A method to determine the planar dose distributions in patient undergone radiotherapy

A 2D-array equipped with 729 vented plane parallel ion-chambers has been calibrated as a portal dose detector for radiotherapy in vivo measurements. The array has been positioned by a radiographic film stand at 120 cm from the source orthogonal to the radiotherapy beam delivered with the gantry angle at 180°. The collision between the 2D-array and the patient's couch have been avoided. In this work, using the measurements of the portal detector, we present a method to reconstruct the dose variations in the patient treated with step and shoot intensity-modulated beams (IMRT) for head-neck tumours. For this treatment morphological changes often occur during the fractionated therapy. In a first step an in-house software supplied the comparison between the measured portal dose and the one computed by a commercial treatment planning system within the field of view of the computed tomography (CT) scanner. For each patient, the percentage Pγ of chambers, where the comparison is in agreement within a selected acceptance criteria, was determined 8 times. At the first radiotherapy fraction the γ-index analysis supplied Pγ values of about 95%, within acceptance criteria in terms of dose-difference, ΔD, and distance-agreement, Δd, that was equal to 5% and 4 mm, respectively. These acceptance criteria were taken into account for small errors in the patient's set-up reproducibility and for the accuracy of the portal dose calculated by the treatment planning system (TPS) in particular when the beam was attenuated by inhomogeneous tissues and the shape of the head-neck body contours were irregular. During the treatment, some patients showed a reduction of the Pγ below 90% because due to radiotherapy treatment there was a change of the patient's morphology. In a second step a method, based on dosimetric measurements that used standard phantoms, supplied the percentage dose variations in a coronal plane of the patient using the percentage dose variations measured by the 2D-array portal detector. The results showed that the dose variations due to the change of the patient's morphology reached 15% and such discrepancies were displayed on the digitally reconstructed radiography of the patient. The dose discrepancies were confirmed by the hybrid plan obtained by the treatment planning system. The good results here reported show that once it is possible to have the portal dose distributions even for other gantry angles, these tests could be introduced in the clinical protocol to have major support to decide when to repeat the patient's CT scan and to re-plan the new IMRT dose calculation. © 2008 Elsevier B.V. All rights reserved