Premium
TH‐A‐9A‐05: Initial Setup Accuracy Comparison Between Frame‐Based and Frameless Stereotactic Radiosurgery
Author(s) -
Tseng T,
Sheu R,
Todorov B,
Green S,
Blacksburg S,
Lo Y
Publication year - 2014
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4889575
Subject(s) - radiosurgery , isocenter , nuclear medicine , linear particle accelerator , medicine , radiation therapy , surgery , imaging phantom , optics , physics , beam (structure)
Purpose: To evaluate initial setup accuracy for stereotactic radiosurgery (SRS) between Brainlab frame‐based and frameless immobilization system, also to discern the magnitude frameless system has on setup parameters. Methods: The correction shifts from the original setup were compared for total 157 SRS cranial treatments (69 frame‐based vs. 88 frameless). All treatments were performed on a Novalis linac with ExacTrac positioning system. Localization box with isocenter overlay was used for initial setup and correction shift was determined by ExacTrac 6D auto‐fusion to achieve submillimeter accuracy for treatment. For frameless treatments, mean time interval between simulation and treatment was 5.7 days (range 0–13). Pearson Chi‐Square was used for univariate analysis. Results: The correctional radial shifts (mean±STD, median) for the frame and frameless system measured by ExacTrac were 1.2±1.2mm, 1.1mm and 3.1±3.3mm, 2.0mm, respectively. Treatments with frameless system had a radial shift >2mm more often than those with frames (51.1% vs. 2.9%; p<.0001). To achieve submillimeter accuracy, 85.5% frame‐based treatments did not require shift and only 23.9% frameless treatment could succeed with initial setup. There was no statistical significant system offset observed in any direction for either system. For frameless treatments, those treated ≥ 3 days from simulation had statistically higher rates of radial shifts between 1–2mm and >2mm compared to patients treated in a shorter amount of time from simulation (34.3% and 56.7% vs. 28.6% and 33.3%, respectively; p=0.006). Conclusion: Although image‐guided positioning system can also achieve submillimeter accuracy for frameless system, users should be cautious regarding the inherent uncertainty of its capability of immobilization. A proper quality assurance procedure for frameless mask manufacturing and a protocol for intra‐fraction imaging verification will be crucial for frameless system. Time interval between simulation and treatment was influential to initial setup accuracy. A shorter time frame for frameless SRS treatment could be helpful in minimizing uncertainties in localization.