Department of Thoracic Surgery, Zuyderland Medical Centre, Heerlen, the Netherlands; Faculty of Health, Medicine and Life Sciences (FHML), School for Oncology and Developmental Biology (GROW), Maastricht, the Netherlands.
3D Lab Radboudumc, Radboud University Medical Centre, Nijmegen, the Netherlands.
J Surg Res. 2021 Mar;259:332-341. doi: 10.1016/j.jss.2020.09.027. Epub 2020 Oct 27.
Three-dimensional (3D) imaging is being used progressively to create models of patients with anterior chest wall deformities. Resulting models are used for clinical decision-making, surgical planning, and analysis. However, given the broad range of 3D imaging systems available and the fact that planning and analysis techniques are often only validated for a single system, it is important to analyze potential intrasystem and intersystem differences. The objective of this study was to investigate the accuracy and reproducibility of three commercially available 3D imaging systems that are used to obtain images of the anterior chest wall.
Among 15 healthy volunteers, 3D images of the anterior chest wall were acquired twice per imaging device. Reproducibility was determined by comparison of consecutive images acquired per device while the true accuracy was calculated by comparison of 3D image derived and calipered anthropometric measurements. A maximum difference of 1.00 mm. was considered clinically acceptable.
All devices demonstrated statistically comparable (P = 0.21) reproducibility with a mean absolute difference of 0.59 mm. (SD: 1.05), 0.54 mm. (SD: 2.08), and 0.48 mm. (SD: 0.60) for the 3dMD, EinScan Pro 2X Plus, and Artec Leo, respectively. The true accuracy was, respectively, 0.89 mm. (SD: 0.66), 1.27 mm. (SD: 0.94), and 0.81 mm. (SD: 0.71) for the 3dMD, EinScan, and Artec device and did not statistically differ (P = 0.085).
Three-dimensional imaging of the anterior chest wall utilizing the 3dMD and Artec Leo is feasible with comparable reproducibility and accuracy, whereas the EinScan Pro 2X Plus is reproducible but not clinically accurate.
三维（3D）成像技术正逐渐应用于前胸部壁畸形患者的模型创建。所生成的模型可用于临床决策、手术规划和分析。然而，鉴于可用的 3D 成像系统种类繁多，且规划和分析技术通常仅针对单一系统进行验证，因此分析潜在的系统内和系统间差异很重要。本研究的目的是研究三种用于获取前胸部壁图像的商业上可用的 3D 成像系统的准确性和可重复性。
在 15 名健康志愿者中，每个成像设备采集两次前胸部壁的 3D 图像。通过比较每个设备连续采集的图像来确定重复性，而通过比较 3D 图像得出的和卡尺测量的人体测量学数据来计算真实准确性。认为 1.00 毫米的最大差异具有临床可接受性。
所有设备的重复性均具有统计学可比性（P=0.21），平均绝对差异分别为 0.59 毫米（SD：1.05）、0.54 毫米（SD：2.08）和 0.48 毫米（SD：0.60），用于 3dMD、EinScan Pro 2X Plus 和 Artec Leo。真实准确性分别为 0.89 毫米（SD：0.66）、1.27 毫米（SD：0.94）和 0.81 毫米（SD：0.71），用于 3dMD、EinScan 和 Artec 设备，且无统计学差异（P=0.085）。
使用 3dMD 和 Artec Leo 对前胸部壁进行 3D 成像具有可重复性和准确性，且具有可比性，而 EinScan Pro 2X Plus 具有可重复性但不具有临床准确性。
