Background/Objectives: 5DCT was first proposed in 2005 as a motion-compensated CT simulation approach for radiotherapy treatment planning to avoid sorting artifacts that arise in 4DCT when patients breathe irregularly. Since March 2019, 5DCT has been clinically implemented for routine use at our institution to leverage this technological advantage. The clinical workflow includes a quality assurance report that describes the output of primary workflow steps. This study reports on the challenges and quality of the clinical 5DCT workflow using these quality assurance reports. Methods: We evaluated all thoracic 5DCT simulation datasets consecutively acquired at our institution between March 2019 and December 2022 for thoracic radiotherapy treatment planning. The 5DCT datasets utilized motion models constructed from 25 fast-helical free-breathing computed tomography (FHFBCTs) with simultaneous respiratory bellows signal monitoring to reconstruct individual, user-specified breathing-phase images (termed 5DCT phase images) for internal target volume contouring. Each 5DCT dataset was accompanied by a structured quality assurance report composed of qualitative and quantitative measures of the breathing pattern, image quality, DIR quality, model fitting accuracy, and a validation process by which the original FHFBCT scans were regenerated with the 5DCT model. Measures of breathing irregularity, image quality, and DIR quality were retrospectively categorized on a grading scale from 1 (regular breathing and accurate registration/modeling) to 4 (irregular breathing and inaccurate registration/modeling). The validation process was graded according to the same scale, and this grade was termed the suitability-for-treatment-planning (STP) grade. We correlated the graded variables to the STP grade. In addition to the quality assurance reports, we reviewed the contour sessions to determine how often 5DCT phase images were used for treatment planning and delivery. Results: There were 169 5DCT simulation datasets available from 156 patients for analysis. The STP was moderately correlated with breathing irregularity, image quality, and DIR quality (Spearman coefficients: 0.26, 0.30, and 0.50, respectively). Multiple linear regression analysis demonstrated that STP was correlated with regular breathing patterns (p= 0.008), image quality (p< 0.001), and better DIR quality (p< 0.001). 5DCT datasets were used for treatment planning in 82% of cases, while in 12% of cases, a backup image process was used. In total, 6% of image datasets were not used for treatment planning due to factors unrelated to the 5DCT workflow quality. Conclusions: The strongest association with STP was with DIR quality grades, as indicated by both Spearman and multiple linear regression analysis, implying that improvements to DIR accuracy and evaluation may be the best route for further improvement to 5DCT. The high rate of 5DCT phase image use for treatment planning showed that the workflow was reliable, and this has encouraged us to continue to develop and improve the workflow steps.
背景/目的：5DCT于2005年首次提出，作为一种用于放射治疗计划的运动补偿CT模拟方法，旨在避免4DCT在患者呼吸不规则时产生的排序伪影。自2019年3月起，我院已在临床常规工作中实施5DCT，以利用这一技术优势。临床工作流程包括一份描述主要工作步骤输出的质量保证报告。本研究基于这些质量保证报告，报告了临床5DCT工作流程面临的挑战及其质量。方法：我们评估了2019年3月至2022年12月期间我院为胸部放射治疗计划连续采集的所有胸部5DCT模拟数据集。5DCT数据集利用从25次快速螺旋自由呼吸计算机断层扫描（FHFBCTs）构建的运动模型，同时监测呼吸气囊信号，以重建用户指定的单个呼吸时相图像（称为5DCT时相图像），用于内靶区勾画。每个5DCT数据集都附有一份结构化的质量保证报告，该报告包含对呼吸模式、图像质量、DIR质量、模型拟合准确性的定性和定量评估，以及一个验证过程，即使用5DCT模型重新生成原始的FHFBCT扫描。对呼吸不规则性、图像质量和DIR质量的评估进行了回顾性分级，分级标准从1级（呼吸规则，配准/建模准确）到4级（呼吸不规则，配准/建模不准确）。验证过程也按照相同的标准进行分级，该等级被称为"适合治疗计划"（STP）等级。我们将分级变量与STP等级进行了关联分析。除了质量保证报告外，我们还审查了勾画环节，以确定5DCT时相图像用于治疗计划和实施的使用频率。结果：共有来自156名患者的169个5DCT模拟数据集可用于分析。STP等级与呼吸不规则性、图像质量和DIR质量呈中等程度相关（Spearman系数分别为：0.26、0.30和0.50）。多元线性回归分析表明，STP等级与规则的呼吸模式（p=0.008）、良好的图像质量（p<0.001）以及更好的DIR质量（p<0.001）相关。在82%的病例中，5DCT数据集被用于治疗计划，而在12%的病例中使用了备用图像流程。总共有6%的图像数据集未用于治疗计划，原因与5DCT工作流程质量无关。结论：根据Spearman相关分析和多元线性回归分析，与STP等级关联最强的是DIR质量等级，这意味着提高DIR准确性和评估可能是进一步改进5DCT的最佳途径。5DCT时相图像用于治疗计划的高使用率表明该工作流程是可靠的，这鼓励我们继续开发和改进工作流程的各个步骤。
A Retrospective Analysis of the First Clinical 5DCT Workflow
肿瘤(癌症)患者之家