INTRODUCTION
Stereotactic radiosurgery (SRS) using the Gamma Knife system is an effective modality for the treatment of brain tumors. The precision of SRS is heavily dependent on accurate imaging, particularly high-resolution magnetic resonance imaging (MRI), for lesion localization and treatment planning. However, MRI can be limited by patient-related factors such as anatomical deformities or mobility restrictions that hinder optimal positioning during scanning. In such cases, inaccurate head placement may lead to distortion or misregistration of anatomical landmarks, thereby compromising the effectiveness and safety of the radiosurgical procedure. Unlike MRI, computed tomography (CT) imaging does not require placement within a rigid head coil, making it feasible even in patients with severe neck motion limitations. This allows acquisition of high-quality stereotactic reference images despite physical constraints that compromise magnetic resonance (MR) positioning.
MR/CT image fusion techniques have been developed to overcome these limitations, combining the superior soft tissue contrast of MRI with the geometric stability of CT. This fusion approach is especially valuable in patients with spinal conditions such as ankylosing spondylitis, which may severely limit cervical motion and interfere with traditional frame-based MRI acquisition. Several studies have demonstrated the improved accuracy of lesion targeting when MR/CT co-registration is applied during Gamma Knife radiosurgery (GKRS) [1-6]. Here, we present a rare case in which MR/CT fusion imaging was essential to overcome positioning limitations and achieve accurate radiosurgical planning.
CASE REPORT
A 79-year-old male with newly diagnosed lung cancer presented with three metastatic brain lesions requiring GKRS. A stereotactic frame (Leksell G-frame, Elekta) was applied prior to imaging; however, the patient had a history of ankylosing spondylitis, evident as bamboo spine on previous spinal CT, which severely limited cervical motion and prevented adequate head extension during MRI acquisition (Fig. 1).
Due to the restricted neck extension, the patient’s head could not be properly positioned within the MRI head coil. During the initial MR-only planning, we noticed fiducial misalignment suggestive of stereotactic geometric inaccuracy (Fig. 2A). Because the rigid MRI head coil constrains head insertion angle and head extension, adequate positioning could not be achieved in this patient with ankylosing spondylitis. In contrast, the open CT gantry allowed neutral positioning with external support, providing a geometrically stable reference for stereotactic registration. A CT scan was subsequently acquired with the stereotactic frame in place, and MR/CT fusion in GammaPlan (Elekta) corrected the spatial discrepancy, thereby enabling accurate target delineation (Fig. 2B).
The MR and CT images were fused using GammaPlan software, enabling high-accuracy co-registration despite the limitations of the initial MRI (Fig. 3). This fusion enabled accurate lesion targeting and facilitated precise treatment planning.
Comparative review of the radiosurgical plans revealed a noticeable discrepancy. To visualize stereotactic registration error, the 50% isodose line prescribed on the MR/CT fusion-based plan was overlaid onto the MR-only dataset: the MR-only-based plan showed inaccurate tumor contouring, while the MR/CT fusion-based plan demonstrated precise tumor targeting and appropriate dose coverage (Fig. 4).
This study was approved by the Institutional Review Board of Kyungpook National University Hospital (2025-05-009) with a waiver of informed consent for anonymized single-patient case reporting.
DISCUSSION
This case highlights several important considerations regarding imaging strategy in SRS for patients with limited cervical mobility. MR/CT fusion imaging effectively mitigates geometric uncertainty that arises from suboptimal MRI positioning. When fiducial misalignment or field-of-view limitations compromise MRI accuracy, rigid or deformable co-registration with CT can restore spatial fidelity. Pappas et al. [1] demonstrated that MR/CT fusion reduced centroid error from over 1.5 mm to less than 0.5 mm in phantom studies. Nakazawa et al. [2] corroborated these findings in a clinical setting, with mean targeting error across 40 targets measured at just 0.3 mm.
Furthermore, prior studies emphasize the need to assess and address image distortion, particularly in 3T MRI systems. Peripheral distortions greater than 1 mm, as reported by Theocharis et al. [3], can significantly impact targeting accuracy if not corrected. Tools such as distortion phantoms and standardized registration protocols, as explored by Damyanovich et al. [4] and Ulin et al. [5], ensure co-registration consistency across institutions and clinical platforms. These methods are essential to maintain the high precision required for SRS.
In addition, the clinical value of MR/CT fusion becomes especially evident in patients with musculoskeletal deformities, such as those with ankylosing spondylitis. These patients, often elderly and frail, may not tolerate the strict positioning needed for standard frame-based MRI protocols. Yu et al. [6] demonstrated that image fusion effectively compensated for such limitations, supporting accurate targeting despite mechanical limitations. Consistent use of fusion techniques or distortion correction improves tumor control rates without increasing toxicity, even in anatomically complex cases.
Taken together, these findings strongly support the integration of MR/CT fusion as a routine option within radiosurgical planning workflows. This approach enhances spatial accuracy, accommodates challenging patient anatomies, and upholds the high precision standards required in GKRS. Future directions may include real-time fusion algorithms and adaptive planning systems to further streamline the treatment process.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
