Acquisition Time-Informed Breast Tumor Segmentation from Dynamic Contrast-Enhanced MRI

2025-11-20

AI Summary

In DCE-MRI, substantial inter-scanner and inter-subject variability in tumor appearance—even within the same contrast-enhancement phase—severely undermines the robustness of cross-center automatic segmentation. To address this, we propose the first time-aware segmentation framework that incorporates image acquisition time as a structured prior. Specifically, we introduce a Feature-wise Linear Modulation (FiLM)-based temporal feature modulation mechanism that dynamically calibrates feature representations across all backbone layers. Our method requires no architectural modifications and is compatible with diverse mainstream backbones. Evaluated on multi-center DCE-MRI data, it consistently improves both intra-domain and cross-domain segmentation performance, achieving an average Dice score gain of 3.2%. Notably, improvements are most pronounced in early enhancement phases and for small tumors. This work establishes an interpretable, plug-and-play paradigm for temporal integration in sequential medical imaging, significantly enhancing clinical generalizability.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays an important role in breast cancer screening, tumor assessment, and treatment planning and monitoring. The dynamic changes in contrast in different tissues help to highlight the tumor in post-contrast images. However, varying acquisition protocols and individual factors result in large variation in the appearance of tissues, even for images acquired in the same phase (e.g., first post-contrast phase), making automated tumor segmentation challenging. Here, we propose a tumor segmentation method that leverages knowledge of the image acquisition time to modulate model features according to the specific acquisition sequence. We incorporate the acquisition times using feature-wise linear modulation (FiLM) layers, a lightweight method for incorporating temporal information that also allows for capitalizing on the full, variables number of images acquired per imaging study. We trained baseline and different configurations for the time-modulated models with varying backbone architectures on a large public multisite breast DCE-MRI dataset. Evaluation on in-domain images and a public out-of-domain dataset showed that incorporating knowledge of phase acquisition time improved tumor segmentation performance and model generalization.

Problem

• Automating breast tumor segmentation in DCE-MRI images
• Addressing tissue appearance variations from acquisition protocols
• Improving segmentation accuracy using acquisition time information

Innovation

• Using acquisition time to modulate model features
• Incorporating temporal data via FiLM layers
• Improving segmentation with phase time knowledge