Estimation of imaging intervals and intrafraction displacement in CyberKnife image‐guided radiotherapy for intracranial lesions

Purpose A recent report by the American Association of Physicists in Medicine Task Group 75 and 180 provided imaging dose estimates for image‐guided CyberKnife radiotherapy. However, to our knowledge, there have been no concrete demonstrations of imaging intervals that are directly linked to exposure dose. We hypothesized that setting a rational standard may be clearer through a balance of treatment accuracy and reducing imaging doses if the margin of the planned treatment volume is controlled through the imaging interval. This study was conducted to simulate the association between the imaging interval and intrafraction displacement and to estimate a reasonable internal margin (IM). Methods We retrospectively analyzed data from 21 shell‐fixed heads of patients treated with CyberKnife G3 using our dedicated monitoring system. This system comprises pressure sensors that can monitor head displacement every 0.2 s in the absence of any imaging dose. First, the root sum square of head displacements was calculated in 76 treatment fractions with an imaging interval of 10–1440 s. The cumulative frequency of a root sum square displacement (which was less than the IM) was evaluated in image verifications that were undertaken 546 274 times for every imaging interval. Results We found that the mean values and SDs of the displacement were larger in proportion to the imaging interval (p 0.056). The cumulative frequencies of displacement of 0.6 and 1.4 mm (i.e., less than an IM) were 99.2% and 99.1% for imaging intervals of 10 and 360 s, respectively. Conclusions In the current study, we found that imaging intervals were directly proportional to intrafraction displacement and that there was no correlation in any combination within 720 s. Imaging intervals for an IM of 0.6 and 1.4 mm were 10 and 360 s, respectively, with a 99% confidence interval of intrafraction displacement. With CyberKnife M6 or a previous version of this system, the imaging dose could be reduced by 0.4760 mSv per 24‐min treatment as the imaging dose ranged from 0.4896 to 0.0136 mSv for imaging intervals of 10 and 360 s with an IM of 0.6 and 1.4 mm, respectively. A rational method that includes X‐ray imaging guidance may be achieved with modulation of the imaging interval via the CyberKnife system.