用大模型+可解释分类器提升高危人群乳腺癌病灶检测准确率
Toward a robust lesion detection model in breast DCE-MRI: adapting foundation models to high-risk women
- 用自监督预训练的医学切片变压器提取图像特征
- 结合KAN分类器实现0.80的AUC性能，优于基线模型
- 通过热力图保持可解释性，适合临床医生信任使用
准确的乳腺MRI病灶检测对早期癌症诊断至关重要，尤其在高危人群中。本文提出一种分类流程，将预训练的基础模型Medical Slice Transformer（MST）适配于动态对比增强MRI（DCE-MRI）的乳腺病灶分类任务。MST基于DINOv2的自监督预训练生成鲁棒的每切片特征嵌入，并用于训练柯尔莫哥洛夫-阿诺德网络（KAN）分类器。KAN通过自适应B样条激活实现局部非线性变换，提供比传统卷积网络更灵活且可解释的分类方案，增强了模型在不平衡、异质临床数据集上区分良恶性病灶的能力。实验表明，MST+KAN流程在基准测试中达到AUC = 0.80 ± 0.02，同时通过注意力热图保持可解释性。研究结果表明，将基础模型嵌入与先进分类策略结合，可构建更稳健、泛化能力强的乳腺MRI分析工具。
原文摘要 · Abstract (English)
Accurate breast MRI lesion detection is critical for early cancer diagnosis, especially in high-risk populations. We present a classification pipeline that adapts a pretrained foundation model, the Medical Slice Transformer (MST), for breast lesion classification using dynamic contrast-enhanced MRI (DCE-MRI). Leveraging DINOv2-based self-supervised pretraining, MST generates robust per-slice feature embeddings, which are then used to train a Kolmogorov--Arnold Network (KAN) classifier. The KAN provides a flexible and interpretable alternative to conventional convolutional networks by enabling localized nonlinear transformations via adaptive B-spline activations. This enhances the model's ability to differentiate benign from malignant lesions in imbalanced and heterogeneous clinical datasets. Experimental results demonstrate that the MST+KAN pipeline outperforms the baseline MST classifier, achieving AUC = 0.80 \pm 0.02 while preserving interpretability through attention-based heatmaps. Our findings highlight the effectiveness of combining foundation model embeddings with advanced classification strategies for building robust and generalizable breast MRI analysis tools.
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。