Kaviani Parisa, Kalra Mannudeep K, Digumarthy Subba R, Gupta Reya V, Dasegowda Giridhar, Jagirdar Ammar, Gupta Salil, Putha Preetham, Mahajan Vidur, Reddy Bhargava, Venugopal Vasanth K, Tadepalli Manoj, Bizzo Bernardo C, Dreyer Keith J
Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, USA.
MGH & BWH Center for Clinical Data Science, Boston, MA 02114, USA.
Diagnostics (Basel). 2022 Sep 30;12(10):2382. doi: 10.3390/diagnostics12102382.
Missed findings in chest X-ray interpretation are common and can have serious consequences.
Our study included 2407 chest radiographs (CXRs) acquired at three Indian and five US sites. To identify CXRs reported as normal, we used a proprietary radiology report search engine based on natural language processing (mPower, Nuance). Two thoracic radiologists reviewed all CXRs and recorded the presence and clinical significance of abnormal findings on a 5-point scale (1-not important; 5-critical importance). All CXRs were processed with the AI model (Qure.ai) and outputs were recorded for the presence of findings. Data were analyzed to obtain area under the ROC curve (AUC).
Of 410 CXRs (410/2407, 18.9%) with unreported/missed findings, 312 (312/410, 76.1%) findings were clinically important: pulmonary nodules ( = 157), consolidation (60), linear opacities (37), mediastinal widening (21), hilar enlargement (17), pleural effusions (11), rib fractures (6) and pneumothoraces (3). AI detected 69 missed findings (69/131, 53%) with an AUC of up to 0.935. The AI model was generalizable across different sites, geographic locations, patient genders and age groups.
A substantial number of important CXR findings are missed; the AI model can help to identify and reduce the frequency of important missed findings in a generalizable manner.
胸部X光解读中漏诊的情况很常见，且可能产生严重后果。
我们的研究纳入了在印度的三个地点和美国的五个地点获取的2407份胸部X光片（CXR）。为了识别报告为正常的CXR，我们使用了基于自然语言处理的专有放射学报告搜索引擎（mPower，Nuance）。两位胸科放射科医生对所有CXR进行了复查，并以5分制记录异常发现的存在情况及其临床意义（1 - 不重要；5 - 至关重要）。所有CXR均由人工智能模型（Qure.ai）进行处理，并记录发现结果的存在情况。对数据进行分析以获得ROC曲线下面积（AUC）。
在410份有未报告/漏诊发现的CXR中（410/2407，18.9%），312项发现（312/410，76.1%）具有临床重要性：肺结节（ = 157）、实变（60）、线状阴影（37）、纵隔增宽（21）、肺门增大（17）、胸腔积液（11）、肋骨骨折（6）和气胸（3）。人工智能检测到69项漏诊发现（69/131，53%），AUC高达0.935。该人工智能模型在不同地点、地理位置、患者性别和年龄组中具有通用性。
大量重要的CXR发现被漏诊；人工智能模型可以以通用的方式帮助识别并减少重要漏诊发现的频率。
