Clinical evaluation of an intra-modality 3D ultrasound IGRT system

In this study prostate localization data is analyzed in order to evaluate the clinical usage of an intramodality 3D ultrasound Image Guided Radiotherapy (IGRT) system. Rigorous phantom-based quality assurance procedures ensure the accuracy of the system, reproducing the manufacturer stated alignment tolerances of ±1 mm and ±2 mm in the CT simulator and treatment room, respectively. The analysis is based on daily couch shift data which give the required alignment between the pre-treatment prostate and its nominal treatment position along the anterior-posterior (AP), right-left (RL), and superior-inferior (SI) directions. A total population of 927 couch shifts are considered, sampled from 28 prostate patients who have completed image guided radiotherapy treatment since the clinical introduction of the US system in September 2008. Systematic and random components of the displacements are computed as mean values and standard deviations, respectively. Mean shift values and standard deviations for the entire data population are calculated to 0.0 ± 4.4 mm AP, 0.9 ± 3.7 mm RL, and 0.2 ± 3.7 mm SI. The deviations from a zero mean value are well within above stated tolerance of ± 2 mm (treatment room), indicating clinically insignificant overall systematic errors in the IGRT setup and alignment procedure. However, individual patients show significant systematic misalignment errors of up to 6.1 mm predominantly in the RL as well as AP directions. Potential sources of the systematic errors are discussed. Variations in prostate movement pattern between treatment fractions of less and greater than 50Gy are well within the system localization uncertainties, although an anterior shift of 0.8 mm is statistically significant for doses in excess of 50 Gy.