The effect of delineation method and observer variability on bladder dose-volume histograms for prostate intensity modulated radiotherapy

Abstract

PurposeTo quantify the effect of delineation method on bladder DVH, observer variability (OV) and contouring time for prostate IMRT plans. Materials and methodsPlanning CT scans and IMRT plans of 30 prostate cancer patients were anonymized. For 20 patients, 1 observer delineated the bladder using 9 methods. The effect of delineation method on the DVH curve, discrete dose levels and delineation time was quantified. For the 10 remaining CTs, 6 observers delineated bladder wall using 4 methods. Observer-based volume variation and intraclass correlation coefficient (ICC) were used to describe the dosimetric effects of OV. ResultsManual delineation of the bladder wall (BW_m) was significantly slower than any other method (mean: 20min vs. ⩽13min) and the dosimetric effect of OV was significantly larger (V70Gy ICC: 0.78 vs. 0.98). Only volumes created using a 2.5mm contraction from the outer surface, and a method providing a consistent wall volume, showed no notable dosimetric differences from BW_m in both absolute and relative volume. ConclusionsAutomatic contractions from the outer surface provide quicker, more reproducible and reasonably accurate substitutes for BW_m. The widespread use of automatic contractions to create a bladder wall volume would assist in the consistent application of IMRT dose constraints and the interpretation of reported dose.