Task & Evaluation¶

Task Description 📝¶

The core task of the RARE 2025 challenge is binary classification: determining whether an endoscopic image of a Barrett’s Esophagus (BE) patient contains early neoplasia or not. The goal is to build AI algorithms that can identify subtle but critical signs of early-stage cancer while maintaining a low false positive rate — a key requirement in real-world clinical use.

Evaluation 📈¶

The evaluation of the algorithms will be done in two phases: the Open Development Phase and the Closed Testing Phase.

In both phases, the key performance metric that will be used to evaluate the AI algorithms is the Positive Predictive Value at 90% Recall / Sensitivity (PPV@90Recall). During the Open Development Phase, participators will also be able to evaluate their algorithms based on the Area Under the Receiver Operating Characteristic Curve (AUC-ROC) and the Area Under the Precision-Recall Curve (AUC-PRC), however, these metrics will not contribute to the final ranking of the algorithms. The results of the Closed Testing Phase will not be shared with the participants before presentation of the results at MICCAI 2025.