Explainable vertebral fracture analysis with uncertainty estimation using differentiable rule-based classification

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Skärström, Victor Wåhlstrand, Johansson, Lisa, Alvén, Jennifer, Lorentzon, Mattias, Häggström, Ida
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917712021684224
author Skärström, Victor Wåhlstrand
Johansson, Lisa
Alvén, Jennifer
Lorentzon, Mattias
Häggström, Ida
author_facet Skärström, Victor Wåhlstrand
Johansson, Lisa
Alvén, Jennifer
Lorentzon, Mattias
Häggström, Ida
contents We present a novel method for explainable vertebral fracture assessment (XVFA) in low-dose radiographs using deep neural networks, incorporating vertebra detection and keypoint localization with uncertainty estimates. We incorporate Genant's semi-quantitative criteria as a differentiable rule-based means of classifying both vertebra fracture grade and morphology. Unlike previous work, XVFA provides explainable classifications relatable to current clinical methodology, as well as uncertainty estimations, while at the same time surpassing state-of-the art methods with a vertebra-level sensitivity of 93% and end-to-end AUC of 97% in a challenging setting. Moreover, we compare intra-reader agreement with model uncertainty estimates, with model reliability on par with human annotators.
format Preprint
id arxiv_https___arxiv_org_abs_2407_02926
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Explainable vertebral fracture analysis with uncertainty estimation using differentiable rule-based classification
Skärström, Victor Wåhlstrand
Johansson, Lisa
Alvén, Jennifer
Lorentzon, Mattias
Häggström, Ida
Image and Video Processing
Computer Vision and Pattern Recognition
I.4.8; I.2.10; J.3
We present a novel method for explainable vertebral fracture assessment (XVFA) in low-dose radiographs using deep neural networks, incorporating vertebra detection and keypoint localization with uncertainty estimates. We incorporate Genant's semi-quantitative criteria as a differentiable rule-based means of classifying both vertebra fracture grade and morphology. Unlike previous work, XVFA provides explainable classifications relatable to current clinical methodology, as well as uncertainty estimations, while at the same time surpassing state-of-the art methods with a vertebra-level sensitivity of 93% and end-to-end AUC of 97% in a challenging setting. Moreover, we compare intra-reader agreement with model uncertainty estimates, with model reliability on par with human annotators.
title Explainable vertebral fracture analysis with uncertainty estimation using differentiable rule-based classification
topic Image and Video Processing
Computer Vision and Pattern Recognition
I.4.8; I.2.10; J.3
url https://arxiv.org/abs/2407.02926