Screener: Self-supervised Pathology Segmentation Model for 3D Medical Images

2025-02-12

AI Summary

Unsupervised segmentation of entire lesions in 3D medical imaging remains challenging—existing supervised methods are constrained by labeled categories and fail to generalize to unseen pathologies. Method: We propose the first Unsupervised Visual Anomaly Segmentation (UVAS) framework tailored for 3D CT. Our approach establishes a dense self-supervised feature extraction paradigm grounded in density estimation, employing learnable, mask-invariant dense features as conditional variables while eliminating handcrafted positional encodings. Contribution/Results: Trained on over 30,000 unlabeled 3D CT volumes, our method significantly outperforms prior UVAS approaches across four large-scale benchmark datasets (1,820 cases). It achieves, for the first time, robust and generalizable detection of multiple classes of unlabeled pathologies—marking a paradigm shift toward fully unsupervised medical image analysis.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Accurate segmentation of all pathological findings in 3D medical images remains a significant challenge, as supervised models are limited to detecting only the few pathology classes annotated in existing datasets. To address this, we frame pathology segmentation as an unsupervised visual anomaly segmentation (UVAS) problem, leveraging the inherent rarity of pathological patterns compared to healthy ones. We enhance the existing density-based UVAS framework with two key innovations: (1) dense self-supervised learning (SSL) for feature extraction, eliminating the need for supervised pre-training, and (2) learned, masking-invariant dense features as conditioning variables, replacing hand-crafted positional encodings. Trained on over 30,000 unlabeled 3D CT volumes, our model, Screener, outperforms existing UVAS methods on four large-scale test datasets comprising 1,820 scans with diverse pathologies. Code and pre-trained models will be made publicly available.

Problem

• Unsupervised pathology segmentation in 3D
• Self-supervised learning for feature extraction
• Improved anomaly detection in medical images

Innovation

• Self-supervised learning for feature extraction
• Masking-invariant dense features
• Unsupervised visual anomaly segmentation