Scalable Segmentation for Ultra-High-Resolution Brain MR Images
This work addresses the problem of segmenting fine-scale anatomical structures in brain MR images for medical imaging researchers, representing an incremental improvement with novel design elements.
The paper tackles the challenge of accurate and efficient segmentation of ultra-high-resolution brain MR images by proposing a framework that uses low-resolution coarse labels as guidance without extra annotation cost, achieving superior performance and scalability compared to conventional methods.
Although deep learning has shown great success in 3D brain MRI segmentation, achieving accurate and efficient segmentation of ultra-high-resolution brain images remains challenging due to the lack of labeled training data for fine-scale anatomical structures and high computational demands. In this work, we propose a novel framework that leverages easily accessible, low-resolution coarse labels as spatial references and guidance, without incurring additional annotation cost. Instead of directly predicting discrete segmentation maps, our approach regresses per-class signed distance transform maps, enabling smooth, boundary-aware supervision. Furthermore, to enhance scalability, generalizability, and efficiency, we introduce a scalable class-conditional segmentation strategy, where the model learns to segment one class at a time conditioned on a class-specific input. This novel design not only reduces memory consumption during both training and testing, but also allows the model to generalize to unseen anatomical classes. We validate our method through comprehensive experiments on both synthetic and real-world datasets, demonstrating its superior performance and scalability compared to conventional segmentation approaches.