EPID-BASED DOSIMETRY

Abstract

The complexity of radiotherapy today demands thorough verification of the dose delivered to the patient. Dosimetry using electronic portal imaging devices (EPIDs) has received considerable attention recently, since with this method dose verification can be performed accurately and efficiently, both pre-treatment and in vivo. Moreover, when EPIDs are used for the verification of the patient position, they are available for the verification of the dose as well.EPID dosimetry can be performed as transmission or non-transmission dosimetry, i.e. with or without a patient/phantom in the beam. Both methods can be used in the forward or the back-projection/dose reconstruction approach, where the comparison with a prediction algorithm takes place at either the position of the EPID or in the patient/phantom. Although detector technology varies for commercially available EPIDs, three steps can be defined in each EPID-based dose-verification procedure:(1) Prediction: An EPID image, a fluence or dose distribution (1D or 2D) at the position of the EPID or a dose distribution (1D, 2D or 3D) in the patient/phantom is calculated per beam. This prediction is done with a treatment planning system (TPS) or an independent algorithm.(2) EPID measurement and processing: The integrated dose of a segment or field is measured with the EPID. In the forward approach, the EPID image is used as is, or is processed to determine the fluence or dose distribution at the EPID position. For this processing a dosimetric calibration of the EPID is needed, which translates the EPID signal into fluence or dose values. Corrections such as for EPID scatter or crosstalk may be required. These dosimetric calibration data are obtained from measurements with reference detectors in phantoms or from Monte Carlo calculations based on EPID characteristics. In the back-projection/dose reconstruction approach, the EPID-based fluence or dose is used as input for a dose-calculation algorithm to calculate the dose in the patient/phantom. Simple measurement-based approaches, Monte Carlo algorithms or TPSs have been used for this purpose.(3) Comparison of prediction and (processed) EPID measurement: The comparison can be performed at points, in planes or volumes, and dose values at specific points, dose-difference maps or gamma evaluations have been used for this. If the total dose is reconstructed in 3D, a 3D comparison is also possible, e.g. using dose-volume histograms.Reported errors include changed patient anatomy, transfer errors, mal-functioning of leaves, incorrectly set TPS parameters. As with other measurement systems, the EPID calibration has regularly to be monitored and adapted. Users must be aware of implementation-specific limitations, such as (in)dependence of the TPS or how treatment-time anatomy is included in the verification. Decision protocols have to be formulated depending on system accuracy, type of comparison (e.g. 2D per beam or 3D total dose) and clinical requirements. Although EPID dosimetry has been shown to be an efficient tool to detect dose delivery errors, it is not a replacement of patient position verification or regular machine QA. Unfortunately, no system is currently commercially available for in vivo EPID dosimetry. It is hoped that vendors will soon offer such systems that provide a practical safety net for radiotherapy.Ref.: van Elmpt W et al. A literature review of electronic portal imaging for radiotherapy dosimetry. Radiother Oncol 2008;88:289-309