Wakker Alexander M, Van Lieshout Esther M M, De Boer A Siebe, Cornelissen Bart M W, Verhofstad Michael H J, Van Walsum Theo, Visser Jacob J, Van Vledder Mark G
Trauma Research Unit, Department of Surgery, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands.
Department of Radiology & Nuclear Medicine, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands.
Quant Imaging Med Surg. 2024 Jun 1;14(6):3778-3788. doi: 10.21037/qims-24-142. Epub 2024 Apr 17.
While current preoperative and postoperative assessment of the fractured and surgically reconstructed calcaneus relies on computed tomography (CT)-imaging, there are no established methods to quantify calcaneus morphology on CT-images. This study aims to develop a semi-automated method for morphological measurements of the calcaneus on three-dimensional (3D) models derived from CT-imaging.
Using CT data, 3D models were created from healthy, fractured, and surgically reconstructed calcanei. Böhler's angle (BA) and Critical angle of Gissane (CAG) were measured on conventional lateral radiographs and corresponding 3D CT reconstructions using a novel point-based method with semi-automatic landmark placement by three observers. Intraobserver and interobserver reliability scores were calculated using intra-class correlation coefficient (ICC). In addition, consensus among observers was calculated for a maximal allowable discrepancy of 5 and 10 degrees for both methods.
Imaging data from 119 feet were obtained (40 healthy, 39 fractured, 40 reconstructed). Semi-automated measurements on 3D models of BA and CAG showed excellent reliability (ICC: 0.87-1.00). The manual measurements on conventional radiographs had a poor-to-excellent reliability (ICC: 0.22-0.96). In addition, the percentage of consensus among observers was much higher for the 3D method when compared to conventional two-dimensional (2D) measurements.
The proposed method enables reliable and reproducible quantification of calcaneus morphology in 3D models of healthy, fractured and reconstructed calcanei.
虽然目前对跟骨骨折和手术重建的术前和术后评估依赖于计算机断层扫描(CT)成像,但尚无在CT图像上量化跟骨形态的既定方法。本研究旨在开发一种半自动方法,用于对从CT成像获得的三维(3D)模型上的跟骨进行形态学测量。
利用CT数据,创建健康、骨折和手术重建跟骨的3D模型。使用一种新型的基于点的方法,由三名观察者半自动放置地标,在传统侧位X线片和相应的3D CT重建图像上测量跟骨角(BA)和吉萨尼临界角(CAG)。使用组内相关系数(ICC)计算观察者内和观察者间的可靠性得分。此外,计算两种方法在最大允许差异为5度和10度时观察者之间的一致性。
获得了119只脚的影像数据(40只健康,39只骨折,40只重建)。对BA和CAG的3D模型进行半自动测量显示出极好的可靠性(ICC:0.87 - 1.00)。在传统X线片上进行的手动测量可靠性较差至极好(ICC:0.22 - 0.9)。此外,与传统二维(2D)测量相比,3D方法观察者之间的一致性百分比要高得多。
所提出的方法能够对健康、骨折和重建跟骨的3D模型中的跟骨形态进行可靠且可重复的量化。