Data-driven determination and validation of an optimised mid-treatment imaging interval for prostate stereotactic body radiotherapy using real-time intrafraction image records

Intrafraction monitoring of prostate position during stereotactic body radiotherapy (SBRT) can correct stochastic displacement but often relies on fiducial markers. This retrospective study employed prostate monitoring data to determine time-based probability of prostate displacement, which was used to recommend an optimised mid-treatment imaging interval for prostate SBRT in patients without fiducials, where the probability of displacement greater than tolerance is less than 5%.

Retrospective monitoring data from 70 patients treated with kilovoltage intrafraction monitoring (KIM) software was used to determine when the probability of displacement exceeding 3 mm first reached 5%, representing the recommended mid-treatment imaging interval, and repeated for displacement magnitudes of 1 and 2 mm. Monitoring data from an additional 10 patients treated with KIM software was analysed to evaluate the imaging intervals determined.

Within the 70 patient cohort (1998 fractions), 5% probability of 3 mm of displacement first occurred at 190 s, while the 5% probability of 1 and 2 mm of displacements occurred at 40 and 95 s, respectively. When correcting the prostate position every 190 s, the likelihood of >3 mm displacement in the 10 patient cohort (50 fractions) remained below 5% for most time points.

This study recommended and validated a prostate SBRT mid-treatment imaging interval of 190 s, derived from time-based probability analysis of large cohort intrafraction prostate displacement data. It is compatible with 3D imaging included on standard teletherapy equipment without requiring fiducials. The generated probability distributions may inform imaging intervals tailored for specific clinical workflows or patient populations.