Department of Forensic Medicine and Imaging, Zurich Institute of Forensic Medicine (ZIFM), University of Zurich, Winterthurerstrasse 190/52, 8057 Zurich, Switzerland.
Horten Centre for Patient Oriented Research and Knowledge Transfer, University of Zurich, Pestalozzistrasse 24, 8091 Zurich, Switzerland.
Forensic Sci Int. 2019 Oct;303:109942. doi: 10.1016/j.forsciint.2019.109942. Epub 2019 Sep 12.
Rib fractures are common and potentially life-threatening. Fast and correct detection as well as comprehensive visual overview of rib fractures are of clinical and forensic importance. This study compared two computed tomography (CT) reformation methods, curved planar reformation (CPR) with conventional multiplanar reformation (MPR), regarding detection of rib fractures in different readers.
Twelve postmortem CT datasets were retrospectively assessed for rib fractures using CPR and MPR. After defining the gold-standard regarding side, level, localization, and quantity of fractures, four reader groups per two readers consisting of radiologists, trauma surgeons, forensic pathologists, and laypersons, were evaluated for sensitivity, proportion of false positives, time to fracture detection, and subjective preference.
Overall sensitivity for fracture detection did not vary significantly between both methods. However, it was significantly higher in trauma surgeons and laypersons when reading CPR compared to MPR (70.7% vs. 62.0%, p=0.038 and 33.7% vs. 22.1%, p=0.003 respectively). It was significantly lower in radiologists (63.8% vs. 76.8%, p=0.001). Forensic pathologists performed similarly with both methods (53.6% vs. 56.5%, p=0.549). All non-radiologists preferred the use of CPR (75%). All readers found CPR to provide better visual overview (100%).
CPR may increase rib fracture detection rates of non-radiologists (i.e. trauma surgeons and laypersons) and provides a better visual overview. However, radiologists achieve higher fracture detection rates when allowed to work with the software tools they are more experienced with. The overall sensitivity was improvable and better visualization methods are warranted in order to avoid misdiagnosis and medicolegal errors regarding rib fracture detection.
肋骨骨折很常见,且有潜在的生命危险。快速准确地检测肋骨骨折,并全面观察肋骨骨折的情况,具有临床和法医学意义。本研究比较了两种计算机断层扫描(CT)重建方法,即曲面重建(CPR)与传统多平面重建(MPR),以评估不同读者对肋骨骨折的检测效果。
回顾性地对 12 例尸检 CT 数据集使用 CPR 和 MPR 进行肋骨骨折检测。在定义骨折的侧别、水平、定位和数量的金标准后,由两名放射科医生、创伤外科医生、法医病理学家和非专业人士组成的四个读者组,分别评估两种方法的敏感性、假阳性比例、骨折检测时间和主观偏好。
两种方法的总体骨折检测敏感性无显著差异。然而,与 MPR 相比,CPR 可显著提高创伤外科医生和非专业人士的骨折检出率(70.7% vs. 62.0%,p=0.038 和 33.7% vs. 22.1%,p=0.003)。而在放射科医生中,CPR 的骨折检出率显著降低(63.8% vs. 76.8%,p=0.001)。法医病理学家使用两种方法的效果相似(53.6% vs. 56.5%,p=0.549)。所有非放射科医生均更喜欢使用 CPR(75%)。所有读者均认为 CPR 可提供更好的整体视图(100%)。
CPR 可提高非放射科医生(即创伤外科医生和非专业人士)对肋骨骨折的检出率,并提供更好的整体视图。然而,放射科医生在使用其更熟悉的软件工具时,可获得更高的骨折检出率。为了避免在肋骨骨折检测中误诊和医疗法律错误,整体敏感性仍有待提高,需要更好的可视化方法。