Purpose: The objective of this study was to measure and compare spatial accuracy of the catheter in TRUS volume images acquired in transverse mode and sagittal mode using computed tomography (CT) as the baseline.
Methods: Three aftermarket tissue-mimicking phantoms that consist of the prostate, urethra, and rectum anatomy were used for this study. A TRUS stepper with a catheter guiding grid system was set up on a CT table similar to the clinical setup. An ultrasound probe was used to scan the phantoms in both stepping-transverse mode and twister-sagittal mode before and after insertion of catheters with a pre-defined pattern. Immediately after the TRUS acquisition, CT images were acquired with the locked grid holding the catheters in the same position. Both ultrasound sound images were registered to their respective CT image based on the prostate and rectum surface. In transverse view, the center coordinate of the catheter in the CT (seen as a hypo-dense circle) and the center of echoes (seen as a hyper-echoic line) for ultrasound images was recorded. In sagittal and coronal views, the bevel of the catheter was identified and recorded in both ultrasound images and the reference CT. The bevel location served as the catheter depth in the phantom. A total of 26 catheters were analyzed. For each cartesian direction, spatial accuracy was compared between stepping-transverse mode and twister-sagittal mode using paired t-test.
Results: The mean difference of spatial accuracy between the two modes were 1.27, -0.07, and 1.23[mm] for left-right, ant-post, and in-depth respectively. Both left-right and in-depth direction showed statistically significant difference (p < 0.05).
Conclusion: Volume image acquired using twister-sagittal mode produced a larger error in the left-right direction. In both modes, the center of the catheter identified using CT was on average 2.35[mm] radially above the hyper-echoic line visualized in ultrasound images.