Wang Tony J C, Brisman Ronald, Lu Zheng Feng, Li Xiang, Isaacson Steven R, Shah Jinesh N, Yoshida Emi J, Liu Tian
Department of Radiation Oncology, Columbia University Medical Center, New York, NY 10032, USA.
Stereotact Funct Neurosurg. 2010;88(4):239-45. doi: 10.1159/000315461. Epub 2010 Jun 8.
BACKGROUND/AIMS: In Gamma Knife radiosurgery, T(1) MRI is most commonly used and is generally sufficient for targeting the trigeminal nerve. For patients whose trigeminal nerves are unclear on T(1) MRI, FIESTA MRI supplements anatomical structure visualization and may improve trigeminal nerve delineation. The purpose of this study was to develop a registration strategy for T(1) and FIESTA MRIs.
We conducted a retrospective study on 54 trigeminal neuralgia patients. All patients were scanned with T(1) and FIESTA MRIs. We evaluated 4 methods of registration: automatic image definition, superior-slice definition, middle-slice definition and inferior-slice definition. Target discrepancies were measured by deviations from an intracranial landmark on T(1) and FIESTA MR images.
The overall range in registration error was 0.10-5.19 mm using superior-, 0.10-1.56 mm using middle- and 0.14-2.89 mm using inferior-slice definition. Registration error >2 mm was observed in 11% of the patients using superior-, 4% using middle- and 7% using inferior-slice FIESTA MRI definition.
Among patients for whom FIESTA and T(1) MRI are used, registration based on middle-slice definition reduces registration error and improves targeting of the trigeminal nerve.
背景/目的:在伽玛刀放射外科治疗中,T(1)磁共振成像(MRI)最为常用,通常足以用于三叉神经的靶向定位。对于T(1) MRI上三叉神经显示不清的患者,稳态进动快速成像(FIESTA)MRI可补充解剖结构可视化,可能改善三叉神经的勾画。本研究的目的是制定一种T(1)和FIESTA MRI的配准策略。
我们对54例三叉神经痛患者进行了一项回顾性研究。所有患者均接受了T(1)和FIESTA MRI扫描。我们评估了4种配准方法:自动图像定义、上层面定义、中间层面定义和下层面定义。通过T(1)和FIESTA MR图像上与颅内标志物的偏差来测量靶点差异。
使用上层面定义时配准误差的总体范围为0.10 - 5.19毫米,使用中间层面定义时为0.10 - 1.56毫米,使用下层面定义时为0.14 - 2.89毫米。在使用上层面FIESTA MRI定义的患者中,11%观察到配准误差>2毫米,使用中间层面定义的患者中为4%,使用下层面定义的患者中为7%。
在使用FIESTA和T(1) MRI的患者中,基于中间层面定义的配准可减少配准误差并改善三叉神经的靶向定位。