Development and evaluation of CADe systems in low-prevalence setting: The RARE25 challenge for early detection of Barrett's neoplasia

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jaspers, Tim J. M., Caetano, Francisco, Claessens, Cris H. B., Kusters, Carolus H. J., van Heslinga, Rixta A. H. van Eijck, Slooter, Floor, Bergman, Jacques J., De With, Peter H. N., Jong, Martijn R., de Groof, Albert J., van der Sommen, Fons
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914468963811328
author Jaspers, Tim J. M.
Caetano, Francisco
Claessens, Cris H. B.
Kusters, Carolus H. J.
van Heslinga, Rixta A. H. van Eijck
Slooter, Floor
Bergman, Jacques J.
De With, Peter H. N.
Jong, Martijn R.
de Groof, Albert J.
van der Sommen, Fons
author_facet Jaspers, Tim J. M.
Caetano, Francisco
Claessens, Cris H. B.
Kusters, Carolus H. J.
van Heslinga, Rixta A. H. van Eijck
Slooter, Floor
Bergman, Jacques J.
De With, Peter H. N.
Jong, Martijn R.
de Groof, Albert J.
van der Sommen, Fons
contents Computer-aided detection (CADe) of early neoplasia in Barrett's esophagus is a low-prevalence surveillance problem in which clinically relevant findings are rare. Although many CADe systems report strong performance on balanced or enriched datasets, their behavior under realistic prevalence remains insufficiently characterized. The RARE25 challenge addresses this gap by introducing a large-scale, prevalence-aware benchmark for neoplasia detection. It includes a public training set and a hidden test set reflecting real-world incidence. Methods were evaluated using operating-point-specific metrics emphasizing high sensitivity and accounting for prevalence. Eleven teams from seven countries submitted approaches using diverse architectures, pretraining, ensembling, and calibration strategies. While several methods achieved strong discriminative performance, positive predictive values remained low, highlighting the difficulty of low-prevalence detection and the risk of overestimating clinical utility when prevalence is ignored. All methods relied on fully supervised classification despite the dominance of normal findings, indicating a lack of prevalence-agnostic approaches such as anomaly detection or one-class learning. By releasing a public dataset and a reproducible evaluation framework, RARE25 aims to support the development of CADe systems robust to prevalence shift and suitable for clinical surveillance workflows.
format Preprint
id arxiv_https___arxiv_org_abs_2604_11171
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Development and evaluation of CADe systems in low-prevalence setting: The RARE25 challenge for early detection of Barrett's neoplasia
Jaspers, Tim J. M.
Caetano, Francisco
Claessens, Cris H. B.
Kusters, Carolus H. J.
van Heslinga, Rixta A. H. van Eijck
Slooter, Floor
Bergman, Jacques J.
De With, Peter H. N.
Jong, Martijn R.
de Groof, Albert J.
van der Sommen, Fons
Computer Vision and Pattern Recognition
Computer-aided detection (CADe) of early neoplasia in Barrett's esophagus is a low-prevalence surveillance problem in which clinically relevant findings are rare. Although many CADe systems report strong performance on balanced or enriched datasets, their behavior under realistic prevalence remains insufficiently characterized. The RARE25 challenge addresses this gap by introducing a large-scale, prevalence-aware benchmark for neoplasia detection. It includes a public training set and a hidden test set reflecting real-world incidence. Methods were evaluated using operating-point-specific metrics emphasizing high sensitivity and accounting for prevalence. Eleven teams from seven countries submitted approaches using diverse architectures, pretraining, ensembling, and calibration strategies. While several methods achieved strong discriminative performance, positive predictive values remained low, highlighting the difficulty of low-prevalence detection and the risk of overestimating clinical utility when prevalence is ignored. All methods relied on fully supervised classification despite the dominance of normal findings, indicating a lack of prevalence-agnostic approaches such as anomaly detection or one-class learning. By releasing a public dataset and a reproducible evaluation framework, RARE25 aims to support the development of CADe systems robust to prevalence shift and suitable for clinical surveillance workflows.
title Development and evaluation of CADe systems in low-prevalence setting: The RARE25 challenge for early detection of Barrett's neoplasia
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.11171