arXiv:2511.03912v2 Announce Type: replace 
Abstract: Unknown anomaly detection in medical imaging remains a fundamental challenge due to the scarcity of labeled anomalies and the high cost of expert supervision. We introduce an unsupervised, oracle-free framework that incrementally expands a trusted set of normal samples without any anomaly labels. Starting from a small, verified seed of normal images, our method alternates between lightweight adapter updates and uncertainty-gated sample admission. A frozen pretrained vision backbone is augmented with tiny convolutional adapters, ensuring rapid domain adaptation with negligible computational overhead. Extracted embeddings are stored in a compact coreset enabling efficient k-nearest neighbor anomaly (k-NN) scoring. Safety during incremental expansion is enforced by dual probabilistic gates, a sample is admitted into the normal memory only if its distance to the existing coreset lies within a calibrated z-score threshold, and its SWAG-based epistemic uncertainty remains below a seed-calibrated bound. This mechanism prevents drift and false inclusions without relying on generative reconstruction or replay buffers. Empirically, our system steadily refines the notion of normality as unlabeled data arrive, producing substantial gains over baselines. On COVID-CXR, ROC-AUC improves from 0.9489 to 0.9982 (F1: 0.8048 to 0.9746); on Pneumonia CXR, ROC-AUC rises from 0.6834 to 0.8968; and on Brain MRI ND-5, ROC-AUC increases from 0.6041 to 0.7269 and PR-AUC from 0.7539 to 0.8211. These results highlight the effectiveness and efficiency of the proposed framework for real-world, label-scarce medical imaging applications.

تم تقديم إطار عمل جديد غير خاضع للإشراف للكشف عن الشذوذات غير المعروفة في التصوير الطبي، مما يعالج تحدي نقص البيانات المعلّمة وارتفاع تكاليف الإشراف من قبل الخبراء. يبني هذا الأسلوب تدريجياً مجموعة موثوقة من العينات الطبيعية، مما يحسن بشكل كبير أداء الكشف عن الشذوذات، كما يتضح من تحسين درجات ROC-AUC عبر مجموعات بيانات التصوير المختلفة.

Se ha introducido un nuevo marco no supervisado para la detección de anomalías desconocidas en la imagenología médica, abordando el desafío de la escasez de datos etiquetados y los altos costos de supervisión por expertos. Este método construye progresivamente un conjunto de muestras normales de confianza, mejorando significativamente el rendimiento de la detección de anomalías, como lo demuestran los puntajes ROC-AUC mejorados en varios conjuntos de datos de imagenología.

Un nouveau cadre non supervisé pour la détection d'anomalies inconnues en imagerie médicale a été introduit, répondant au défi des données étiquetées limitées et des coûts élevés de supervision par des experts. Cette méthode construit progressivement un ensemble de samples normaux de confiance, améliorant considérablement les performances de détection d'anomalies, comme en témoignent les scores ROC-AUC améliorés sur divers ensembles de données d'imagerie.

A new unsupervised framework for detecting unknown anomalies in medical imaging has been introduced, addressing the challenge of limited labeled data and high expert supervision costs. This method incrementally builds a trusted set of normal samples, enhancing anomaly detection performance significantly, as evidenced by improved ROC-AUC scores across various imaging datasets.

I Detect What I Don't Know: Incremental Anomaly Learning with Stochastic Weight Averaging-Gaussian for Oracle-Free Medical Imaging

Was this article worth reading? Share it

Ready to build your own newsroom?