Standardized Evaluation of Automatic Methods for Perivascular Spaces Segmentation in MRI -- MICCAI 2024 Challenge Results

2025-12-19

AI Summary

Automatic segmentation of perivascular spaces (PVS), particularly enlarged PVS (EPVS), faces challenges including small size, morphological variability, ambiguity with pathological lesions, and scarcity of expert annotations. Method: We organized the MICCAI 2024 EPVS Challenge and introduced the first multi-center, multi-sequence, whole-brain MRI benchmark dataset (100/50/50 cases for training/validation/testing), with whole-brain EPVS annotations standardized according to the STRIVE guidelines. We evaluated methods based on U-Net and MedNeXt architectures, incorporating 2D/3D dual-path networks, multimodal input, ensemble learning, and Transformer modules. Contribution/Results: The top-performing method achieved a Dice score of 0.68; however, cross-site generalization dropped by over 30%, revealing domain shift as the primary bottleneck in current EPVS segmentation. This benchmark establishes a rigorous, community-validated evaluation framework and provides critical insights for developing robust imaging biomarkers of cerebral small vessel disease.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Perivascular spaces (PVS), when abnormally enlarged and visible in magnetic resonance imaging (MRI) structural sequences, are important imaging markers of cerebral small vessel disease and potential indicators of neurodegenerative conditions. Despite their clinical significance, automatic enlarged PVS (EPVS) segmentation remains challenging due to their small size, variable morphology, similarity with other pathological features, and limited annotated datasets. This paper presents the EPVS Challenge organized at MICCAI 2024, which aims to advance the development of automated algorithms for EPVS segmentation across multi-site data. We provided a diverse dataset comprising 100 training, 50 validation, and 50 testing scans collected from multiple international sites (UK, Singapore, and China) with varying MRI protocols and demographics. All annotations followed the STRIVE protocol to ensure standardized ground truth and covered the full brain parenchyma. Seven teams completed the full challenge, implementing various deep learning approaches primarily based on U-Net architectures with innovations in multi-modal processing, ensemble strategies, and transformer-based components. Performance was evaluated using dice similarity coefficient, absolute volume difference, recall, and precision metrics. The winning method employed MedNeXt architecture with a dual 2D/3D strategy for handling varying slice thicknesses. The top solutions showed relatively good performance on test data from seen datasets, but significant degradation of performance was observed on the previously unseen Shanghai cohort, highlighting cross-site generalization challenges due to domain shift. This challenge establishes an important benchmark for EPVS segmentation methods and underscores the need for the continued development of robust algorithms that can generalize in diverse clinical settings.

Problem

• Develops automated segmentation for enlarged perivascular spaces in MRI
• Addresses challenges from small size and limited annotated datasets
• Evaluates cross-site generalization to overcome domain shift issues

Innovation

• U-Net architectures with multi-modal processing
• Ensemble strategies and transformer-based components
• MedNeXt architecture with dual 2D/3D strategy