视网膜血管分割中的观察者选择与阈值选取：一种受试者分离的评估
Observer Choice and Threshold Selection in Retinal Vessel Segmentation: A Subject-Separated Evaluation
浏览论文内容
中文总结 AI 辅助
本研究通过受试者分离的评估协议，系统比较五种阈值策略对视网膜血管分割的影响，发现最大最小调优虽改变多数阈值但无准确性收益，强调应明确报告阈值选择与评估参考。
中文摘要 AI 辅助
用于选择分割阈值的标注是评估协议的一部分，但其影响很容易与模型质量混淆。我们使用全部28张CHASE DB1图像和两位人工标注，检验了视网膜血管分割中的这一选择。一个固定的七折协议将14名受试者的双眼保持在一起。随机森林和Extra Trees针对观察者1的标注进行拟合，使用三个随机种子，共产生42个拟合结果。五种阈值策略共享相同的评分图：固定0.50、观察者1调优、观察者2调优、平均观察者调优，以及逐图像较低观察者Dice的最大最小调优。对于随机森林，最大最小调优在21个拟合中的19个改变了阈值，但最差观察者Dice从70.53%下降到70.45%。配对差异为-0.073个百分点，条件受试者自助法的95%置信区间为[-0.384, 0.238]。Extra Trees显示出相同的趋势。相同的观察者1调优随机森林掩膜在观察者1上的得分为73.66%，在观察者2上的得分为71.06%。结果支持明确报告阈值选择参考和评估参考；但不支持在该队列中最大最小调优带来准确性收益。所有划分、原始预测、指标和代码均已提供。AI辅助已披露。
英文摘要
The annotation used to select a segmentation threshold is part of the evaluation protocol, yet its effect is easily conflated with model quality. We examine this choice for retinal vessel segmentation using all 28 CHASE DB1 images and both human annotations. A fixed seven-fold protocol keeps both eyes of each of the 14 subjects together. Random forests and Extra Trees are fitted against observer 1 with three random seeds, yielding 42 fits. Five threshold policies share identical score maps: fixed 0.50, observer-1 tuning, observer-2 tuning, mean-observer tuning, and maximin tuning of the per-image lower observer Dice. For random forests, maximin changes the threshold in 19 of 21 fits, but worst-observer Dice decreases from 70.53 percent to 70.45 percent. The paired difference is -0.073 percentage points, with a conditional subject-bootstrap 95 percent interval of [-0.384, 0.238]. Extra Trees shows the same direction. Identical observer-1-tuned random-forest masks score 73.66 percent against observer 1 and 71.06 percent against observer 2. The results support explicit reporting of both the threshold-selection reference and evaluation reference; they do not support an accuracy benefit from maximin tuning in this cohort. All splits, raw predictions, metrics and code are supplied. AI assistance is disclosed.
发表机构
- Zhengzhou Police University（郑州警察学院）
- School of Artificial Intelligence, Beijing University of Posts and Telecommunications（北京邮电大学人工智能学院）
- Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）
- Qilu University of Technology and Jinan Supercomputing Center（齐鲁工业大学与济南超算中心）
机构由 AI 辅助整理，请以论文原文为准。
