A light-guide optimization for proof-of-principle of a megavoltage orthogonal ray imaging prototype

Abstract

Radiotherapy (RT) is nowadays, after surgery, the most frequently used cancer treatment. Modern RT techniques provide increasingly higher conformality, a potential invaluable clinical benefit to the patient. Consequently, patient misalignments and changing internal anatomy (e.g. tissue swelling, edema, inflammation or tumor shrinkage/growth) are also becoming more critical since higher conformality may equally represent a higher risk of target underdosage or organ-at-risk overdosage. Therefore, state-of-the-art image-guided radiotherapy (IGRT) is the modern technique that aims at providing feedback to the radiation oncologist in regard to these matters, sometimes at the cost of increased dosage or treatment fraction time (e.g. kilo and megavoltage cone-beam computed tomography), other times providing insufficient clinical information. Our group is investigating a novel imaging system especially designed for assisting RT treatments. Such system, termed OrthoCT, consists in operating a dedicated X-ray detection system specially built for collecting selected patient-scattered radiation. Our team has shown by simulation [1, 2] and experiments [2, 3] that collecting such radiation allows for a rotation-free, 3D imaging of the inner morphology of the target (patient). The rotation-free and low dose 3D imaging capability of OrthoCT renders it very attractive due to its usefulness for so-called “on-board” patient imaging. A small-scale, bi-dimensional system is under construction for proof-of-principle validation. It consists of 4 crystals lines separated by collimator slices. Each crystal line is composed by 50 gadolinium oxyorthosilicate (GSO) crystals with a front-area of 4 × 4 mm2, forming a total line length of 200 mm. As light detector, a photomultiplier tube (PMT) will be used. Since the sensitive area of the PMT is 50 × 50 mm2, a light-guide to drive the light between the GSO crystals and the PMT is required. The aim of this work was the optimization, by means of Monte Carlo simulation, of the light-guide dimensions. The setup implemented into Geant4 is shown schematically in Figure 1, left. Such apparatus was tested for different theta angles (i.e. different light-guide length L). In Figure 1, right the count profiles obtained for different theta values are presented. The length L corresponding to θ = 7.5, 10, 12.5, and 15 degrees was 570, 425, 338, and 280 mm, respectively. Thus, smaller angles require larger light-guide lengths and, consequently, light losses in the path length are also higher. By analyzing the plots, the profile obtained for θ = 12.5 degrees shows a great compromise between the fraction of photons detected and the guide-light dimensions. For angles θ > 15 degrees the detection of the light coming from peripheric crystals is compromised. In concluding, the length of the guide-light must be approximately 340 mm.