Room: AAPM ePoster Library
To create a systematic procedure that provides improved quantifiable evaluation of the variation in the dose distribution of brachytherapy LDR prostate radioactive seed implant as a function of the seed position uncertainty.
In-house window-based software (BrachyVIC) has been coded (Python) to calculate the 3D dose distribution of brachytherapy LDR radioactive seeds arrangements, using AAPM TG43 formalism. The algorithm has been designed to calculate the 3D dose distribution and ROI DVHs at approximate rate of 100 times in 25 seconds. Plans generated in the treatment planning system are exported to BrachyVIC which randomly shifts the seeds up to 2mm in any direction and recalculate the dose distribution. The process of seed shifting and dose calculation is done 100 times. It further calculates the mean dose and the standard deviation (STD) of each ROI dosimetry alert (PTV_V100, PTV_V150, Pros_V100, Pros_V150, Urethra_V125 and Rectum_D1cc). The probability to meet each dosimetry alert is then calculated using the probability density function of the normal distribution.
In average, the standard deviation of PTV_V100 and PTV_V150 dosimetry alerts are 0.3% and 1.6%, respectively. Similar quantities are observed for prostate volumes and rectum but urethra goes to 18%. Different source arrangements mainly affect the standard deviation of PTV_V150, Pros_V150, and Urethra_125 but PTV_V100 and Pros_V100 have a consistent and stable metrics regardless of the source/seed distribution. Plan robustness is more correlated to the standard deviation of the dose alert but the probability to meet the objectives is a function of both the planned dosimetry alerts and the standard deviation.
Giving the dose uncertainty as a function of position uncertainty provides a more accurate and precise overview of the dose goals. Probability to meet objectives is a comprehensive metric to evaluate/compare LDR plans and provides an improved prediction of dosimetry outcome after seeds implant.