文献速递技术与安全US

深度学习自动检测超声探头均匀性伪影:基于临床中位图像的研究

Detecting Uniformity Artifacts in Ultrasound Transducers: Insights from Clinical Median Images and Deep Learning for Automatic Detection.
PubMed 文献 · J Ultrasound Med · 2026-04-01

本研究提出基于临床中位图像的深度学习框架,用于检测线阵和凸阵探头的均匀性伪影,以补充传统质量控制。模型在首个测试集准确率达100%,在独立体模验证集上准确率87.4%、敏感度0.84、特异度0.88。该框架可自动标记含伪影的中位图像,约减少80%人工审核时间。

要点

  • 临床中位图像可聚合探头层面的均匀性伪影,适合深度学习自动质控。
  • 模型在独立体模验证集准确率87.4%,敏感度0.84、特异度0.88。
  • 自动标记伪影图像可减少约80%人工审核时间,适合融入临床质控流程。

核心结论

  • 基于临床中位图像的深度学习框架能稳健检测线阵和凸阵探头的均匀性伪影,并自动标记以减少人工审核时间。

研究设计与样本

  • 研究为深度学习模型开发与验证,使用预训练ResNet-18在线阵和凸阵探头的临床中位图像上微调,训练集、验证集和测试集按探头不重叠划分。
  • 模型另在396幅经体模验证的线阵和凸阵图像组成的独立测试集上评估泛化能力。

关键影像表现/征象

  • 遮挡敏感图显示模型关注区域与均匀性伪影区域一致。

诊断标准与关键数值

  • 在首个数据集测试集上准确率为100%。
  • 在独立测试集上准确率87.4%,敏感度0.84,特异度0.88。

临床应用与阅片要点

  • 可将模型集成到临床质控流程,自动标记分类为存在伪影的中位图像,优先人工复核。
  • 模型关注区域与伪影区域一致,提示其判断依据具有可解释性。

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

OBJECTIVES: Uniformity artifacts caused by defective transducer elements or scanner malfunctions degrade diagnostic image quality. Traditional quality control (QC) methods, such as phantom testing and visual or quantitative image analysis, can be labor-intensive and limited in test frequency. This study aims to develop a deep learning framework to detect uniformity artifacts and complement traditional QC.

METHODS: Clinical median images were generated by aggregating co-registered grayscale ultrasound images acquired by each transducer and computing median intensity values across the image stack. A pretrained ResNet-18 model was fine-tuned on a dataset consisting of clinical median images from linear and curvilinear transducers. The dataset was divided into training, validation, and testing subsets, ensuring no overlap between training and test transducers. To assess generalizability, the model was also evaluated on an independent test set of 396 phantom-validated images from linear and curvilinear transducers.

RESULTS: The model achieved 100% accuracy on the first dataset's test set. On the independent test set, it attained 87.4% accuracy with high sensitivity (0.84) and specificity (0.88), demonstrating robust generalization. Occlusion sensitivity maps confirmed the model's attention to uniformity artifact regions.

CONCLUSION: The deep learning framework using clinical median images demonstrated robust performance across several linear and curvilinear transducer models. It could be integrated into the clinical QC workflow by automating artifact detection in an effective and timely manner. In our practice, it can flag median images classified as artifact-present, reducing human review time by approximately 80% while preserving detection accuracy.

来源:J Ultrasound Med,查看原文。版权归原作者所有。

引用

  1. Gu C, Brom K, Stekel S, et al. Detecting Uniformity Artifacts in Ultrasound Transducers: Insights from Clinical Median Images and Deep Learning for Automatic Detection. J Ultrasound Med. 2026. PMID: 41919451. doi:10.1002/jum.70239

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。