3207 - Inter- and Intra-Observer Variability in Catheter Localization in High-Dose-Rate Brachytherapy for Prostate Cancer with Stepping Transverse and Twister Sagittal Volume Scanning

S. K. J. Kim, P. H. Stone, S. Hoopingarner, X. Zhao, A. Wong, R. K. Valicenti, J. Matney, S. Dieterich, S. H. Benedict, and P. Park; UC Davis, Department of Radiation Oncology, Sacramento, CA

Purpose/Objective(s): Accurate needle localization technique is vital for a successful prostate HDR implant as it directly affects the plan quality and accurate dose delivery. Modern transrectal ultrasound and stepper systems provide two distinct volume acquisition modes: stepping transverse (ST) mode and twister sagittal mode (TS). The aim of this study is to quantify and compare the intra- and inter-observer variability (Intra-OV and Inter-OV) of catheter localization accuracy with both acquisition modes. Materials/

Methods: A total of 329 paired points of catheter tip from both ST and TS modes were defined by 7 observers (2 Physicians, 5 Physicists) from retrospective clinical plans. Each mode of images was presented, and the observers were asked to define the tip of catheters using 3D slice viewing software. Their coordinates were recorded along the Left-Right (LR), Anterior-Posterior (AP), and Superior-Inferior (SI, or depth) directions. For Intra-OV, the difference between the catheter tip positions to the reference position inferred by the physically measured free length positions were compared between the ST and TS modes for each observer. For Inter-OV quantification, the statistical variance in tip positions along all three cartesian directions defined by 7 observers were computed, and the values between the ST and TS modes were compared using a one-sided paired t-Test.
Results: Statistically significant differences in Inter-OV were observed in SI and LR directions only. For defining the depth of the needle, the sagittal mode reduced variability by 29.0% compared to the transverse mode (0.56mm vs 0.75mm, p < 0.05). For delineation of the left-right positioning of the needle, the transverse mode was more consistent than the sagittal mode, with a 22.2% reduction in mean Inter-OV (0.36mm vs 0.45mm, p < 0.05). For Intra-OV, the mean differences between the user defined tip points to the physical free length measurements were smaller for the sagittal mode than the transverse mode (0.03mm vs 0.69mm for TS and ST, respectively).
Conclusion: The different volume acquisition modes in transrectal ultrasound produce statistically significant differences in how the catheter is reconstructed during prostate HDR brachytherapy procedures. Each mode showed superiority in reducing observer variation along the different directions, with ST mode improving LR delineation and TS mode improving depth delineation. Both modes of image acquisition are complimentary for more accurate and precise catheter reconstruction for prostate HDR brachytherapy planning. Further investigation in how these variances affect dose distribution will be important to improving patient outcomes.