文献速递胸部技术与安全DR

市售胸片人工智能工具检测气腔病变、气胸和胸腔积液的诊断准确性

Commercially Available Chest Radiograph AI Tools for Detecting Airspace Disease, Pneumothorax, and Pleural Effusion.
PubMed 文献 · Radiology · 2023-01-01

这项回顾性研究纳入丹麦四家医院2020年1月连续接受胸片检查的2040例成人患者,由两名(分歧时三名)胸部放射科医师参考全部既往和后续影像独立标注参考标准,评估四种市售人工智能工具检测气腔病变、气胸和胸腔积液的诊断准确性,发现其对三种目标病变的AUC为0.83-0.97、敏感度为62%-95%,阴性预测值为92%-100%,但假阳性多于放射科报告,且对较小病变和多发病变的表现下降。

要点

  • 四种市售AI工具检测气腔病变、气胸和胸腔积液的AUC分别为0.83-0.88、0.89-0.97和0.94-0.97,阴性预测值达92%-100%。
  • 正常或单一表现的胸片上AI特异度较高(气腔病变85%-96%、气胸99%-100%、胸腔积液95%-100%),四种及以上表现时明显降低。
  • AI对模糊气腔病变和较小气胸或胸腔积液的敏感度低于较大病变,且假阳性发现多于放射科报告。

核心结论

  • 当前代AI工具检测胸片上气腔病变、气胸和胸腔积液具有中等至高度敏感度,但假阳性发现多于放射科报告,且对较小目标病变和存在多种病变时表现下降。

研究设计与样本

  • 回顾性研究纳入2020年1月丹麦四家医院连续接受胸片检查的成人患者2040例(中位年龄72岁,IQR 58-81岁;1033例女性),由两名胸部放射科医师(分歧时三名)参考全部既往和后续影像独立标注参考标准,计算AUC、敏感度和特异度,并按病变严重程度、胸片表现数量和投照体位分层,采用χ2检验和McNemar检验比较。

诊断标准与关键数值

  • 四种市售AI工具检测气腔病变的AUC为0.83-0.88、敏感度72%-91%,检测气胸的AUC为0.89-0.97、敏感度63%-90%,检测胸腔积液的AUC为0.94-0.97、敏感度62%-95%;三种目标病变的阴性预测值均为92%-100%。

临床应用与阅片要点

  • 阅片和报告时需注意AI对模糊气腔病变(敏感度33%-61%)及较小气胸或胸腔积液(敏感度9%-94%)的敏感度低于较大病变(81%-100%),且对含四种及以上表现的胸片特异度明显降低(气腔病变27%-69%、气胸96%-99%、胸腔积液65%-92%),AI假阳性发现多于放射科报告。

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

Background Commercially available artificial intelligence (AI) tools can assist radiologists in interpreting chest radiographs, but their real-life diagnostic accuracy remains unclear. Purpose To evaluate the diagnostic accuracy of four commercially available AI tools for detection of airspace disease, pneumothorax, and pleural effusion on chest radiographs. Materials and Methods This retrospective study included consecutive adult patients who underwent chest radiography at one of four Danish hospitals in January 2020. Two thoracic radiologists (or three, in cases of disagreement) who had access to all previous and future imaging labeled chest radiographs independently for the reference standard. Area under the receiver operating characteristic curve, sensitivity, and specificity were calculated. Sensitivity and specificity were additionally stratified according to the severity of findings, number of findings on chest radiographs, and radiographic projection. The χ2 and McNemar tests were used for comparisons. Results The data set comprised 2040 patients (median age, 72 years [IQR, 58-81 years]; 1033 female), of whom 669 (32.8%) had target findings. The AI tools demonstrated areas under the receiver operating characteristic curve ranging 0.83-0.88 for airspace disease, 0.89-0.97 for pneumothorax, and 0.94-0.97 for pleural effusion. Sensitivities ranged 72%-91% for airspace disease, 63%-90% for pneumothorax, and 62%-95% for pleural effusion. Negative predictive values ranged 92%-100% for all target findings. In airspace disease, pneumothorax, and pleural effusion, specificity was high for chest radiographs with normal or single findings (range, 85%-96%, 99%-100%, and 95%-100%, respectively) and markedly lower for chest radiographs with four or more findings (range, 27%-69%, 96%-99%, 65%-92%, respectively) (P < .001). AI sensitivity was lower for vague airspace disease (range, 33%-61%) and small pneumothorax or pleural effusion (range, 9%-94%) compared with larger findings (range, 81%-100%; P value range, > .99 to < .001). Conclusion Current-generation AI tools showed moderate to high sensitivity for detecting airspace disease, pneumothorax, and pleural effusion on chest radiographs. However, they produced more false-positive findings than radiology reports, and their performance decreased for smaller-sized target findings and when multiple findings were present. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Yanagawa and Tomiyama in this issue.

来源:Radiology,查看原文。版权归原作者所有。

引用

  1. Lind Plesner L, Müller FC, Brejnebøl MW, et al. Commercially Available Chest Radiograph AI Tools for Detecting Airspace Disease, Pneumothorax, and Pleural Effusion. Radiology. 2023. PMID: 37750768. doi:10.1148/radiol.231236

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。