Examining the relationship between pre- and postimplant geometry in prostate low-dose-rate brachytherapy and its...

This is a retrospective study in which we define multiple metrics for similarity and then inquire on the relationship between similarity and currently used dosimetric quantities describing preimplant and postimplant plans.

We analyzed a unique cohort of 94 consecutively performed prostate seed implant patients, associated with excellent dosimetric and clinical outcomes. For each patient, an ultrasound (US) preimplant and two CT postimplant (Day 0 and Day 30) studies were available. Measures for similarity were created and computed using feature vectors based on two classes of moments: first, invariant to rotation and translation, and the second polar-radius moments invariant to rotation, translation, and scaling. Both similarity measures were calibrated using controlled perturbations (random and systematic) of seed positions and contours in different size implants, thus producing meaningful numerical threshold values used in the clinical analysis.

An important finding is that similarity, for both seed distributions and contours, improves significantly when scaling invariance is added to translation and rotation. No correlation between seed and contours similarity was found. In the setting of preplanned prostate seed implants using preloaded needles, based on our data, similarity between preimplant and postimplant plans does not correlate with either minimum dose to 90% of the volume of the prostate or analogous similarity metrics for prostate contours.

We have developed novel tools and metrics, which will allow practitioners to better understand the relationship between preimplant and postimplant plans. Geometrical similarity between a preplan and an actual implant, although useful, does not seem to be necessary to achieve minimum dose to 90% of the volume of the prostate-good dosimetric implants.