学术文献库

Purpose: Automatic planning for focal low-dose-rate prostate brachytherapy (LDR-PB) boost. Methods: A simulated annealing approach was utilized for creating automatic focal LDR-PB boost plans using both geometric (i.e. adherence of isodose contours to structure contours) and dosimetric constraints. In 46 patients, four unilateral focal lesion scenarios were considered as the boost region, in the anterior/posterior and superior/inferior quadrants. Plans consisted of seeds with two air kerma strengths (0.740U and 0.351 or 0.354U) to achieve required dose coverage on these smaller focal targets. Plan characteristics (including PTV V100 and V150, Rectum V50, Urethra V150, and number of needles and seeds), were compared in terms of type of needle loading (dual vs. single air kerma needles), boost quadrant and prostate volume. Plans were blindly ranked on 20 consecutive cases by two experienced radiation oncologists. Results: Differences in plans generated using the two needle loading techniques were not clinically significant, particularly in terms of the PTV V100 and V150 and Urethra V150, with single air kerma loaded needles potentially being more practical. A higher rectal dose was observed in the posterior boost quadrants (V50 >0.18), particularly in larger prostates (>40 cc). In smaller prostates (<25 cc), obtaining adequate dose coverage while sparing the urethra was more challenging. 88% of the plans were considered clinically acceptable without (46%) or with (mostly minor) modifications. Conclusions: The novel approach of using dual air kerma seeds combined with geometric constraints within a simulated annealing-based automatic planning framework can lead to focal LDR-PB plans with high acceptability.