arXiv cs.AI· Sami Alnawali, Muhammad Atif Saeed, Abdul Hadi, Azka Atiq, Akhtar Jamil, Ahmad Din·· 4 小时前AI 评分29
基于医学影像的可解释深度学习乳腺癌诊断研究
Interpretable Deep Learning Models for Breast Cancer Diagnosis Using Medical Imaging
AI 导读
一项研究在 BreakHis 乳腺组织病理学数据集上以患者级划分基准测试了 10 个 ImageNet 预训练模型，并用 Grad-CAM、LIME、Integrated Gradients 和 LRP 解释预测。
正文 · 原文
Abstract:Deep learning shows promise for breast cancer diagnosis, but limited interpretability hinders clinical adoption. We benchmark ten ImageNet-pretrained models (ResNet50, VGG19, InceptionV3, DenseNet121, Xception, MobileNetV2, EfficientNet, ConvNeXt, ViT, and Swin Transformer) for benign-versus-malignant classification on the BreakHis histopathology dataset using a patient-level split, and explain their predictions with Grad-CAM, LIME, Integrated Gradients, and LRP. DenseNet121 performed best, with 99.4% test accuracy and an AUC-ROC of 0.999, and significantly outperformed all other models (p < 0.001). Transformer-based models trailed the CNNs. The explanations consistently highlighted nuclei-rich and glandular regions of diagnostic relevance. Combining transfer learning with explainable AI can thus yield accurate, interpretable tools for breast cancer diagnosis.
| Comments: | 37 page, 45 Figures |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.37026 [cs.CV] |
| (or arXiv:2610.37026v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.37026 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Muhammad Atif Saeed Mr. [view email]
[v1]
Tue, 27 Oct 2026 11:01:50 UTC (33,532 KB)
来源：arXiv cs.AI · arxiv.org