Room: ePoster Forums
Purpose: To quantify the random uncertainty levels due to patient movement during beam on and after CBCT correction during frameless Gamma Knife treatment and then identify potential correlative factors with high levels of uncertainty per patient.
Methods: IFMM log files were extracted for all treatments of 30 patients (IRB 2017-0216). These files were analyzed using Python for data extraction. The data of interest was displacement of fiducial at the time of CBCT and new reference position approval (post-CBCT residual error) and the fiducial displacement during “beam-on� time (patient movement). For the residual error, the X, Y, Z, and radial displacement were averaged over all acquired CBCT’s for all treatment fractions and then averaged for all patients. For movement error, the fiducial displacement over time is recorded in the log files for each displacement 0.2 mm greater than the last value. Since this data is displacement over time, the area-under-the-curve was calculated and then divided by the total IFMM tracking time of the patient in treatment position. Since the treatment is gated when displacement goes above a threshold, the values were capped at the set threshold value for the patient (usually 1.5 mm). After average displacement was calculated for each patient over all treatments it was summed with the average residual CBCT error for a patient-specific total random uncertainty. Bivariate statistics were then performed to identify any statistically significant relationships between patient and target factors and uncertainty.
Results: The only statistically significant relationship found was between BMI and the patient-specific total random uncertainty. The strength was moderate with r = 0.393, p = 0.032.
Conclusion: Due to a moderate, but significant correlation between BMI and total random uncertainty during frameless Gamma Knife treatment, perhaps practical steps could be taken to mediate potential, additional uncertainty experienced by patients with high BMI values.