Department of Radiation Oncology, University of Pennsylvania, Philadelphia, PA.
L'Unità Operativa di Protonterapia, Azienda Provinciale per i Servizi Sanitari, Trento, Italy.
Int J Radiat Oncol Biol Phys. 2015 Mar 15;91(4):825-31. doi: 10.1016/j.ijrobp.2014.11.036.
To assess the feasibility of radiation therapy treatment planning 4-dimensional computed tomography (4DCT) and deep-inspiration breath-hold (DIBH) CT to accurately contour the left anterior descending artery (LAD), a primary indicator of radiation-induced cardiac toxicity for patients undergoing radiation therapy.
Ten subjects were prospectively imaged with a cardiac-gated MRI protocol to determine cardiac motion effects, including the displacement of a region of interest comprising the LAD. A series of planar views were obtained and resampled to create a 3-dimensional (3D) volume. A 3D optical flow deformable image registration algorithm determined tissue displacement during the cardiac cycle. The measured motion was then used as a spatial boundary to characterize motion blurring of the radiologist-delineated LAD structure for a cohort of 10 consecutive patients enrolled prospectively on a breast study including 4DCT and DIBH scans. Coronary motion-induced blurring artifacts were quantified by applying an unsharp filter to accentuate the LAD structure despite the presence of motion blurring. The 4DCT maximum inhalation and exhalation respiratory phases were coregistered to determine the LAD displacement during tidal respiration, as visualized in 4DCT.
The average 90th percentile heart motion for the region of interest was 0.7 ± 0.1 mm (left-right [LR]), 1.3 ± 0.6 mm (superior-inferior [SI]), and 0.6 ± 0.2 mm (anterior-posterior [AP]) in the cardiac-gated MRI cohort. The average relative increase in the number of voxels comprising the LAD contour was 69.4% ± 4.5% for the DIBH. The LAD volume overestimation had the dosimetric impact of decreasing the reported mean LAD dose by 23% ± 9% on average in the DIBH. During tidal respiration the average relative LAD contour increase was 69.3% ± 5.9% and 67.9% ± 4.6% for inhalation and exhalation respiratory phases, respectively. The average 90th percentile LAD motion was 4.8 ± 1.1 mm (LR), 0.9 ± 0.4 mm (SI), and 1.9 ± 0.6 mm (AP) for the 4DCT cohort, in the absence of cardiac gating.
An anisotropic margin of 2.7 mm (LR), 4.1 mm (SI), and 2.4 mm (AP) was quantitatively determined to account for motion blurring and patient setup error while placing minimum constraint on the plan optimization.
评估 4 维计算机断层扫描（4DCT）和深吸气屏气（DIBH）CT 在放射治疗中准确勾画左前降支（LAD）的可行性，LAD 是放射诱导心脏毒性的主要指标之一。
对 10 例患者进行前瞻性心脏门控 MRI 检查，以确定心脏运动的影响，包括包含 LAD 的感兴趣区的位移。获得一系列平面视图并重新采样以创建三维（3D）体积。3D 光流变形图像配准算法确定了心动周期中的组织位移。然后，将测量到的运动用作空间边界，以描述 10 例连续患者的放射科医生勾画的 LAD 结构的运动模糊，这些患者前瞻性地纳入包括 4DCT 和 DIBH 扫描的乳房研究。通过应用非锐化滤波器来突出 LAD 结构，尽管存在运动模糊，但仍可以量化冠状动脉运动引起的模糊伪影。将 4DCT 的最大吸气和呼气呼吸阶段配准，以确定在 4DCT 中可见的潮汐呼吸期间 LAD 的位移。
在心脏门控 MRI 队列中，感兴趣区域的平均 90 百分位数心脏运动为 0.7 ± 0.1mm（左右[LR]）、1.3 ± 0.6mm（上下[SI]）和 0.6 ± 0.2mm（前后[AP]）。DIBH 中 LAD 轮廓组成的体素数量平均增加 69.4%±4.5%。DIBH 中，LAD 体积的高估平均使报告的平均 LAD 剂量降低 23%±9%。在潮汐呼吸期间，LAD 轮廓的平均相对增加分别为吸气和呼气呼吸阶段的 69.3%±5.9%和 67.9%±4.6%。在没有心脏门控的情况下，4DCT 队列中 LAD 的平均 90 百分位数为 4.8 ± 1.1mm（LR）、0.9 ± 0.4mm（SI）和 1.9 ± 0.6mm（AP）。
定量确定了 2.7mm（LR）、4.1mm（SI）和 2.4mm（AP）的各向异性边界，以在最小限制计划优化的同时，考虑运动模糊和患者设置误差。
