用高效视觉变压器区分癌细胞有丝分裂的正常与异常，准确率超85%。
Normal and Atypical Mitosis Image Classifier using Efficient Vision Transformer
- 采用混合CNN-ViT架构，结合染色解卷积增强数据
- 在7种癌症类型上达到0.859平衡准确率和0.942的AUC
- 适合病理图像分析与医学影像智能诊断研究者
我们针对MIDOG 2025挑战赛中的正常与异常有丝分裂分类问题，采用EfficientViT-L2这一兼顾精度与效率的混合CNN-ViT架构。使用包含13,938个细胞核的统一数据集（MIDOG++和AMi-Br），其中异常有丝分裂占比约15%。为评估领域泛化能力，采用留一癌症类型排除的五折集成交叉验证，并通过染色解卷积进行图像增强。挑战提交版本在相同五折划分下对所有癌症类型进行训练。在预评阶段，该模型取得0.859的平衡准确率、0.942的ROC AUC和0.85的原始准确率，各项指标表现优异且均衡。
原文摘要 · Abstract (English)
We tackle atypical versus normal mitosis classification in the MIDOG 2025 challenge using EfficientViT-L2, a hybrid CNN--ViT architecture optimized for accuracy and efficiency. A unified dataset of 13,938 nuclei from seven cancer types (MIDOG++ and AMi-Br) was used, with atypical mitoses comprising ~15. To assess domain generalization, we applied leave-one-cancer-type-out cross-validation with 5-fold ensembles, using stain-deconvolution for image augmentation. For challenge submissions, we trained an ensemble with the same 5-fold split but on all cancer types. In the preliminary evaluation phase, this model achieved balanced accuracy of 0.859, ROC AUC of 0.942, and raw accuracy of 0.85, demonstrating competitive and well-balanced performance across metrics.
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。