arXiv cs.AI· Zhenye Lou, Qing Xu, Wenting Duan, Zhen Chen·· 10 小时前AI 评分18
COMPASS-METS：面向脑转移瘤分割的区域级 Logit 融合与效用门控细化方法
Region-Wise Logit Fusion and Utility-Gated Refinement for Brain Metastasis Segmentation
AI 导读
COMPASS-METS 将三个五折 nnU-Net 配置组合为对齐的区域概率集成，通过结构化概率锚点完成区域组件决策、TC 边界补全、RC 过滤与层级保持转换，并用效用门控仅接纳满足病灶存在与几何安全概率的孤立候选。
正文 · 原文
Abstract:BraTS-METS 2026 Task 1 evaluates segmentation of brain metastases and postoperative resection cavities on heterogeneous multiparametric MRI, requiring detection of small enhancing foci while preserving the geometry and hierarchy of enhancing tumor (ET), tumor core (TC), whole tumor (WT), and resection cavity (RC). We present COMPASS-METS (Component-Aware Multi-Trainer Probability Assembly with Structured Segmentation), which combines three five-fold nnU-Net configurations into aligned regional probability ensembles. A structured-probability anchor applies region-specific component decisions, TC boundary completion, RC filtering, and hierarchy-preserving conversion. Omitted ET candidates are recovered additively using learned component confidence and deterministic multi-trainer support for ET and its TC/WT parents; a final utility gate admits only disconnected candidates with sufficient lesion-existence and geometry-safety probabilities. On the reused 179-case challenge development cohort, COMPASS-METS (Team: MicroBT) achieved Small F1 0.354, lesion-wise DSC 0.712, NSD 0.714, All F1 0.779, and HD95 56.6 mm. Relative to ResEncXL, it improved DSC, NSD, and All F1 and reduced HD95, while Small F1 changed from 0.367 to 0.354. Stagewise analysis attributes the principal geometric gains to structured probability construction and shows that guarded ET refinement recovers most of the associated small-lesion detection loss. These 179-case results remain selection-aware development estimates. The source code is publicly available at this https URL.
| Comments: | Accepted at the Brain Tumor Segmentation Challenge - Metastases (BraTS-METS), held in conjunction with MICCAI 2026 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.13884 [cs.CV] |
| (or arXiv:2611.13884v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13884 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Wenting Duan [view email]
[v1]
Tue, 10 Nov 2026 10:44:47 UTC (554 KB)
来源：arXiv cs.AI · arxiv.org