SU‐E‐T‐207: Local and Limited Projection Tomography Reconstructions for Contrast‐Enhanced Synchrotron Stereotactic Radiotherapy in Vivo Dosimetry

Abstract

Purpose: Contrast‐enhanced stereotactic synchrotron radiation therapy (SSRT) is an innovative technique based on localized dose‐enhancement effects obtained by reinforced photoelectric absorption in the target. Medium energy monochromatic x‐rays (50 –100 keV) are used for irradiating tumors previously loaded with high‐Z elements. SSRT clinical trials are being prepared at the ESRF. The first patients (scheduled in summer‐autumn 2011) should be treated at 80 keV, with 10 conformational beams. The treatment plan will use CT images acquired prior and after systemic iodine injection. However the exact dose delivered to the patient relies on the exact 3D iodine biodistribution in the tumor during the treatment itself. An in vivo dosimetry methodology based on portal imaging was developed to retrieve the absolute 3D iodine concentrations reached during the treatment. Methods: The proposed algorithm uses the images acquired with a high‐purity germanium detector located behind the patient. The proposed methodology has to solve a tomography reconstruction problem with a restricted field of view (local tomography) and a limited number of projections (10 at maximum). None of the currently available algorithms can solve both local and limited projection tomography problems at once while being quantitative. However, in SSRT, a large amount of a priori information is available (three CT‐scans acquired prior the treatment, with and without iodine). Thus the proposed algorithm proposed in this paper is based on filtered backprojection and uses extensively the a priori information related to patient morphology, with a few geometrical hypothesis. Results: The results obtained from numerical simulation on a human head phantom show that the proposed algorithm is able to retrieve the iodine concentrations with a relative error lower than 2%. Conclusions: The in vivo dosimetry methodology for SSRT has been validated with numerical simulations. It has now to be tested on real patient data before the first patient is treated.