ABSTRACT
In order to assign appropriate planning target volume [PTV] margins, each centre should measure the patient positioning deviations for their set-up techniques. At the Royal Marsden Hospital, UK, a conformal shell [cast] system is used when a stereotactic frame is not suitable. In this paper, we report on a series of measurements with the aim of obtaining the systematic and random components of positioning error when using the above-mentioned shell system. The verification protocol was based on orthogonal pairs of anterior-posterior and lateral electronic portal images [EPIs] used to check the isocentre position. The isocentre verification results of paediatric patients were analysed. A practical [off-line] patient set-up correction strategy had been used with the aim of reducing systematic errors. The verification protocol involved EPI acquisition on the first three fractions and then on a weekly basis. Additional images were taken if an isocentre movement was applied based on a 3 mm tolerance for a consistent 1D discrepancy. Four patients required isocentre corrections ranging between 2 mm and 4 mm. Following the offline corrections, the residual systematic errors in each direction were within 0.5 mm while the 1D random variation was about 1.0 mm. The head fixation system in conjunction with the correction strategy successfully kept the random and systematic positioning errors within an acceptable level well within the 3 mm tolerance. The measured components of positioning error can be used to define appropriate PTV margins