对比三种模型在不同医院数据上的表现，发现视觉变压器更适应跨机构应用。
Do Diabetic Foot Ulcer Segmentation Models Generalize? A Cross-Dataset Benchmark of CNN and Transformer Architectures
- 统一测试流程下比较卷积与变压器架构的跨数据集泛化能力
- 变压器模型在两个外部数据集上分割准确率均领先，最高达0.786
- 模型复杂度不是关键，架构类型才决定能否跨医院稳定使用
针对糖尿病足溃疡（DFU）分割的深度学习模型通常报告高精度，但几乎都在同一数据集上训练和测试，对其在不同临床来源数据上的表现缺乏评估。本文采用一致且防泄露的测试协议，对三种代表性分割架构——U-Net和DeepLabV3+（卷积型）以及SegFormer-B2（Transformer）——进行基准测试：在合并的FUSeg/AZH伤口数据上训练，不微调直接评估在两个独立外部数据集（DFUC2022和Medetec）上的表现。所有模型在本域内性能良好（Dice 0.80–0.83），但在跨数据集时显著下降。然而，退化程度因架构而异：SegFormer-B2在两个外部数据集上泛化最佳（DFUC2022 Dice 0.557，Medetec Dice 0.786），优于两个卷积模型；更复杂的DeepLabV3+反而比更简单的U-Net泛化更差。对2,160张图像的逐图失败分析表明，SegFormer-B2在DFUC2022上出现灾难性失败的比例最低（31.1%），低于U-Net（38.5%）和DeepLabV3+（43.0%）。在两个独立外部数据集上的一致排名经威尔科克森符号秩检验验证（两组p < 0.001），说明架构类别而非模型复杂度是影响跨院泛化能力的关键因素。
原文摘要 · Abstract (English)
Deep learning models for diabetic foot ulcer (DFU) segmentation routinely report high accuracy, but they are almost always trained and tested on the same dataset, leaving their behaviour on data from a different clinical source largely unmeasured. We benchmark three representative segmentation architectures -- U-Net and DeepLabV3+ (convolutional) and SegFormer-B2 (Transformer) -- under an identical, leakage-screened protocol: training on the combined FUSeg/AZH wound data and evaluating, without fine-tuning, on two independent external datasets (DFUC2022 and Medetec). All models achieve strong in-domain performance (Dice 0.80--0.83) but degrade substantially across datasets. The degradation is, however, architecture-dependent: SegFormer-B2 generalizes best on both external sets (DFUC2022 Dice 0.557, Medetec Dice 0.786), outperforming both convolutional models, while the more complex DeepLabV3+ generalizes worse than the simpler U-Net. Per-image failure analysis on 2,160 images across both external test sets confirms that SegFormer-B2 produces the fewest catastrophic failures on DFUC2022 (31.1%), compared with U-Net (38.5%) and DeepLabV3+ (43.0%). The consistent ranking across two independent external sources, confirmed by Wilcoxon signed-rank tests (p < 0.001 on both datasets), indicates that architecture family, not model complexity, drives cross-hospital generalization.
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。