Mallio Carlo Augusto, Napolitano Andrea, Castiello Gennaro, Giordano Francesco Maria, D'Alessio Pasquale, Iozzino Mario, Sun Yipeng, Angeletti Silvia, Russano Marco, Santini Daniele, Tonini Giuseppe, Zobel Bruno Beomonte, Vincenzi Bruno, Quattrocchi Carlo Cosimo
Departmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, Italy.
Departmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, Italy.
Cancers (Basel). 2021 Feb 6;13(4):652. doi: 10.3390/cancers13040652.
Coronavirus disease 2019 (COVID-19) pneumonia and immune checkpoint inhibitor (ICI) therapy-related pneumonitis share common features. The aim of this study was to determine on chest computed tomography (CT) images whether a deep convolutional neural network algorithm is able to solve the challenge of differential diagnosis between COVID-19 pneumonia and ICI therapy-related pneumonitis.
We enrolled three groups: a pneumonia-free group ( = 30), a COVID-19 group ( = 34), and a group of patients with ICI therapy-related pneumonitis ( = 21). Computed tomography images were analyzed with an artificial intelligence (AI) algorithm based on a deep convolutional neural network structure. Statistical analysis included the Mann-Whitney U test (significance threshold at < 0.05) and the receiver operating characteristic curve (ROC curve).
The algorithm showed low specificity in distinguishing COVID-19 from ICI therapy-related pneumonitis (sensitivity 97.1%, specificity 14.3%, area under the curve (AUC) = 0.62). ICI therapy-related pneumonitis was identified by the AI when compared to pneumonia-free controls (sensitivity = 85.7%, specificity 100%, AUC = 0.97).
The deep learning algorithm is not able to distinguish between COVID-19 pneumonia and ICI therapy-related pneumonitis. Awareness must be increased among clinicians about imaging similarities between COVID-19 and ICI therapy-related pneumonitis. ICI therapy-related pneumonitis can be applied as a challenge population for cross-validation to test the robustness of AI models used to analyze interstitial pneumonias of variable etiology.
2019冠状病毒病（COVID-19）肺炎与免疫检查点抑制剂（ICI）治疗相关肺炎具有共同特征。本研究的目的是在胸部计算机断层扫描（CT）图像上确定深度卷积神经网络算法是否能够解决COVID-19肺炎与ICI治疗相关肺炎的鉴别诊断难题。
我们纳入了三组：无肺炎组（n = 30）、COVID-19组（n = 34）和ICI治疗相关肺炎患者组（n = 21）。使用基于深度卷积神经网络结构的人工智能（AI）算法对计算机断层扫描图像进行分析。统计分析包括曼-惠特尼U检验（显著性阈值为P < 0.05）和受试者操作特征曲线（ROC曲线）。
该算法在区分COVID-19与ICI治疗相关肺炎方面表现出较低的特异性（敏感性97.1%，特异性14.3%，曲线下面积（AUC） = 0.62）。与无肺炎对照组相比，AI识别出了ICI治疗相关肺炎（敏感性 = 85.7%，特异性100%，AUC = 0.97）。
深度学习算法无法区分COVID-19肺炎与ICI治疗相关肺炎。临床医生必须提高对COVID-19与ICI治疗相关肺炎影像学相似性的认识。ICI治疗相关肺炎可作为交叉验证的挑战人群，以测试用于分析不同病因间质性肺炎的AI模型的稳健性。
