文献速递骨肌技术与安全US

通过半监督学习、焦点损失与帧间关系增强超声图像中的多目标检测

Enhancing Multi-Object Detection in Ultrasound Images Through Semi-Supervised Learning, Focal Loss and Relation of Frame.
PubMed 文献 · Ultrasound Med Biol · 2024-09-21

本研究设计基于深度学习的自动标注系统用于实时识别肌肉骨骼解剖结构,提出半监督学习(SSL)减少标注时间,采用焦点损失(FL)提升困难结构识别精度,并在推理阶段利用视频帧时间连续性避免检测不连续。结果显示仅用30%训练数据即可达到与监督学习相当的性能,帧间关系可减少推理时漏检帧数并提高检测置信度。

要点

  • 半监督学习(SSL)联合焦点损失(FL)仅用30%训练数据即可达到与监督学习相当的性能
  • 推理阶段利用视频帧时间连续性可减少漏检帧数并提高检测置信度
  • 该系统旨在辅助医疗专业人员实时识别肌肉骨骼解剖结构

核心结论

  • 所提出的系统有潜力辅助医疗专业人员高效、有效地诊断肌肉骨骼疾病,最终改善患者预后。

研究设计与样本

  • 研究设计为基于深度学习神经网络的自动标注系统开发与验证,旨在辅助实时识别解剖结构。
  • 训练采用半监督学习(SSL)与焦点损失(FL)相结合的方法,并与监督学习进行性能比较。
  • 推理阶段利用视频帧的时间连续性,即利用先前帧的信息辅助当前图像的結構识别。

关键影像表现/征象

  • 前臂组织检测通过合理配置半监督学习(SSL)方法(包括焦点损失(FL)和过滤阈值)得到验证。
  • 仅使用30%的训练数据即达到与监督学习相当的性能。

诊断标准与关键数值

  • 原文未报告敏感度、特异度、AUC、阈值或分级标准等诊断准确性指标。

临床应用与阅片要点

  • 原文未据此给出推荐。

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: To identify musculoskeletal anatomical structures in real time by using deep learning techniques.

METHODS: An automated annotation system based on deep learning neural networks was designed to aid in the real-time identification of anatomical structures. Additionally, novel algorithms aimed at diminishing model training duration while enhancing accuracy were introduced. In this study, we proposed a semi-supervised learning (SSL) approach that substantially reduced annotation time. We also adopted the focal loss (FL) method to enhance the accuracy of challenging structures. Additionally, during the inference stage, we harnessed the temporal continuity of video frames, which involved leveraging information from preceding frames to facilitate recognition of structures in the current image. Training the model through a combination of SSL and FL yielded superior performance compared with supervised learning, while also substantially mitigating any expense linked to annotations. During inference, the incorporation of frame continuity helped to avoid discontinuity and bolster accuracy.

RESULTS: Forearm tissue detection was demonstrated by properly configuring the SSL approach, including FL and the filtering threshold. Comparable performance with supervised learning was achieved while only using 30% of the training data. The real-time experimental results also demonstrated that implementing relation of frame reduced the number of missing frames during inference and successfully increased the confidence scores of detected objects.

CONCLUSION: This proposed system has the potential to aid medical professionals in efficiently and effectively diagnosing musculoskeletal disorders, ultimately leading to enhanced patient outcomes.

来源:Ultrasound Med Biol,查看原文。版权归原作者所有。

引用

  1. Chu HY, Wu CH, Chen PX, et al. Enhancing Multi-Object Detection in Ultrasound Images Through Semi-Supervised Learning, Focal Loss and Relation of Frame. Ultrasound Med Biol. 2024. PMID: 39307679. doi:10.1016/j.ultrasmedbio.2024.08.012

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。