文献速递胸部CT

基于自动深度学习的CT肺肿瘤检测与分割

Automated Deep Learning-Based Detection and Segmentation of Lung Tumors at CT.
PubMed 文献 · Radiology · 2025-01-01

这项回顾性研究利用1504例放疗计划的CT模拟扫描和临床肺肿瘤分割数据,训练基于3D U-Net的图像多分辨率集成模型,在内部和两个附属医疗中心的外部测试集上评估。模型在150例CT测试集上检测肺肿瘤的敏感度为92%(92/100)、特异度为82%(41/50);100例单发肺肿瘤的中位模型-医师DSC为0.77(IQR 0.65-0.83),医师间DSC为0.80(IQR 0.72-0.86);模型分割时间短于医师(均值76.6秒 vs 166.1-187.7秒,p<0.001)。

要点

  • 模型在150例CT测试集检测肺肿瘤敏感度92%(92/100)、特异度82%(41/50)。
  • 单发肺肿瘤模型-医师中位DSC 0.77(IQR 0.65-0.83),医师间0.80(IQR 0.72-0.86)。
  • 模型分割时间均值76.6秒,短于医师的166.1-187.7秒(p<0.001)。

核心结论

  • 常规收集的放疗数据可用于模型训练,模型关键优势包括平衡体积上下文与分辨率的3D U-Net集成方法、稳健的肿瘤检测与分割性能以及泛化至外部中心的能力。

研究设计与样本

  • 回顾性研究,2019年7月至2024年11月间使用放疗计划中的CT模拟扫描和临床肺肿瘤分割大数据集训练基于3D U-Net的图像多分辨率集成模型。
  • 训练集为1504例CT扫描及临床肺肿瘤分割;测试集为150例CT模拟扫描及两个附属医疗中心的肺肿瘤分割,包括单发原发性和转移性肺肿瘤。
  • 采用敏感度、特异度、假阳性率和Dice相似系数(DSC)评估性能,模型预测肿瘤体积与医师勾画体积比较,组间比较用Wilcoxon符号秩检验或单因素方差分析,P<0.05为差异有统计学意义。

关键影像表现/征象

  • 模型在合并150例CT测试集上检测肺肿瘤的敏感度为92%(92/100),特异度为82%(41/50)。
  • 100例单发肺肿瘤CT中,模型-医师中位DSC为0.77(IQR 0.65-0.83),医师间中位DSC为0.80(IQR 0.72-0.86)。
  • 模型分割时间均值76.6秒,医师分割时间均值166.1-187.7秒,模型短于医师(p<0.001)。

诊断标准与关键数值

  • 检测性能以敏感度92%(92/100)、特异度82%(41/50)报告,对象为合并150例CT测试集中的肺肿瘤检测。
  • 分割性能以DSC报告,100例单发肺肿瘤中模型-医师中位DSC 0.77(IQR 0.65-0.83),医师间中位DSC 0.80(IQR 0.72-0.86)。

临床应用与阅片要点

  • 模型可自动识别和分割CT上的肺肿瘤,输出结果与医师勾画体积进行比较,分割时间短于医师手动勾画。
  • 模型在内部和外部测试集上评估,涵盖单发原发性和转移性肺肿瘤,原文报告模型能泛化至外部中心。

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

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Background Detection and segmentation of lung tumors on CT scans are critical for monitoring cancer progression, evaluating treatment responses, and planning radiation therapy; however, manual delineation is labor-intensive and subject to physician variability. Purpose To develop and evaluate an ensemble deep learning model for automating identification and segmentation of lung tumors on CT scans. Materials and Methods A retrospective study was conducted between July 2019 and November 2024 using a large dataset of CT simulation scans and clinical lung tumor segmentations from radiotherapy plans. This dataset was used to train a 3D U-Net-based, image-multiresolution ensemble model to detect and segment lung tumors on CT scans. Model performance was evaluated on internal and external test sets composed of CT simulation scans and lung tumor segmentations from two affiliated medical centers, including single primary and metastatic lung tumors. Performance metrics included sensitivity, specificity, false positive rate, and Dice similarity coefficient (DSC). Model-predicted tumor volumes were compared with physician-delineated volumes. Group comparisons were made with Wilcoxon signed-rank test or one-way ANOVA. P < 0.05 indicated statistical significance. Results The model, trained on 1,504 CT scans with clinical lung tumor segmentations, achieved 92% sensitivity (92/100) and 82% specificity (41/50) in detecting lung tumors on the combined 150-CT scan test set. For a subset of 100 CT scans with a single lung tumor each, the model achieved a median model-physician DSC of 0.77 (IQR: 0.65-0.83) and an interphysician DSC of 0.80 (IQR: 0.72-0.86). Segmentation time was shorter for the model than for physicians (mean 76.6 vs. 166.1-187.7 seconds; p<0.001). Conclusion Routinely collected radiotherapy data were useful for model training. The key strengths of the model include a 3D U-Net ensemble approach for balancing volumetric context with resolution, robust tumor detection and segmentation performance, and the ability to generalize to an external site.

来源:Radiology,查看原文。版权归原作者所有。

引用

  1. Kashyap M, Wang X, Panjwani N, et al. Automated Deep Learning-Based Detection and Segmentation of Lung Tumors at CT. Radiology. 2025. PMID: 39835976. doi:10.1148/radiol.233029

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。