Real-time catheter tracking for high-dose-rate prostate brachytherapy using an electromagnetic 3D-guidance device: a preliminary performance study.

PURPOSE: In order to increase the accuracy and speed of catheter reconstruction in a high-dose-rate (HDR) prostate implant procedure, an automatic tracking system has been developed using an electromagnetic (EM) device (trakSTAR, Ascension Technology, VT). The performance of the system, including the accuracy and noise level with various tracking parameters and conditions, were investigated.
METHODS: A direct current (dc) EM transmitter (midrange model) and a sensor with diameter of 1.3 mm (Model 130) were used in the trakSTAR system for tracking catheter position during HDR prostate brachytherapy. Localization accuracy was assessed under both static and dynamic analyses conditions. For the static analysis, a calibration phantom was used to investigate error dependency on operating room (OR) table height (bottom vs midposition vs top), sensor position (distal tip of catheter vs connector end of catheter), direction [left-right (LR) vs anterior-posterior (AP) vs superior-inferior (SI)], sampling frequency (40 vs 80 vs 120 Hz), and interference from OR equipment (present vs absent). The mean and standard deviation of the localization offset in each direction and the corresponding error vectors were calculated. For dynamic analysis, the paths of five straight catheters were tracked to study the effects of directions, sampling frequency, and interference of EM field. Statistical analysis was conducted to compare the results in different configurations.
RESULTS: When interference was present in the static analysis, the error vectors were significantly higher at the top table position (3.3 ± 1.3 vs 1.8 ± 0.9 mm at bottom and 1.7 ± 1.0 mm at middle, p < 0.001), at catheter end position (3.1 ± 1.1 vs 1.4 ± 0.7 mm at the tip position, p < 0.001), and at 40 Hz sampling frequency (2.6 ± 1.1 vs 2.4 ± 1.5 mm at 80 Hz and 1.8 ± 1.1 at 160 Hz, p < 0.001). So did the mean offset errors in the LR direction (-1.7 ± 1.4 vs 0.4 ± 0.5 mm in AP and 0.8 ± 0.8 mm in SI directions, p < 0.001). The error vectors were significantly higher with surrounding interference (2.2 ± 1.3 mm) vs without interference (1.0 ± 0.7 mm, p < 0.001). An accuracy of 1.6 ± 0.2 mm can be reached when using optimum configuration (160 Hz at middle table position). When interference was present in the dynamic tracking, the mean tracking errors in LR direction (1.4 ± 0.5 mm) was significantly higher than that in AP direction (0.3 ± 0.2 mm, p < 0.001). So did the mean vector errors at 40 Hz (2.1 ± 0.2 mm vs 1.3 ± 0.2 mm at 80 Hz and 0.9 ± 0.2 mm at 160 Hz, p < 0.05). However, when interference was absent, they were comparable in the both directions and at all sampling frequencies. An accuracy of 0.9 ± 0.2 mm was obtained for the dynamic tracking when using optimum configuration.
CONCLUSIONS: The performance of an EM tracking system depends highly on the system configuration and surrounding environment. The accuracy of EM tracking for catheter reconstruction in a prostate HDR brachytherapy procedure can be improved by reducing interference from surrounding equipment, decreasing distance from transmitter to tracking area, and choosing appropriated sampling frequency. A calibration scheme is needed to further reduce the tracking error when the interference is high.