PGR-Net：用于脑肿瘤 MRI 分割的先验引导 ROI 推理网络
PGR-Net: Prior-Guided ROI Reasoning Network for Brain Tumor MRI Segmentation
把「病灶只占体数据很小一部分」这一临床空间先验显式建模成 ROI 决策机制（分层 Top-K 候选区域选择 + 高斯窗口衰减引导图），可直接迁移到其他稀疏目标分割任务（如小器官、微转移灶、肺结节）以削减背景冗余计算。
原文依据 a hierarchical Top-K ROI decision mechanism that progressively selects the most confident lesion candidate regions across encoder layers
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论脑肿瘤 MRI 分割对临床诊断与治疗规划至关重要，可用于准确检测病灶并勾画放疗靶区。然而，肿瘤病灶仅占据体数据空间的一小部分，造成严重的空间稀疏性；同时现有分割网络往往忽略临床上观察到的肿瘤发生空间先验，导致在大量背景区域上进行冗余的特征计算。为解决该问题，我们提出 PGR-Net（先验引导 ROI 推理网络）——一个显式的 ROI 感知框架，它引入数据驱动的空间先验集合来刻画肿瘤病灶的分布与尺度特性，为更稳定的分割提供全局引导。借助这些先验，PGR-Net 引入分层 Top-K ROI 决策机制，在编码器各层中逐步选出置信度最高的病灶候选区域，以提升定位精度。我们进一步设计了 WinGS-ROI（窗口化高斯-空间衰减 ROI）模块，利用多窗口高斯模板与空间衰减函数生成中心增强的引导图，从而在整个网络中引导特征学习。基于这些 ROI 特征，采用窗口化 RetNet 骨干网络以增强定位的可靠性。在 BraTS 2019/2023 与 MSD Task01 上的实验表明，PGR-Net 在仅使用 8.64M 参数的情况下持续优于现有方法，在 Whole Tumor 区域上分别取得 89.02%、91.82% 和 89.67% 的 Dice 分数。代码已开源：https://github.com/CNU-MedAI-Lab/PGR-Net。
英文原文摘要 1598 字符
Brain tumor MRI segmentation is essential for clinical diagnosis and treatment planning, enabling accurate lesion detection and radiotherapy target delineation. However, tumor lesions occupy only a small fraction of the volumetric space, resulting in severe spatial sparsity, while existing segmentation networks often overlook clinically observed spatial priors of tumor occurrence, leading to redundant feature computation over extensive background regions. To address this issue, we propose PGR-Net (Prior-Guided ROI Reasoning Network) - an explicit ROI-aware framework that incorporates a data-driven spatial prior set to capture the distribution and scale characteristics of tumor lesions, providing global guidance for more stable segmentation. Leveraging these priors, PGR-Net introduces a hierarchical Top-K ROI decision mechanism that progressively selects the most confident lesion candidate regions across encoder layers to improve localization precision. We further develop the WinGS-ROI (Windowed Gaussian-Spatial Decay ROI) module, which uses multi-window Gaussian templates with a spatial decay function to produce center-enhanced guidance maps, thus directing feature learning throughout the network. With these ROI features, a windowed RetNet backbone is adopted to enhance localization reliability. Experiments on BraTS 2019/2023 and MSD Task01 show that PGR-Net consistently outperforms existing approaches while using only 8.64M Params, achieving Dice scores of 89.02%, 91.82%, and 89.67% on the Whole Tumor region. Code is available at https://github.com/CNU-MedAI-Lab/PGR-Net.