Intra- and inter-fractional analysis of setup errors in the thoracic region using electronic portal imaging

Introduction: Recent studies on patient setup errors performed by our group, as well as others, have shown considerable problems in setting up patients to be treated in the pelvic region using radiation therapy. The extent, incidence and type of errors are well known in this region due to the extensive use of electronic portal imaging devices. Also various strategies for correcting these errors have been developed. In the thoracic region however there is little data available, not in the least due to the difficulty to find appropriate landmarks from which the patients position with respect to the field can be quantified. Methods and Materials: A patient study was initiated to determine the intra- as well as the inter-fractional variation of the patient setup. The treatment consists of standard Anterior-Posterior (AP) fields as well as oblique fields. For the AP-fields stable landmarks were identified and used to perform intra-fractional treatment correction. This allowed us to estimate before and after correction setup errors, finally yielding an estimate of the accuracy with which setup errors could be detected and corrected. For oblique fields only setup errors where measured using overlay techniques. So far 10 patients have been evaluated throughout their treatment consisting of AP as well as oblique fields. Results: For AP-fields the overall standard deviation of setup errors was 4.53mm. While the SD measured after correction amounted to 2.41 mm. Oblique fields had a setup error SD of 4.95mm. Statistical analysis showed that the errors for AP fields where partly attributed to systematic errors and partly to random errors. In the case of oblique fields the systematic errors showed a tendency to dominate. Finally time measurement showed an average increase in treatment time of 63%, when using intra-fractional correction