文献速递乳腺技术与安全US

深度学习降低超声BI-RADS 4A乳腺病变恶性率

Application of Deep Learning to Reduce the Rate of Malignancy Among BI-RADS 4A Breast Lesions Based on Ultrasonography.
PubMed 文献 · Ultrasound Med Biol · 2022-08-30

本研究基于479个术前超声诊断为BI-RADS 4A的乳腺病变(恶性率24.4%),开发了四种深度学习模型以区分良恶性。MobileNet模型表现最优,在测试集中AUROC为0.897、准确率0.913、敏感度0.926、特异度0.899。借助该模型,预计约14.4%的患者可被升级为BI-RADS 4B,有助于降低4A病变中的恶性率。

要点

  • MobileNet模型在测试集中AUROC达0.897,敏感度0.926,特异度0.899。
  • 模型辅助下约14.4%的BI-RADS 4A病变可升级为4B,可能降低4A恶性率。
  • 研究基于术前超声图像,经术后病理证实良恶性,训练集与测试集按4:1划分。

核心结论

  • 基于超声图像的MobileNet深度学习模型能有效区分BI-RADS 4A乳腺病变的良恶性,有助于降低该类别中的恶性率。

研究设计与样本

  • 回顾性研究纳入479个术前超声诊断为BI-RADS 4A的乳腺病变,其中良性362个、恶性117个,恶性率24.4%,均经术后病理证实。
  • 超声图像随机按4:1分为训练集和测试集,开发了MobileNet、DenseNet121、Xception和Inception V3四种深度学习模型。
  • 采用AUROC、准确率、敏感度、特异度、PPV和NPV比较模型性能,并通过五折交叉验证评估稳健性。

诊断标准与关键数值

  • MobileNet模型在测试集中分类BI-RADS 4A病变良恶性的AUROC为0.897,准确率为0.913,敏感度为0.926,特异度为0.899,PPV为0.958,NPV为0.784。
  • 借助MobileNet模型,约14.4%的患者预期在超声中被升级为BI-RADS 4B。

临床应用与阅片要点

  • 对于术前超声诊断为BI-RADS 4A的乳腺病变,可参考MobileNet模型输出辅助判断,模型预测恶性概率高的病变可考虑升级为4B以提示进一步处理。
  • 模型具有较高的敏感度和PPV,但NPV为0.784,阴性预测价值有限,阅片时仍需结合常规超声特征综合评估。

局限性

  • 研究为单中心回顾性设计,样本量有限(479个病变),外部泛化能力需进一步验证。
  • 原文未报告模型在外部验证集或前瞻性队列中的性能。

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

The aim of the work described here was to develop an ultrasound (US) image-based deep learning model to reduce the rate of malignancy among breast lesions diagnosed as category 4A of the Breast Imaging-Reporting and Data System (BI-RADS) during the pre-operative US examination. A total of 479 breast lesions diagnosed as BI-RADS 4A in pre-operative US examination were enrolled. There were 362 benign lesions and 117 malignant lesions confirmed by postoperative pathology with a malignancy rate of 24.4%. US images were collected from the database server. They were then randomly divided into training and testing cohorts at a ratio of 4:1. To correctly classify malignant and benign tumors diagnosed as BI-RADS 4A in US, four deep learning models, including MobileNet, DenseNet121, Xception and Inception V3, were developed. The performance of deep learning models was compared using the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Meanwhile, the robustness of the models was evaluated by five-fold cross-validation. Among the four models, the MobileNet model turned to be the optimal model with the best performance in classifying benign and malignant lesions among BI-RADS 4A breast lesions. The AUROC, accuracy, sensitivity, specificity, PPV and NPV of the optimal model in the testing cohort were 0.897, 0.913, 0.926, 0.899, 0.958 and 0.784, respectively. About 14.4% of patients were expected to be upgraded to BI-RADS 4B in US with the assistance of the MobileNet model. The deep learning model MobileNet can help to reduce the rate of malignancy among BI-RADS 4A breast lesions in pre-operative US examinations, which is valuable to clinicians in tailoring treatment for suspicious breast lesions identified on US.

来源:Ultrasound Med Biol,查看原文。版权归原作者所有。

引用

  1. Zhao, Hou, Li, et al. Application of Deep Learning to Reduce the Rate of Malignancy Among BI-RADS 4A Breast Lesions Based on Ultrasonography. Ultrasound Med Biol. 2022. PMID: 36055860. doi:10.1016/j.ultrasmedbio.2022.06.019

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。