CSWinUNETR: Segmentation of Thin Anatomical Structures in Medical Images
This work addresses the persistent challenge of accurately segmenting thin, tortuous structures in medical images, which is critical for clinical diagnosis and treatment planning.
CSWinUNETR proposes a cross-shaped stripe self-attention mechanism and sparse-control dynamic snake convolution to improve segmentation of thin anatomical structures, achieving state-of-the-art results on four benchmarks across ophthalmology, neurovascular imaging, and dermatology without task-specific post-processing.
Accurate segmentation of thin, tortuous anatomical structures, such as retinal vessels, cerebral vasculature, and facial wrinkles, remains challenging due to low contrast, frequent discontinuities, and severe class imbalance. Although recent convolutional and Transformer-based models have improved performance, they often yield fragmented predictions and fail to recover fine branches. We propose CSWinUNETR, a general-purpose backbone for 2D and 3D thin-structure segmentation. It employs cross-shaped stripe self-attention to model long-range principal-axis context and incorporates cyclic shifts to enhance information exchange across stripes. To better preserve fine-grained details, we further introduce a detail-enhanced multi-scale self-attention module that aggregates contextual features from multi-resolution representations. In addition, we propose sparse-control dynamic snake convolution, which reconstructs reliable dense curvilinear kernels from sparsely predicted control points to better follow tortuous geometry. Extensive experiments on four benchmarks across ophthalmology, neurovascular imaging, and dermatology demonstrate that CSWinUNETR consistently outperforms state-of-the-art methods without task-specific post-processing or topology-aware losses. The code is available at https://github.com/labhai/CSWinUNETR.