement in correlation on average with expert disagreements at the image and pixel levels, respectively, better calibration, and competitive segmentation performance compared to the state-of-the-art deep ensembles, requiring only a single forward pass.
title EDUE: Expert Disagreement-Guided One-Pass Uncertainty Estimation for Medical Image Segmentation
topic Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2403.16594