[39236286]
In order to assure the civco protura robotic couch was working correctly, we came up with a daily quality assurance test that we felt confident would test the entire system. The test involved placing a cube phantom with an offset rotation of -2 degrees in the pitch and roll directions onto the couch at a known translational offset from the treatment isocenter, imaging the phantom with cone-beam ct, performing an image registration to determine the necessary couch translations and rotations in order to correctly bring the cube phantom to treatment isocenter, and then visually inspecting where the treatment isocenter lasers fall on the cube. Our tolerance for this machine, which is used for stereotactic body radiation therapy (sbrt) treatments and therefore requires tighter tolerances, is that all hardware/software involved in the above steps must bring the phantom to within 1mm of the treatment isocenter; this is the tolerance required by the (b)(6) report for stereotactic treatment units. However, when performing the aboe test with translational offsets of between 3-5 cm in all three directions, we found that the end result consistently was that the cube phantom was 2. 5 mm off in the lateral direction and the 1. 5 mm off in the v vertical direction. We performed several tests to try to isolate the problem and submitted the results of those tests to civco for analysis. After civco investigated the issue, they discovered that under certain circumstances - specifically when "split shifts" are enabled, meaning that the varian pedestal is used for translations while the protura robotics are used for rotations - tht protura's rotation algorithm is assuming the rotations determined in the (b)(6) software are performed around the couch position before the shifts instead of after the shifts are applied ) i. E. Original position instead of target position). Because the result of a rotation does depend on the isocenter it is rotating around, rotating around the wrong isocenter results in an incorrect positon - as was seen in the test that we performed. The larger the discrepancy between the before shift and after shift couch positon, the larger the error. At our site we do have patients that occasionally must be scanned offset (sometimes up to 8 cm) from where the treatment isocenter ultimately will be, so this is a real concern. Fortunately, we have not had any such patients since the protura robotic couch went live (and we have been rescanning post-shifts to confirm correct positioning during this time that we have been trying to vet our quality assurance testing of the system). We have discontinued the use of the protura robotic couch until this issue is resolved. This product is a robotic couch used in the radiation oncology department of many hospitals, to position patients for their radiation treatments. After being released for treatment, it became clear that the product was not working correctly. No patients were harmed at our site.
Patient Sequence No: 1, Text Type: D, B5