Improving Vessel Segmentation with Multi-Task Learning and Auxiliary Data Available Only During Model Training

2025-09-04

AI Summary

Accurate hepatic vessel segmentation in non-contrast-enhanced MRI remains challenging due to low contrast and severe scarcity of pixel-level annotations. Method: We propose a multi-task learning framework that leverages paired contrast-enhanced MRI images *only during training* as an auxiliary modality. A shared encoder jointly optimizes vessel segmentation and contrast-enhanced image reconstruction, eliminating the need for auxiliary modality access during inference. Contribution/Results: Our approach establishes a novel “training-only auxiliary, inference-free dependency” weakly supervised paradigm, supporting partial labeling (i.e., some training samples lack vessel annotations). Experiments demonstrate significant improvements in segmentation accuracy—particularly under limited-label regimes—and strong cross-domain generalizability, as validated by successful transfer to brain tumor segmentation.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Liver vessel segmentation in magnetic resonance imaging data is important for the computational analysis of vascular remodelling, associated with a wide spectrum of diffuse liver diseases. Existing approaches rely on contrast enhanced imaging data, but the necessary dedicated imaging sequences are not uniformly acquired. Images without contrast enhancement are acquired more frequently, but vessel segmentation is challenging, and requires large-scale annotated data. We propose a multi-task learning framework to segment vessels in liver MRI without contrast. It exploits auxiliary contrast enhanced MRI data available only during training to reduce the need for annotated training examples. Our approach draws on paired native and contrast enhanced data with and without vessel annotations for model training. Results show that auxiliary data improves the accuracy of vessel segmentation, even if they are not available during inference. The advantage is most pronounced if only few annotations are available for training, since the feature representation benefits from the shared task structure. A validation of this approach to augment a model for brain tumor segmentation confirms its benefits across different domains. An auxiliary informative imaging modality can augment expert annotations even if it is only available during training.

Problem

• Segmenting liver vessels in non-contrast MRI images
• Reducing need for annotated training data with auxiliary contrast-enhanced MRI
• Improving segmentation accuracy when few annotations are available

Innovation

• Multi-task learning with auxiliary contrast-enhanced MRI
• Exploiting paired native and contrast data training
• Shared task structure improves feature representation accuracy