Evaluation of interobserver differences in postimplant dosimetry following prostate brachytherapy and the efficacy of CT/MRI fusion imaging

Abstract

Interobserver differences in postimplant dosimetry based on computed tomography (CT) and CT/magnetic resonance imaging (MRI) fusion images were assessed to evaluate the efficacy of the fusion image. In addition, the part of the prostate contour responsible for the interobserver differences in CT was identified. In June 2008, 1-month postimplant CT data from two patients who underwent (125)I prostate brachytherapy were sent to 90 institutions for postimplant dosimetry. Subsequently, MRI data from the same patients were sent for fusion-based postimplant dosimetry. The variance of the difference between MRIbased D90 and CT-based or fusion-based D90 was compared. Prostate volume on CT was plotted on the y-axis against the position of the most cranial and caudal slices in the prostate contour delineated at each institution to analyze interobserver differences. The prostate volume from CT was significantly greater than from the CT/MRI fusion image (P = 0.0014). Fusion-based variance was significantly greater than CT-based variance (P < 0.01). CT-based postimplant dosimetry showed that 88%-96% of the institutions had an apical and basal position within a range of 5 mm. There were marked interobserver differences in CT/MRI fusion-based postimplant dosimetry.