Measurement of setup deviation in IMRT head and neck patients using an on-board kilovoltage imager

Abstract

The purpose of this study was to measure interfractional and intrafractional setup deviations, using a kilovoltage imaging device, in patients with head and neck or brain cancers receiving IMRT (intensity-modulated radiotherapy) treatment. Seven patients were imaged approximately 3 times per week, before and after IMRT treatment, using the Varian On-Board Imager (OBI), a kilovoltage imaging device permanently mounted on the gantry of a Varian 21EX machine (Varian Medical Systems, Palo Alto California).   Due to commissioning of the remote couch correction of the OBI during this study, online setup corrections were performed on 2 patients, and for the other 5 patients weekly corrections were made based on a sliding average of the measured data.  From these data, we determined the interfractional setup deviation, defined as the shift from the original set-up position suggested by the daily image, the residual error associated with the weekly correction protocol, as well as the intrafractional setup deviation, defined as the difference between the post-treatment and pre-treatment images.   Interfractional and intrafractional rotational deviations were also determined from the images using our own image-registration software based on the template matching method.  In addition, we evaluated the influence of inter-observer variation on our results, and whether the use of different registration techniques introduced differences.  Finally, translational data was compared with rotational data to search for correlations.Translational setup errors from all the data were 0.0±0.2, -0.1±0.3, and -0.2±0.3 cm in the right-left (RL), anterior-posterior (AP), and superior-inferior (SI) directions, respectively.  Residual error for the 5 patients with a weekly correction protocol was -0.1±0.2 (RL), 0.0±0.3 (AP), and 0.0±0.2 (SI). Intrafractional translation errors were small, amounting to 0.0±0.1, -0.1±0.2, and 0.0±0.1 cm in the RL, AP, and SI directions, respectively.  Inter-fractional rotational errors were -1.1±1.7 and -0.5±0.9 degrees, and intrafractional rotational error 0.3±0.6 and 0.2±0.5 degrees, in the sagittal and coronal views respectively.  No significant correlation was seen between translational and rotational data.OBI image data was used to study setup error in the head and neck.  Nonzero systematic errors were seen in the interfractional translational and rotational data but not in the intrafractional data indicating that the mask is better at maintaining head position than in reproducing it.