Intraobserver and interobserver variability of MR imaging- and CT-derived prostate volumes after transperineal interstitial permanent prostate brachytherapy

PURPOSE: To evaluate the relative accuracy and precision of magnetic resonance (MR) imaging and computed tomography (CT) in the assessment of postimplantation prostate volume by determining intraobserver, interobserver, and intermodality variations. MATERIALS AND METHODS: CT and MR images of 41 consecutive patients, after transperineal interstitial permanent prostate brachytherapy, were evaluated by two physicians to determine interobserver and intermodality variability in prostate volume measurements. Repeat evaluation in five randomly selected patients was used to determine intraobserver variability. RESULTS: Observer 1 versus 2 CT-determined mean prostate volume difference was statistically significant (-8.5 cm(3)+/-9.7[standard deviation],P <.001);observer 1 versus 2 MR-determined mean prostate volume difference was not significant (1.9cm(3)+/-11.7,P = .492). CT intraobserver range of dimensional errors was 3.5 and 11.4 times that of MR imaging. Observer 1 CT and MR volumes were significantly different (P = .001);observer 2 CT and MR volumes were not significantly different (P = .079). CONCLUSIONS: With both CT and MR imaging, variation is less when evaluations are conducted by one observer. Variation in one observer may be further reduced by using MR imaging in place of CT.