Abstract：
OBJECTIVE: Some patients with intracranial aneurysms (IAs) cannot undergo three-dimensional computed tomography angiography (3D-CTA) or digital subtraction angiography due to contraindications to contrast agents or radiation. Time-of-flight magnetic resonance angiography (TOF-MRA) offers a contrast-free alternative but lacks cranial bone detail critical for surgical planning. This study evaluates the feasibility of using 3D Slicer to fuse TOF-MRA with thin-section CT images to generate synthetic images resembling CTA for surgical clipping planning.
METHODS: This prospective study included 22 patients with unruptured IAs and 8 with ruptured IAs undergoing aneurysm clipping surgery (≥3 mm). TOF-MRA and CT/3D-CTA scans were fused using 3D Slicer. Neuroradiologists and neurosurgeons independently assessed 3D-CTA and synthetic TOF-MRA-CT images for aneurysm detection rates, morphology, and dimensions. Evaluation metrics included dice similarity coefficient and 95% Hausdorff distance.
RESULTS: Evaluation of aneurysm detection rates, morphology, and dimensions showed no significant differences between synthetic TOF-MRA-CT fusion images and 3D-CTA (all P > 0.05). Neuroradiologist assessments revealed strong concordance in aneurysm morphology between synthetic TOF-MRA-CT fusion images and 3D-CTA (κ = 0.867, P < 0.001). The dice similarity coefficient (0.937 ± 0.012) and Hausdorff distance (4.54 ± 0.26) indicated a high degree of image overlap between synthetic TOF-MRA-CT fusion images and 3D-CTA. Surgeons rated the consistency of aneurysm morphology between synthetic TOF-MRA-CT fusion images and intraoperative findings as strongly concordant (κ = 0.873, P < 0.001).
CONCLUSIONS: Synthetic TOF-MRA-CT fusion images closely match 3D-CTA for ≥3 mm aneurysms, demonstrating comparable diagnostic and surgical clipping planning effectiveness. They represent a promising alternative for personalized preoperative planning, particularly when contrast agents are contraindicated.
METHODS: This prospective study included 22 patients with unruptured IAs and 8 with ruptured IAs undergoing aneurysm clipping surgery (≥3 mm). TOF-MRA and CT/3D-CTA scans were fused using 3D Slicer. Neuroradiologists and neurosurgeons independently assessed 3D-CTA and synthetic TOF-MRA-CT images for aneurysm detection rates, morphology, and dimensions. Evaluation metrics included dice similarity coefficient and 95% Hausdorff distance.
RESULTS: Evaluation of aneurysm detection rates, morphology, and dimensions showed no significant differences between synthetic TOF-MRA-CT fusion images and 3D-CTA (all P > 0.05). Neuroradiologist assessments revealed strong concordance in aneurysm morphology between synthetic TOF-MRA-CT fusion images and 3D-CTA (κ = 0.867, P < 0.001). The dice similarity coefficient (0.937 ± 0.012) and Hausdorff distance (4.54 ± 0.26) indicated a high degree of image overlap between synthetic TOF-MRA-CT fusion images and 3D-CTA. Surgeons rated the consistency of aneurysm morphology between synthetic TOF-MRA-CT fusion images and intraoperative findings as strongly concordant (κ = 0.873, P < 0.001).
CONCLUSIONS: Synthetic TOF-MRA-CT fusion images closely match 3D-CTA for ≥3 mm aneurysms, demonstrating comparable diagnostic and surgical clipping planning effectiveness. They represent a promising alternative for personalized preoperative planning, particularly when contrast agents are contraindicated.
摘要：
目的：一些颅内动脉瘤（IA）患者由于造影剂或放射线的禁忌症而无法进行三维计算机断层扫描血管造影（3D-CTA）或数字减影血管造影（DSA）。飞行时间磁共振血管造影（TOF-MRA）提供了无对比的替代方法，但缺乏对手术计划至关重要的颅骨细节。这项研究评估了使用3DSlicer将TOF-MRA与薄层CT图像融合以生成类似于CTA的合成图像以进行手术剪裁计划的可行性。
方法：这项前瞻性研究包括22例接受动脉瘤夹闭手术（≥3mm）的未破裂IAs（UIA）和8例破裂IAs（RIA）患者。使用3DSlicer融合TOF-MRA和CT/3D-CTA扫描。神经放射科医生和神经外科医生独立评估3D-CTA和合成TOF-MRA-CT图像的动脉瘤检出率，形态学,和尺寸。评估指标包括骰子相似系数(DSC)和95%Hausdorff距离(HD)。
结果：动脉瘤检出率的评估，形态学,合成TOF-MRA-CT融合图像与3D-CTA的尺寸差异均无统计学意义（均p>0.05）。神经放射学家的评估显示，合成TOF-MRA-CT融合图像与3D-CTA之间的动脉瘤形态具有很强的一致性（κ=0.867，p<0.001）。DSC（0.937±0.012）和HD（4.54±0.26）表明合成TOF-MRA-CT融合图像与3D-CTA之间存在高度图像重叠。外科医生将合成TOF-MRA-CT融合图像与术中发现之间的动脉瘤形态一致性评价为高度一致(κ=0.873,p<0.001)。
结论：合成TOF-MRA-CT融合图像与≥3mm动脉瘤的3D-CTA紧密匹配，证明具有可比性的诊断和手术夹闭计划的有效性。它们代表了个性化术前计划的有希望的替代方案，特别是当造影剂是禁忌的。
方法：这项前瞻性研究包括22例接受动脉瘤夹闭手术（≥3mm）的未破裂IAs（UIA）和8例破裂IAs（RIA）患者。使用3DSlicer融合TOF-MRA和CT/3D-CTA扫描。神经放射科医生和神经外科医生独立评估3D-CTA和合成TOF-MRA-CT图像的动脉瘤检出率，形态学,和尺寸。评估指标包括骰子相似系数(DSC)和95%Hausdorff距离(HD)。
结果：动脉瘤检出率的评估，形态学,合成TOF-MRA-CT融合图像与3D-CTA的尺寸差异均无统计学意义（均p>0.05）。神经放射学家的评估显示，合成TOF-MRA-CT融合图像与3D-CTA之间的动脉瘤形态具有很强的一致性（κ=0.867，p<0.001）。DSC（0.937±0.012）和HD（4.54±0.26）表明合成TOF-MRA-CT融合图像与3D-CTA之间存在高度图像重叠。外科医生将合成TOF-MRA-CT融合图像与术中发现之间的动脉瘤形态一致性评价为高度一致(κ=0.873,p<0.001)。
结论：合成TOF-MRA-CT融合图像与≥3mm动脉瘤的3D-CTA紧密匹配，证明具有可比性的诊断和手术夹闭计划的有效性。它们代表了个性化术前计划的有希望的替代方案，特别是当造影剂是禁忌的。
