文献速递外周血管技术与安全US

基于多序列注意力卷积神经网络的血管超声自发显影自动检测

Automated Spontaneous Echo Contrast Detection Using a Multisequence Attention Convolutional Neural Network.
PubMed 文献 · Ultrasound Med Biol · 2024-03-08

研究纳入201例患者沿股静脉的801份存档超声采集数据,构建以ResNetv2为骨干的多序列卷积神经网络并用软注意力可视化关键帧,按80/20划分训练与测试评估自发显影检测性能;带软注意力时AUC 0.74、敏感度0.73、特异度0.68,不带软注意力时AUC 0.69、敏感度0.71、特异度0.60。

要点

  • 多序列CNN联合软注意力检测股静脉自发显影,AUC 0.74、敏感度0.73、特异度0.68。
  • 去掉软注意力后AUC降至0.69、敏感度0.71、特异度0.60。
  • 模型对含更多血管腔的超声帧赋予更高注意力分数。

核心结论

  • 多序列CNN模型可从超声关键帧中识别自发显影的存在,AUC为0.74,可能支持筛查应用并发现更多自发显影数据。
  • 该模型无需专家判定或额外临床医生报告时间,而这两者是目前自发显影应用的重要障碍。

研究设计与样本

  • 研究为模型开发与评估,数据集为来自201例患者的801份沿股静脉的存档超声采集。
  • 采用以ResNetv2为骨干的多序列卷积神经网络,并用软注意力可视化关键帧重要性。
  • 按80/20划分训练集与测试集评估自发显影预测性能,报告ROC-AUC及Youden阈值对应的敏感度、特异度、F1分数、真阴性、假阴性、假阳性与真阳性。
  • 模型可回顾性应用,不偏离常规临床实践。

关键影像表现/征象

  • 模型对含更多血管腔的超声帧赋予更高的注意力分数,并提供注意力可视化。

诊断标准与关键数值

  • 带软注意力:检测自发显影的AUC 0.74、敏感度0.73、特异度0.68。
  • 不带软注意力:AUC 0.69、敏感度0.71、特异度0.60。
  • 性能评估采用80/20训练/测试划分,敏感度与特异度对应Youden阈值。

临床应用与阅片要点

  • 模型可回顾性应用于存档超声数据,无需偏离常规临床实践。
  • 模型无需专家判定或额外临床医生报告时间。
  • 未来工作可扫描存档数据,将自发显影发现与已记录的临床结局进行机会性关联。
  • 模型与处理后的数据集已在 https://github.com/Ouwen/automatic-spontaneous-echo-contrast 公开。

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

OBJECTIVE: Spontaneous echo contrast (SEC) is a vascular ultrasound finding associated with increased thromboembolism risk. However, identification requires expert determination and clinician time to report. We developed a deep learning model that can automatically identify SEC. Our model can be applied retrospectively without deviating from routine clinical practice. The retrospective nature of our model means future works could scan archival data to opportunistically correlate SEC findings with documented clinical outcomes.

METHODS: We curated a data set of 801 archival acquisitions along the femoral vein from 201 patients. We used a multisequence convolutional neural network (CNN) with ResNetv2 backbone and visualized keyframe importance using soft attention. We evaluated SEC prediction performance using an 80/20 train/test split. We report receiver operating characteristic area under the curve (ROC-AUC), along with the Youden threshold-associated sensitivity, specificity, F1 score, true negative, false negative, false positive and true positive.

RESULTS: Using soft attention, we can identify SEC with an AUC of 0.74, sensitivity of 0.73 and specificity of 0.68. Without soft attention, our model achieves an AUC of 0.69, sensitivity of 0.71 and specificity of 0.60. Additionally, we provide attention visualizations and note that our model assigns higher attention score to ultrasound frames containing more vessel lumen.

CONCLUSION: Our multisequence CNN model can identify the presence of SEC from ultrasound keyframes with an AUC of 0.74, which could enable screening applications and enable more SEC data discovery. The model does not require the expert intervention or additional clinician reporting time that are currently significant barriers to SEC adoption. Model and processed data sets are publicly available at https://github.com/Ouwen/automatic-spontaneous-echo-contrast.

来源:Ultrasound Med Biol,查看原文。版权归原作者所有。

引用

  1. Huang O, Shi Z, Garg N, et al. Automated Spontaneous Echo Contrast Detection Using a Multisequence Attention Convolutional Neural Network. Ultrasound Med Biol. 2024. PMID: 38461036. doi:10.1016/j.ultrasmedbio.2024.01.016

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。