Investigations on absolute portal dosimetry for the verification of static and dynamic dose delivery in radiotherapy (in English)

Abstract

New in vivo dosimetry procedures using electronic portal imaging devices (EPIDs) either compare measured to predicted transmission dose, or infer patient dose from measured transmission dose images. These procedures demand a careful calibration of the portal imager. The EPID used in this research was a scanning liquid ionization chamber (SLIC). The first investigation was to see if a dose calibration of the EPID, performed on the central axis, is sufficiently accurate to estimate the dose for off‐axis points. The incident photon energy fluence distribution generated from the phase space of a linear accelerator was first convolved with Monte Carlo calculated dose spread kernels in water, which serves as the reference medium, and then in the detector. The kernels in water and the detector differ considerably at photon energies lower than 1 MeV or higher than 6 MeV. Transmission doses were then calculated in the SLIC‐EPID and in water. The dose differences are due to changing beam quality. It is shown that depending on the type of calibration, the difference can be as large as 5%, which is unacceptable for accurate transmission dose measurements. To investigate the applicability of the EPID for dynamic dose delivery techniques, the dose response to a photon beam with variable intensity was measured and theoretically modeled. It was found that the accuracy using the detector for verification of the dose from intensity modulated treatment fields delivered by a dynamically modulated multileaf technique is limited. This is due to the detector's nonlinear response. However, under certain acquisition conditions, the leaf movement during delivery can be precisely verified.