A general methodology for three-dimensional analysis of variation in target volume delineation

A generic method for three-dimensional (3-D) evaluation of target volume de lineation in multiple imaging modalities is presented. The evaluation inclu des geometrical and statistical methods to estimate observer differences an d variability in defining the Gross Tumor Volume (GTV) in relation to the d iagnostic CT and MRI modalities. The geometrical method is based on mapping the 3-D shape of the target volume to a scalar representation, thus enabli ng a one-dimensional statistical analysis. The statistical method distingui shes observer and modality related uncertainties, which are expressed in te rms of three error components: random observer deviations, systematic obser ver differences, and systematic modality differences. Monte Carlo simulatio ns demonstrate that the standard errors of each of the three model paramete rs are inversely proportional to the square root of the product of the pati ent group size and the number of observers and proportional to the intraobs erver variation. For 18 patients and 3 observers the standard errors of the estimated systematic modality and observer differences are 19% and 14% of the intraobserver standard deviation, respectively. A scalar representation of the shape of the prostate, delineated by 3 observers for 18 patients, w as obtained by sampling the distance between the average center of gravity of the prostate in CT and the prostate surface for a large number of direct ions (2500), using polar coordinates. Observer variability and differences were obtained by applying the statistical method to the samples independent ly. The intraobserver variation for CT was largest in regions near the semi nal vesicles (s.d: 3 mm) and the apex (s.d: 3 mm). The systematic observer variation in CT was largest in a region near the plexus Santorini, at the c audal-anterior side of the prostate (s.d.: 2 mm). The sensitivity for the c hoice of origin was tested by using the average center of gravity from axia l MRI instead of CT. The results were almost identical. The polar map measu res distances in the scanning directions. A correction procedure to get The variability in directions perpendicular to the surface of the prostate yie lded variations that were a factor of 0.85 smaller for all directions. It i s concluded that by separating the shape evaluation in a geometrical and a statistical part, the complexity of the analysis of 3-D shape differences c an be significantly reduced. The method was successfully applied to a group of prostate patients, where we demonstrated that delineation variability i s nonhomogeneous, with the largest variations occurring near the seminal ve sicles and the apex. (C) 1999 American Association of Physicists in Medicin e.