ResNet与视觉Transformer融合模型用于预测HIFU治疗子宫肌瘤的疗效
本研究纳入中心A(训练集272例、内部验证集92例)和中心B(外部测试集125例)接受HIFU治疗的子宫肌瘤患者,利用术前T2加权MRI构建Res-ViT融合模型预测术后即刻无灌注体积比(NPVR)≥80%,Res-ViT在内部测试集AUC为0.895(95%CI 0.857-0.987)、外部测试集AUC为0.853(95%CI 0.776-0.921),优于独立影像组学、ResNet-18和ViT模型。
要点
- Res-ViT融合局部纹理与全局空间特征,用于预测HIFU治疗子宫肌瘤后NPVR≥80%。
- 外部测试集AUC为0.853(95%CI 0.776-0.921),内部测试集AUC为0.895(95%CI 0.857-0.987)。
- SHAP分析显示ResNet分支为主要决策成分,特征贡献为55.4%。
核心结论
- Res-ViT模型表明局部与全局特征融合策略可有效量化子宫肌瘤异质性,显著提升HIFU疗效预测准确性。
研究设计与样本
- 回顾性多中心研究,训练集来自中心A(N=272)、内部验证集来自中心A(N=92)、外部测试集来自中心B(N=125),基于术前T2加权MRI构建Res-ViT模型,参考标准为术后即刻NPVR≥80%,并与独立影像组学、ResNet-18和ViT模型比较。
诊断标准与关键数值
- 以术后即刻无灌注体积比(NPVR)≥80%为预测目标;Res-ViT内部测试集AUC=0.895(95%CI 0.857-0.987),外部测试集AUC=0.853(95%CI 0.776-0.921)。
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: High-intensity focused ultrasound (HIFU) is a non-invasive technique for treating uterine fibroids, and the accurate prediction of its therapeutic efficacy depends on precise quantification of the intratumoral heterogeneity. However, existing methods still have limitations in characterizing intratumoral heterogeneity, which restricts the accuracy of efficacy prediction. To this end, this study proposes a deep learning model with a parallel architecture of ResNet and ViT (Res-ViT) to verify whether the synergistic characterization of local texture and global spatial features can improve the accuracy of HIFU efficacy prediction.
MATERIALS AND METHODS: This study enrolled patients with uterine fibroids who underwent HIFU treatment from Center A (training set: N = 272; internal validation set: N = 92) and Center B (external test set: N = 125). Preoperative T2-weighted magnetic resonance images were used to develop the Res-ViT model for predicting immediate post-treatment non-perfused volume ratio (NPVR) ≥ 80%. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and compared against independent Radiomics, ResNet-18, and ViT models.
RESULTS: The Res-ViT model outperformed all standalone models across both internal (AUC = 0.895, 95% CI: 0.857-0.987) and external (AUC = 0.853, 95% CI: 0.776-0.921) test sets. SHAP analysis identified the ResNet branch as the predominant decision-making component (feature contribution: 55.4%). The visualization of Gradient-weighted Class Activation Mapping (Grad-CAM) shows that the key regions attended by Res-ViT have higher spatial overlap with the postoperative non-ablated fibroid tissue.
CONCLUSION: The proposed Res-ViT model demonstrates that the fusion strategy of local and global features is an effective method for quantifying uterine fibroid heterogeneity, significantly enhancing the accuracy of HIFU efficacy prediction.
引用
- Zhou Y, Xu H, Jiang W, et al. A Fusion Model of ResNet and Vision Transformer for Efficacy Prediction of HIFU Treatment of Uterine Fibroids. Acad Radiol. 2025. PMID: 40935773. doi:10.1016/j.acra.2025.08.054
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。