A Multi-Source Ultrasound Benchmark Revealing the Limits of Contemporary Self-Supervised Anomaly Detection Methods

Marco Riedenauer; Daniel Kienzle; Pratik Mayekar; Rainer Lienhart

cs.CVOct 7, 2026

A Multi-Source Ultrasound Benchmark Revealing the Limits of Contemporary Self-Supervised Anomaly Detection Methods

Authors: Marco Riedenauer, Daniel Kienzle, Pratik Mayekar, Rainer Lienhart

Organizations: University of Augsburg

Abstract

Self-supervised anomaly detection is a promising paradigm for medical ultrasound, as normal images are often easier to obtain than exhaustive annotations of all possible pathologies. However, most existing evaluations are limited to a single anatomy or task, making it unclear whether models learn a robust notion of normal ultrasound appearance or only a source-specific representation. We introduce the SADUSI benchmark, a multi-source ultrasound dataset designed to train and evaluate anomaly detection methods across a broad range of anatomical regions, views, and acquisition protocols. The goal of SADUSI is to provide a diverse normal ultrasound distribution and a benchmark for visible structural anomalies that can be assessed from single images. We evaluate representative self-supervised anomaly detection methods and find that current approaches struggle in this setting. In particular, reconstruction-based diffusion methods such as AnoDDPM and DeCo-Diff achieve pixel-level AUROC values of 0.56-0.72 and maximum F1 scores of 0.10-0.26, indicating limited separation of pathology from normal image regions. Feature-based PatchCore variants perform better, reaching pixel-level AUROC values of 0.76-0.83, but remain limited with maximum F1 scores of 0.14-0.40. These findings suggest that broad multi-source ultrasound anomaly detection remains an open challenge and that SADUSI can serve as a resource for developing methods that generalize beyond anatomy-specific settings.

Figures & tables

Source	Anatomy	Type	$\|\mathcal{D}_{n}\|$ / $\|\mathcal{D}^{\mathrm{f}}_{n}\|$ / $\|\mathcal{D}_{a}\|$	Use
AbdomenUS [ 19 ]	Abdominal	Images	616 / 616 / 0	Train
AUL [ 20 ]	Liver	Images	100 / 100 / 434	Train / Test
Breast-Lesions-USG [ 21 , 22 ]	Breast	Images	4 / 4 / 0	Train
BUS-UCLM [ 23 , 10 ]	Breast	Images	405 / 405 / 0	Train
BUSI [ 9 , 24 ]	Breast	Images	131 / 131 / 512	Train / Test
COVID-19 [ 25 ]	Lung	Images	62 / 62 / 0	Train

TABLE I: Source datasets included in SADUSI .

Eval. set	Training	Method	AUROC	AP	AUPRO	F1	Prec.	Rec.
TN3K	Filtered	AnoDDPM	0.59	0.14	0.17	0.23	0.14	0.82
TN3K	Filtered	DeCo-Diff	0.66	0.16	0.20	0.26	0.16	0.64
TN3K	Filtered	PatchCore-WRN	0.76	0.24	0.40	0.33	0.22	0.62
TN3K	Filtered	PatchCore-DINOv3	0.79	0.26	0.29	0.36	0.26	0.61
TN3K	Filtered	PatchCore-DINOv3 (A)	0.76	0.24	0.37	0.33	0.22	0.62
TN3K	Filtered	EfficientAD	0.50	0.11	0.07	0.20	0.11	0.94

TABLE II: Pixel-level anomaly localization results. “(A)” denotes data augmentation.

Fig. 1: Scaled pathology-background anomaly-map contrast by evaluation dataset, model, and training pool. Violin plots show the distributions and embedded box plots show their medians and interquartile ranges.

Panel of Figure 4 — Fig. 2: Patchwise DINOv3 feature-space analysis. From left to right: balanced normal–normal and normal–anomalous patch distances; cross-image rank-1 distance from each anomalous patch to its nearest normal patch; and fraction of normal patches in its cross-image $k=10$ neighborhood.

Fig. 3: Distribution of the image area covered by annotated pathology in the three evaluation datasets.

Explore similar work

May 25, 2026cs.CV

Anatomy-Anchored Self-Supervision: Distilling Vision Foundation Models for Invariant Ultrasound Representation

Self-supervised pre-training paradigm has gained increasing prominence for learning transferable representations in medical imaging, yet existing methods for ultrasound (US) images operate at the image or frame level, overlooking the anatomical context for clinical-aligned representation learning. In this work, we propose an anatomy-anchored ultrasound self-supervision framework ANAUS that shifts representation learning from generic visual regions to clinically meaningful anatomical structures. Utilizing a learnable latent prompt engine alongside a one-time domain adaptation on existing public image-mask pairs, we empower the LP-SAM module to achieve annotation-free anatomy delineation at scale. Building upon this anatomical grounding, we propose a dual-policy self-supervised learning paradigm consisting of inter-view semantics-aware anatomy-separating alignment and contextual core-region prediction to enhance representation learning. Specifically, the former enforces feature invariance within identical anatomical regions while promoting discriminability across distinct structures; the latter compels the model to reconstruct corrupted regions, thereby capturing fine-grained structural details. Extensive evaluations on six public datasets demonstrate that ANAUS consistently outstrips current state-of-the-art methods while maintaining the computational efficiency essential for clinical deployment. Code is available at https://github.com/zhcz328/ANAUS.

Chunzheng Zhu, Yijun Wang, Jianxin Lin +5

Hunan University, Changsha, China · Shenzhen Maternity and Child Healthcare Hospital, Shenzhen, China

Ultrasound Self-Supervised Learning Representation Learning Recent Vision Foundation Models

Ask PDF

Sep 21, 2026cs.AI

Unsupervised Brain Anomaly Detection as a Bayesian Inverse Problem with Diffusion Prior

Unsupervised anomaly detection (UAD) aims to localize abnormal regions in medical scans without pixel-level annotations. A typical strategy seeks to reconstruct a pseudo-healthy image that preserves subject-specific anatomy. Recently, diffusion models have been proposed to perform UAD. However, these methods rely on heuristic noise schedules or synthetic corruptions to balance subject-specificity and anomaly removal. In this work, we propose an alternative formulation of UAD as a Bayesian inverse problem under a diffusion prior. First, we introduce a latent spatial anomaly mask that models pixel-wise consistency between a test image and its latent corresponding pseudo-healthy image. Then, we propose an approximation of the unknown generation process that links healthy anatomy, anomalies, and the observed image, enabling a well-defined likelihood within the Bayesian framework. Building on recent advances in diffusion-based inverse problem methods, we jointly infer the pseudo-healthy image and the anomaly mask via annealed posterior sampling. We evaluate our approach on FDG PET (ADNI) and FLAIR MRI (BraTS 2021), demonstrating improved anomaly localization performance compared to other diffusion-based approaches and validating the contribution of our introduced model. Our code is available at https://github.com/HuguesRoy/UAD_DAPS.

Hugues Roy, Reuben Dorent, Ninon Burgos

Sorbonne Université, Institut du Cerveau - Paris Brain Institute - ICM, CNRS, Inria, Inserm, AP-HP, Hôpital de la Pitié Salpêtrière, F-75013, Paris, France

Graph Anomaly Detection Diffusion Priors Anatomical Priors Bayesian Inverse Problems

Ask PDF

Aug 1, 2026cs.CV

Reconstruction-Shift Discrimination via Mask-Guided Latent Diffusion for Medical Anomaly Detection

Unsupervised medical anomaly detection learns normal anatomical patterns from healthy training images and identifies deviations at test time. Reconstruction-based and diffusion-based methods commonly use the difference between an input image and its reconstruction as anomaly evidence. However, this residual can be ambiguous. Expressive models may preserve pathological structures, while benign anatomical variation, imaging noise, and acquisition differences may also produce large reconstruction errors. We propose discriminative mask-guided diffusion (DMD), a medical anomaly detection framework that complements residual-based localization with reconstruction-shift discrimination. DMD first learns a compact quantized latent representation of normal images. Localized masks then perturb selected latent regions, and a latent diffusion model reconstructs the perturbed representations. The resulting reconstructions are paired with their original normal images to define a self-supervised classification task. At inference, the classifier provides a learned image-level anomaly score, while the residual between the input and its diffusion-based reconstruction yields a pixel-level anomaly map. Experiments on five datasets spanning brain MRI, breast ultrasound, and chest radiography show that DMD achieves the best overall performance among the state-of-the-art baseline methods.

Yibo Wan, Jinyu Cai, Yunhe Zhang +2

School of Computing, National University of Singapore, Singapore · Institute of Data Science, National University of Singapore, Singapore · Department of Computer and Information Science, SKL-IOTSC, University of Macau, China +1

Identity-Conditioned Latent Diffusion Models

Ask PDF