基于动态超声视频的深度学习乳腺病灶分类
研究采集888例患者的1000个乳腺病灶,每个病灶含2幅静态图像和2段动态视频,按7:2:1随机分为训练、验证和测试集,分别基于3D Resnet-50和2D Resnet-50构建DL-video与DL-image模型,并在测试集比较两模型与6名不同年资放射科医师的诊断性能。DL-video模型AUC为0.969,显著高于DL-image模型的0.925(P=0.0172)及6名医师的0.779-0.912(P<0.05);所有医师阅读动态视频的表现均优于静态图像,且年资越高表现越好。
要点
- DL-video模型AUC 0.969,显著高于DL-image模型0.925(P=0.0172)。
- DL-video模型AUC 0.969,显著高于6名放射科医师的0.779-0.912(P<0.05)。
- 所有放射科医师评估动态视频的表现均优于静态图像,且年资越高表现越好。
核心结论
- DL-video模型较传统DL-image模型和放射科医师能辨别更详细的空时信息,对乳腺病灶实现更准确分类,其临床应用可进一步改善乳腺癌诊断。
研究设计与样本
- 收集2020年5月至2021年12月888例患者的1000个乳腺病灶,每个病灶含2幅静态图像和2段动态视频,按7:2:1随机分为训练、验证和测试集;分别基于3D Resnet-50和2D Resnet-50用2000段动态视频和2000幅静态图像构建DL-video与DL-image模型,在测试集比较两模型与6名不同年资放射科医师的诊断性能。
诊断标准与关键数值
- DL-video模型AUC为0.969,DL-image模型AUC为0.925,两者差异P=0.0172。
- DL-video模型AUC 0.969 vs 6名放射科医师AUC 0.779-0.912,P<0.05。
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
PURPOSE: We intended to develop a deep-learning-based classification model based on breast ultrasound dynamic video, then evaluate its diagnostic performance in comparison with the classic model based on ultrasound static image and that of different radiologists.
METHOD: We collected 1000 breast lesions from 888 patients from May 2020 to December 2021. Each lesion contained two static images and two dynamic videos. We divided these lesions randomly into training, validation, and test sets by the ratio of 7:2:1. Two deep learning (DL) models, namely DL-video and DL-image, were developed based on 3D Resnet-50 and 2D Resnet-50 using 2000 dynamic videos and 2000 static images, respectively. Lesions in the test set were evaluated to compare the diagnostic performance of two models and six radiologists with different seniority.
RESULTS: The area under the curve of the DL-video model was significantly higher than those of the DL-image model (0.969 vs. 0.925, P = 0.0172) and six radiologists (0.969 vs. 0.779-0.912, P < 0.05). All radiologists performed better when evaluating the dynamic videos compared to the static images. Furthermore, radiologists performed better with increased seniority both in reading images and videos.
CONCLUSIONS: The DL-video model can discern more detailed spatial and temporal information for accurate classification of breast lesions than the conventional DL-image model and radiologists, and its clinical application can further improve the diagnosis of breast cancer.
引用
- Zhao, Kong, Xu, et al. Deep learning-based classification of breast lesions using dynamic ultrasound video. Eur J Radiol. 2023. PMID: 37290361. doi:10.1016/j.ejrad.2023.110885
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。