文献速递胸部CT

基于深度学习的肺结节检测系统在中国肺癌筛查项目中作为替代阅片者的性能

Performance of a deep learning-based lung nodule detection system as an alternative reader in a Chinese lung cancer screening program.
PubMed 文献 · Eur J Radiol · 2021-11-24

在中国低剂量CT肺癌筛查项目中,将180例基线筛查有肺结节的受试者与无结节受试者按1:1随机混合共360例,所有扫描经双读片后由学术型深度学习计算机辅助检测(DL-CAD)系统处理,并由两位高年资放射科医师确定参考标准。DL-CAD检测肺结节的敏感度为90.1%、每扫描1.0个假阳性,双读片敏感度为76.0%、每扫描0.04个假阳性(P=0.001),DL-CAD对Lung-RADS 2结节的敏感度更高。

要点

  • DL-CAD检测任意大小和类型肺结节的敏感度为90.1%,每扫描1.0个假阳性。
  • 双读片检测肺结节的敏感度为76.0%,每扫描0.04个假阳性。
  • DL-CAD对4–6 mm结节的敏感度为86.3%,高于双读片的58.9%(P=0.001)。

核心结论

  • DL-CAD系统能在低剂量CT上准确检测肺结节,假阳性率为每扫描1个结节,检测性能高于双读片,可能辅助放射科医师在低剂量CT肺癌筛查中进行结节检测。

研究设计与样本

  • 研究将180例基线低剂量CT肺癌筛查扫描有肺结节的个体与无结节的筛查者按1:1随机混合(共360例),所有扫描经双读片评估后由学术型DL-CAD系统处理,双读片与DL-CAD的发现再由两位高年资放射科医师评估以确定参考标准。

诊断标准与关键数值

  • 参考标准包含196例个体的262个≥4 mm结节,359个发现被判定为假阳性。
  • DL-CAD检测任意大小或类型肺结节的敏感度为90.1%,每扫描1.0个假阳性;双读片敏感度为76.0%,每扫描0.04个假阳性(P=0.001)。
  • DL-CAD检测≥4–≤6 mm结节的敏感度为86.3%,双读片为58.9%(P=0.001)。
  • DL-CAD与双读片检测Lung-RADS 3结节的敏感度分别为94.3%与90.0%(P=0.549),检测Lung-RADS 4结节分别为100.0%与97.0%(P=1.000)。
  • DL-CAD检测Lung-RADS 2结节的敏感度为86.2%,双读片为65.4%(P<0.001)。

临床应用与阅片要点

  • DL-CAD系统可能辅助放射科医师在低剂量CT肺癌筛查中进行结节检测。
  • 63个结节仅由DL-CAD系统识别,27个结节仅由双读片发现。

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 evaluate the performance of a deep learning-based computer-aided detection (DL-CAD) system in a Chinese low-dose CT (LDCT) lung cancer screening program.

MATERIALS AND METHODS: One-hundred-and-eighty individuals with a lung nodule on their baseline LDCT lung cancer screening scan were randomly mixed with screenees without nodules in a 1:1 ratio (total: 360 individuals). All scans were assessed by double reading and subsequently processed by an academic DL-CAD system. The findings of double reading and the DL-CAD system were then evaluated by two senior radiologists to derive the reference standard. The detection performance was evaluated by the Free Response Operating Characteristic curve, sensitivity and false-positive (FP) rate. The senior radiologists categorized nodules according to nodule diameter, type (solid, part-solid, non-solid) and Lung-RADS.

RESULTS: The reference standard consisted of 262 nodules ≥ 4 mm in 196 individuals; 359 findings were considered false positives. The DL-CAD system achieved a sensitivity of 90.1% with 1.0 FP/scan for detection of lung nodules regardless of size or type, whereas double reading had a sensitivity of 76.0% with 0.04 FP/scan (P = 0.001). The sensitivity for detection of nodules ≥ 4 - ≤ 6 mm was significantly higher with DL-CAD than with double reading (86.3% vs. 58.9% respectively; P = 0.001). Sixty-three nodules were only identified by the DL-CAD system, and 27 nodules only found by double reading. The DL-CAD system reached similar performance compared to double reading in Lung-RADS 3 (94.3% vs. 90.0%, P = 0.549) and Lung-RADS 4 nodules (100.0% vs. 97.0%, P = 1.000), but showed a higher sensitivity in Lung-RADS 2 (86.2% vs. 65.4%, P < 0.001).

CONCLUSIONS: The DL-CAD system can accurately detect pulmonary nodules on LDCT, with an acceptable false-positive rate of 1 nodule per scan and has higher detection performance than double reading. This DL-CAD system may assist radiologists in nodule detection in LDCT lung cancer screening.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Cui, Zheng, Heuvelmans, et al. Performance of a deep learning-based lung nodule detection system as an alternative reader in a Chinese lung cancer screening program. Eur J Radiol. 2021. PMID: 34871936. doi:10.1016/j.ejrad.2021.110068

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。