[Assessment of the accuracy of position in breast irradiation with isocentric technique using an electronic system to acquire portal images].

This study investigated the accuracy of daily patient positioning in an isocentric set-up for breast irradiation. This was achieved through the assessment of both systematic and random errors measured as discrepancies between reference simulation images and portal images obtained with an on-line electronic portal imaging device (EPID). To this end, 10 portal images for each of the tangential fields were obtained in 12 consecutive patients and the images were compared to reference simulation films with dedicated software tools provided with the EPID. The discrepancies measured for each set of images were analyzed statistically. Most variations in each series both in terms of cranio-caudal and of lateral displacement appeared to be random, with mean standard deviations of 2.3 and 2.8 mm, respectively. Such variations between reference fields and portal images tend to counteract one another and do not usually play a significant role in the overall accuracy of patient set-up. In some cases, however, the distribution of the variations occurred in a well-defined pattern indicating a systematic error in patient positioning. This study shows that our set-up technique for breast irradiation is relatively accurate and reproducible, and most of the observed errors were below the 5 mm cut-off level, but it stresses the need of accurate quality assurance programs in radiotherapy. Electronic portal imaging devices are excellent tools for fast portal image acquisition, and permit the accurate assessment of discrepancies with respect to reference images.