基于两家机构DCE-MRI多序列融合的深度学习模型预测乳腺癌分子亚型
这项回顾性研究纳入两家机构366例乳腺癌患者,构建多分支卷积神经网络(MBCNN)并融合DCE-MRI多序列信息预测分子亚型;在中间融合与80像素ROI下表现最优,四分类准确率为0.64,优于CNN的0.48和CLSTM的0.52。
要点
- MBCNN在中间融合与ROI 80像素时预测各分子亚型表现最优。
- 四亚型预测准确率MBCNN为0.64,高于CNN的0.48和CLSTM的0.52。
- 预测TN亚型时MBCNN的AUC显著优于CNN(p=0.03)。
核心结论
- 融合两家机构DCE-MRI特征提取与融合的深度学习模型可实现乳腺癌分子亚型的术前预测,诊断性能较高。
研究设计与样本
- 回顾性研究,纳入两家机构366例乳腺癌患者,分为训练集292例、验证集49例和测试集25例,以DCE-MRI为输入、分子亚型为预测目标。
诊断标准与关键数值
- MBCNN在中间融合与ROI 80像素并作外观变换时达最优性能。
- 测试集预测luminal B的AUC:MBCNN 0.8182 vs CNN 0.7208(p=0.44)vs CLSTM 0.7922(p=0.80)。
- 测试集预测HER2-enriched的AUC:MBCNN 0.8500 vs CNN 0.7300(p=0.36)vs CLSTM 0.8200(p=0.70)。
- 测试集预测TN的AUC:MBCNN 0.8900 vs CNN 0.7600(p=0.03)vs CLSTM 0.8300(p=0.63)。
- 测试集预测luminal A的AUC:MBCNN 0.8571 vs CNN 0.7619(p=0.08),与CLSTM相当;四亚型预测ACC:MBCNN 0.64 vs CNN 0.48 vs CLSTM 0.52。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: To evaluate the performance of deep learning (DL) in predicting different breast cancer molecular subtypes using DCE-MRI from two institutes.
MATERIALS AND METHODS: This retrospective study included 366 breast cancer patients from two institutes, divided into training (n = 292), validation (n = 49) and testing (n = 25) sets. We first transformed the public DCE-MRI appearance to ours to alleviate small-data-size and class-imbalance issues. Second, we developed a multi-branch convolutional-neural-network (MBCNN) to perform molecular subtype prediction. Third, we assessed the MBCNN with different regions of interest (ROIs) and fusion strategies, and compared it to previous DL models. Area under the curve (AUC) and accuracy (ACC) were used to assess different models. Delong-test was used for the comparison of different groups.
RESULTS: MBCNN achieved the optimal performance under intermediate fusion and ROI size of 80 pixels with appearance transformation. It outperformed CNN and convolutional long-short-term-memory (CLSTM) in predicting luminal B, HER2-enriched and TN subtypes, but without demonstrating statistical significance except against CNN in TN subtypes, with testing AUCs of 0.8182 vs. [0.7208, 0.7922] (p=0.44, 0.80), 0.8500 vs. [0.7300, 0.8200] (p=0.36, 0.70) and 0.8900 vs. [0.7600, 0.8300] (p=0.03, 0.63), respectively. When predicting luminal A, MBCNN outperformed CNN with AUCs of 0.8571 vs. 0.7619 (p=0.08) without achieving statistical significance, and is comparable to CLSTM. For four-subtype prediction, MBCNN achieved an ACC of 0.64, better than CNN and CLSTM models with ACCs of 0.48 and 0.52, respectively.
CONCLUSION: Developed DL model with the feature extraction and fusion of DCE-MRI from two institutes enabled preoperative prediction of breast cancer molecular subtypes with high diagnostic performance.
引用
- Xie X, Zhou H, Ma M, et al. A Deep Learning Model for Predicting Molecular Subtype of Breast Cancer by Fusing Multiple Sequences of DCE-MRI From Two Institutes. Acad Radiol. 2024. PMID: 38637240. doi:10.1016/j.acra.2024.03.002
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。