Panda: Unsupervised Pelvic Anomaly Detection for Real-Time MR Imaging
2026-07-27 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed a computer method to detect unusual problems in female pelvic MRI scans without needing labeled examples of diseases. Their system learns what normal looks like using healthy scans and then spots differences that might be abnormalities. It runs fast enough to give real-time hints during MRI scans, helping doctors make quicker decisions. Tests on uterine images showed the method works well at pinpointing problems accurately.
pelvic MRIunsupervised anomaly detectionDINOv3 Vision TransformerMLP bottleneckLinear Attentioncosine distanceUterine Myoma DatasetAUROCreal-time imagingimage-guided procedures
Authors
Anika Knupfer, Maximilian Lindholz, Johanna Paula Müller, Jordina Aviles Verdera, Smiti Tripathy, Susanne Schulz-Heise, Jana Hutter
Abstract
Female pelvic diseases remain an under researched area characterized by often delayed diagnosis. While pelvic MRI offers superior soft-tissue contrast for diagnosis and image-guided procedures, real-time anomaly detection remains challenging due to physiological motion, tissue deformation, and instrument artifacts. Existing supervised approaches are impractical, as adverse events are rare, heterogeneous, and difficult to annotate. We present a Dinomaly-based unsupervised anomaly detection framework adapted for pelvic MRI that learns normative representations from healthy cases and flags deviations without requiring labels. Our approach leverages a frozen DINOv3 Vision Transformer encoder combined with a noisy MLP bottleneck and Linear Attention decoder to prevent identity mapping while maintaining computational efficiency. Anomalies are localized via per-token cosine distance between encoder and decoder representations, yielding spatial anomaly maps that provide immediate feedback at the scanner to support radiologist decision-making and adaptive protocol adjustment. Evaluated on a curated subset of the Uterine Myoma Dataset, the framework achieves a pixel-level AUROC of 88.06% and high specificity (95.45%) at frame level at 40.5 slices/s, meeting real-time clinical deployment requirements. The spatial anomaly maps and frame-level scores provide immediate, localized feedback at the scanner to support radiologist decision-making and adaptive protocol adjustment during active procedures.