Department of Diagnostic Radiology and Nuclear Medicine, R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, 22 S Greene St, Baltimore, MD, 21201, USA.
Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, MD, USA.
Emerg Radiol. 2022 Dec;29(6):995-1002. doi: 10.1007/s10140-022-02087-5. Epub 2022 Aug 16.
We employ nnU-Net, a state-of-the-art self-configuring deep learning-based semantic segmentation method for quantitative visualization of hemothorax (HTX) in trauma patients, and assess performance using a combination of overlap and volume-based metrics. The accuracy of hemothorax volumes for predicting a composite of hemorrhage-related outcomes - massive transfusion (MT) and in-hospital mortality (IHM) not related to traumatic brain injury - is assessed and compared to subjective expert consensus grading by an experienced chest and emergency radiologist.
The study included manually labeled admission chest CTs from 77 consecutive adult patients with non-negligible (≥ 50 mL) traumatic HTX between 2016 and 2018 from one trauma center. DL results of ensembled nnU-Net were determined from fivefold cross-validation and compared to individual 2D, 3D, and cascaded 3D nnU-Net results using the Dice similarity coefficient (DSC) and volume similarity index. Pearson's r, intraclass correlation coefficient (ICC), and mean bias were also determined for the best performing model. Manual and automated hemothorax volumes and subjective hemothorax volume grades were analyzed as predictors of MT and IHM using AUC comparison. Volume cut-offs yielding sensitivity or specificity ≥ 90% were determined from ROC analysis.
Ensembled nnU-Net achieved a mean DSC of 0.75 (SD: ± 0.12), and mean volume similarity of 0.91 (SD: ± 0.10), Pearson r of 0.93, and ICC of 0.92. Mean overmeasurement bias was only 1.7 mL despite a range of manual HTX volumes from 35 to 1503 mL (median: 178 mL). AUC of automated volumes for the composite outcome was 0.74 (95%CI: 0.58-0.91), compared to 0.76 (95%CI: 0.58-0.93) for manual volumes, and 0.76 (95%CI: 0.62-0.90) for consensus expert grading (p = 0.93). Automated volume cut-offs of 77 mL and 334 mL predicted the outcome with 93% sensitivity and 90% specificity respectively.
Automated HTX volumetry had high method validity, yielded interpretable visual results, and had similar performance for the hemorrhage-related outcomes assessed compared to manual volumes and expert consensus grading. The results suggest promising avenues for automated HTX volumetry in research and clinical care.
我们使用 nnU-Net，一种最先进的自配置深度学习语义分割方法，对创伤患者的血胸（HTX）进行定量可视化，并使用基于重叠和基于体积的度量标准组合来评估性能。评估并比较了血胸体积对预测与创伤性脑损伤无关的大量输血（MT）和院内死亡率（IHM）等出血相关结果的复合指标的准确性，这些结果是由一位经验丰富的胸部和急诊放射科医生进行的主观专家共识分级得出的。
这项研究包括 2016 年至 2018 年期间，从一家创伤中心的 77 名连续成人非小量（≥50ml）创伤性 HTX 患者的入院胸部 CT 手动标记。通过五重交叉验证确定了集成 nnU-Net 的深度学习结果，并使用迪奇相似系数（DSC）和体积相似性指数比较了单个 2D、3D 和级联 3D nnU-Net 的结果。还确定了最佳模型的 Pearson r、组内相关系数（ICC）和平均偏差。使用 AUC 比较分析手动和自动血胸体积以及主观血胸体积分级作为 MT 和 IHM 的预测因子。通过 ROC 分析确定产生敏感性或特异性≥90%的体积截止值。
集成 nnU-Net 的平均 DSC 为 0.75（SD：±0.12），平均体积相似性为 0.91（SD：±0.10），Pearson r 为 0.93，ICC 为 0.92。尽管手动 HTX 体积范围从 35 到 1503ml（中位数：178ml），但平均过度测量偏差仅为 1.7ml。自动体积对复合结果的 AUC 为 0.74（95%CI：0.58-0.91），与手动体积的 0.76（95%CI：0.58-0.93）和专家共识分级的 0.76（95%CI：0.62-0.90）相比（p=0.93）。自动体积截止值为 77ml 和 334ml 时，预测结果的敏感性分别为 93%和特异性为 90%。
自动 HTX 体积测量具有较高的方法有效性，产生了可解释的视觉结果，并且与手动体积和专家共识分级相比，对评估的出血相关结果具有相似的性能。结果表明，在研究和临床护理中，自动 HTX 体积测量具有广阔的应用前景。
