Author: Tatiana Bejarano, Omer Gal, Alonso N. Gutierrez, Matthew D. Hall, Rupesh Kotecha, Yongsook C. Lee, Michael W. McDermott, Minesh P. Mehta, Robert H. Press, Carlos M. Rivera, Ranjini P. Tolakanahalli, D. Jay J. Wieczorek π¨βπ¬
Affiliation: Miami Cancer Institute, Baptist Health South Florida π
Purpose:
Gamma Knife (GK) cone beam computed tomography-based stereotactic reference (CBCTRef) combined with pre-planning MRI, provides a flexible alternative to the historical frame-based localization. This study evaluates the accuracy of CBCT-based stereotactic definition, paired with pre-plan MR images acquired with and without frame fixation for trigeminal neuralgia (TN) patients.
Methods:
Nine (n=9) patients treated with G-frame fixation were evaluated. Patients were prescribed 43 Gy to the 50% isodose line using a 4 mm shot placed to deliver a maximum dose of 86 Gy to the trigeminal nerveβs root entry zone. Shot positions defined with historical frame-based localization on MRI (WFH, gold standard) were compared to two alternative workflows: Workflow-1 [WF1] used CBCTRef with shots placed on co-registered MRI acquired with frame; and Workflow-2 [WF2] used CBCTRef with shots placed on co-registered MRI acquired without frame, simulated by applying a 5-degree pitch and yaw rotation before co-registration. Euclidean distances between shot positions using the three workflows were calculated. Shot placement variability was assessed with ten repeated placements per patient. Statistical significance was evaluated using the exact Wilcoxon Signed-Rank test.
Results:
Maximum and mean errors of fiducial position (Median[Range]) using WFH were 0.9[0.7-1.3] mm, and 0.5[0.4-0.7] mm, respectively. Maximum shot position differences in the [X, Y, and Z] directions between WF1 and WF2 compared to WFH were [0.9mm, 1.4mm, and 1.0mm] and [0.8mm, 1.8mm, and 0.7mm], respectively. Median euclidean distances between WF1 and WF2 compared to WFH were 1.0[0.3-1.7] mm and 0.9[0.7-2.0] mm, respectively (p=0.36). Repeatability analysis showed mean differences of 0.17[Ο =0.11] mm for WF1 and 0.23[Ο=0.09] mm WF2.
Conclusion:
Both CBCTRef workflows demonstrated statistically similar accuracy relative to historical workflow. CBCTRef, with pre-planning MRI without frame fixation, offers an accurate, alternate stereotactic solution while maintaining patient comfort and efficiency.