Prototype Memory-Guided Training-Free Anomaly Classification and Localization in Prenatal Ultrasound

2026-07-01

AI Summary

This work addresses the challenge of multi-class anomaly classification and localization in prenatal ultrasound, where abnormalities are rare, highly heterogeneous, and lack large-scale annotated data. The authors propose the first training-free, zero-shot anomaly detection framework that requires only a few reference images per class. By constructing a multi-granularity prototype memory bank, designing a prototype-driven soft fusion mechanism, and introducing a category-aware prediction refinement strategy, the method jointly models semantic and anomalous features. Evaluated on a multi-center dataset comprising 1,149 cases, 2,357 images, and nine anomaly classes, the approach significantly outperforms existing methods, demonstrating strong effectiveness and generalization capability.

Abstract

Prenatal anomaly classification and localization is of critical importance for fetal health and pregnancy management. Although ultrasound (US) is the primary modality for prenatal screening, accurate diagnosis remains challenging due to the low prevalence and high heterogeneity of anomalies. Existing deep learning methods for prenatal tasks rely on large-scale annotated datasets, which are difficult to obtain in practice. Although few-shot learning alleviates data scarcity, it typically requires fine-tuning for new categories, limiting its practicality in resource-limited clinical settings. To address these challenges, we propose a training-free framework for multi-class prenatal US anomaly classification and localization that operates with only a few reference images per class, representing the first exploration of this setting. Our framework comprises three key components: (1) a memory bank with multi-granular prototypes that explicitly models both class-level semantics and anomaly characteristics; (2) a prototype-driven soft merging mechanism that aggregates discriminative features to detect the anomaly region; and (3) a class-aware refinement strategy that leverages prototype consistency to improve category prediction. Extensively validated on a multi-center prenatal US dataset containing 1,149 cases, with a total of 2,357 images and 9 categories, our proposed method outperforms the competitors.

Problem

• prenatal ultrasound
• anomaly classification
• anomaly localization
• training-free
• few-shot learning

Innovation

• training-free
• prototype memory
• anomaly localization
• few-shot learning
• prenatal ultrasound