针对病理图像分析，定制化模型能有效避免过拟合。
Overfitting in Histopathology Model Training: The Need for Customized Architectures
- 为病理图像设计专用模型，而非直接套用自然图像模型。
- 简单专用架构在有限数据下表现优于复杂模型。
- 适合小样本病理分析研究者参考使用。
本研究探讨深度学习模型在病理图像分析中过拟合的严重问题。结果表明，直接采用为自然图像设计的大规模模型并进行微调，往往导致性能不佳和显著过拟合。通过在多种架构（包括ResNet变体和Vision Transformers）上进行大量实验，我们发现增加模型容量并不必然提升病理数据集上的表现。研究强调需为病理图像分析设计专用架构，尤其在数据量有限时更为重要。基于食管腺癌公开数据集（Oesophageal Adenocarcinomas），我们证明：更简单的领域特定架构可在保持甚至超越现有性能的同时，显著降低过拟合风险。
原文摘要 · Abstract (English)
This study investigates the critical problem of overfitting in deep learning models applied to histopathology image analysis. We show that simply adopting and fine-tuning large-scale models designed for natural image analysis often leads to suboptimal performance and significant overfitting when applied to histopathology tasks. Through extensive experiments with various model architectures, including ResNet variants and Vision Transformers (ViT), we show that increasing model capacity does not necessarily improve performance on histopathology datasets. Our findings emphasize the need for customized architectures specifically designed for histopathology image analysis, particularly when working with limited datasets. Using Oesophageal Adenocarcinomas public dataset, we demonstrate that simpler, domain-specific architectures can achieve comparable or better performance while minimizing overfitting.
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。