一个基于多类放射学方法的世卫组织严重程度尺度,用于从CT扫描中改善COVID-19患者评估和疾病特征
John Anderson Garcia Henao1, Arno Depotter, Danielle V Bower
1From the ARTORG Center for Biomedical Research, University of Bern, Bern, Switzerland (J.A.G.H., M.R.); Department of Diagnostic, Interventional, and Pediatric Radiology, Inselspital Bern, University of Bern, Bern, Switzerland (A.D., D.V.B., H.B., P.T.T., H.S.-J., M.C.B., H.M.B., A.P.); Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT (J.H., L.H.S., C.G., J.S.D.); Department of Biomedical Engineering, Yale University, New Haven, CT (J.H., J.Y., L.H.S., J.S.D.); Department of Electrical Engineering, Yale University, New Haven, CT (C.Y.); Section of "Scienze Radiologiche," Diagnostic Department, University Hospital of Parma, Parma, Italy (R.E.L., M.S., N.S.); Department of Medicine and Surgery, University of Parma, Italy (R.E.L., N.S.); Ricerca Clinica ed Epidemiologica, University Hospital of Parma, Parma, Italy (C.C.); Department of Radiology at Mayo Clinic College of Medicine and Science, Florida, Jacksonville, FL (I.O.C.); Section of Pulmonary, Critical Care, and Sleep Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT (C.S.D.C.); Department of Emergency Medicine, Inselspital University Hospital, University of Bern, Bern, Switzerland (W.H.); Campusradiologie, Department of Radiological Diagnostics, Lindenhofspital Bern, Bern, Switzerland (H.M.B.).; Campus Stiftung Lindenhof Bern, Bern, Switzerland (H.M.B.); and Department of Radiation Oncology, Inselspital, Bern University Hospital, Bern, Switzerland (M.R.).
一个新的AI模型AssessNet-19使用多类肺病变分析准确评估COVID-19的严重程度. 这种人工智能方法在胸部CT扫描中优于放射科医生和单一类模型.
科学领域:
- 放射学和人工智能的人工智能
- 医学成像分析 医学成像分析
- 计算病理学计算病理学
背景情况:
- 胸部计算机断层扫描 (CT) 对于评估COVID-19严重程度至关重要.
- 肺病变的准确细分和分类对于疾病的分期至关重要.
- 现有的模型经常使用单类细分,可能会限制准确性.
研究的目的:
- 开发和评估AssessNet-19,用于多类肺病变细分和COVID-19严重程度分类的AI模型.
- 将AssessNet-19的性能与单一类模型和专家放射科医生的性能进行比较.
- 为了验证模型能够根据世界卫生组织临床进展量表预测疾病严重程度的能力.
主要方法:
- 一个2D-U-Net网络被训练为四个COVID-19引起的肺病变的多类细分.
- 放射性特征被提取,使用LASSO回归减少,并输入到XGBoost分类器中.
- 该模型在两个独立的多中心队列上进行了验证,用于手动和自动评估.
主要成果:
- 与放射科医生 (0.63) 和单一类型模型 (0.64) 相比,AssessNet-19在严重性分类中获得了0.76的优异F1得分.
- 自动细分实现了不同的损伤类型的0.30到0.70的Dice分数.
- 观察到AssessNet-19和放射科医生在量化疾病程度方面达成高度一致 (科恩 κ: 0.920.95).
结论:
- 开发的人工智能多类放射学模型 (AssessNet-19) 准确评估了COVID-19疾病的严重程度.
- 在胸部CT分析中,AssessNet-19在单类模型和专家放射科医生中表现出优异的性能.
- 这种人工智能工具为客观和精确的COVID-19严重程度评估提供了一个有希望的方法.
相关概念视频
Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT
Radiological Investigation I: X-ray and CT
Imaging Studies for Cardiovascular System V: CT
Radiological Investigation II: MRI and Ventilation Perfusion Scan
Magnetic Resonance Imaging (MRI) and Ventilation Perfusion Scans are two radiological investigations that offer detailed diagnostic images of the body, particularly lung structures.
MRI
MRI uses magnetic fields and radiofrequency signals to distinguish between normal and abnormal tissues. This technology provides a more detailed diagnostic image than CT scans, enabling it to characterize pulmonary nodules, stage bronchogenic carcinoma, and evaluate inflammatory activity in...
Radiological Investigation III: Pulmonary Angiogram and PET Scan
Pulmonary Angiogram
A Pulmonary Angiogram is an invasive procedure involving injecting a contrast medium through a catheter threaded into the pulmonary artery or the right side of the heart to visualize the pulmonary vasculature. Computed Tomography (CT) scans have mainly replaced this...
Imaging Studies III: Computed Tomography


