Problem
Research questions and friction points this paper is trying to address.
CT encoders
coarse semantic understanding
fine-grained anatomical structures
self-supervised pretraining
B-MIM: Biased Masked Image Modeling for Generalizable Segmentation of Fine-Grained Anatomical Structures
🤖 AI Summary
为提高细粒度解剖结构分割的泛化能力，本文提出B-MIM方法，通过减少全局语义对齐来优先局部补丁重建，从而增强3D Swin Transformer编码器捕捉高频形态细节的能力。
📝 Abstract
Self-supervised pretraining enables transferable representations for medical imaging, yet most CT encoders remain biased toward coarse semantic understanding, limiting their sensitivity to fine-grained anatomical structures such as vessels or small tumors. In this paper, we introduce Biased Masked Image Modeling (B-MIM), a modification of the iBOT objective that stochastically reduces global semantic alignment to prioritize local patch reconstruction. This bias encourages the encoder to capture high-frequency morphological details and structural continuity. We curate a multi-institutional CT abdominal dataset of 9,955 filtered studies from 17 public sources and pretrain a 3D Swin Transformer backbone using B-MIM. Across inter-dataset experiments on liver vessel segmentation, the proposed encoder improves topological fidelity (clDice) and achieves competitive Dice scores in tumor segmentation, compared to fully fine-tuned baselines, despite updating only a fraction of the parameters. Our results suggest that reducing global semantic pressure during pretraining enhances generalization to intricate anatomical structures.
🔎 Similar Papers
No similar papers found.
