通过自适应后处理提升脑卒中病灶实例检测效果
Adaptive Post-Processing Drives Instance-Level Detection in Stroke Lesion Segmentation
- 根据病灶负担动态调整组件大小阈值，实现病例自适应后处理
- 使病灶检测F1提升0.032，效果是模型架构改进的6倍
- 特别提升小病灶检测率，适合关注临床实用性的研究者
实例级病灶检测在医学图像分割中日益重要，但多数流程仍仅优化体素重叠。尤其对小病灶，近似匹配（重叠略低于阈值）与完全漏检得分相同，造成评估偏差。我们提出的体积条件自适应后处理（VCAP）根据每个病例的预测病灶负荷动态调整组件阈值，在5折交叉验证的1,453例数据集上，使病灶F1提升0.032（无偏估计），约为模型架构改进效果的6倍。采用针对小病灶设计的分辨率感知注意力架构（Viola2Plus），虽未改变小病灶Dice，但检测率提高3.7%，此效果仅靠体素重叠指标无法发现。最终两模型集成经后处理后，达到Dice 0.651、Lesion-F1 0.614，优于未处理单模型基线（Dice 0.644，Lesion-F1 0.573）。
原文摘要 · Abstract (English)
Instance-level lesion detection has been an increasingly larger focal point in medical image segmentation besides the more standard voxel-level overlap. Still, most pipelines are trained and post-processed for voxel overlap alone. In particular, the mismatch is most pronounced for small lesions, where a near-miss prediction---substantial overlap that falls just short of the instance-matching threshold---scores the same as a complete miss. In our ISLES'26 submission, we found that closing this gap mattered far more in post-processing than in architecture design. Our Volume-Conditioned Adaptive Post-Processing (VCAP) scheme adjusts component-size thresholds to each case's predicted lesion burden, improving Lesion-F1 by 0.032 (unbiased cross-fold estimate)---approximately 6 times larger than any architectural change we tested. A resolution-aware attention architecture (Viola2Plus), designed for small-lesion segmentation, shows why the distinction matters: it left small-lesion Dice unchanged but raised small-lesion detection rate by 3.7\%, a real effect voxel-overlap metrics alone would have missed. Under 5-fold cross-validation on the 1,453-case training set, our post-processed two-architecture ensemble achieves Dice 0.651 and Lesion-F1 0.614, versus 0.644 and 0.573 for the unprocessed single-model baseline.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。