ESTRO 37 Abstract book

S991

ESTRO 37

Purpose or Objective Nuclear medicine techniques are used for diagnostic and therapeutic purposes and make up about 25% of the per capita effective dose from medical sources in the United States (NCRP Report 160, 2006). Existing methods for performing nuclear medicine dosimetry are mostly based on simplified human anatomical models developed in the 1980s which approximate the human body using geometric shapes such as cylinders and spheres. Newer dosimetry methods based on more realistic anatomical models have since been developed, but are not appropriate for dosimetry in large-scale epidemiological investigations. The over-simplified anatomy in the old models may cause significant dosimetric error for real patient anatomies. We developed a new computational tool to overcome these shortcomings. Material and Methods We adopted the most advanced type of human models, the pediatric and adult hybrid computational phantom series (Lee, 2010). Comprehensive energy absorption factors, called Specific Absorbed Fractions (SAFs), were established for various energies of photons and electrons ranging from 10 keV to 10 MeV, and for more than 50 source and target organs, by using a general-purpose Monte Carlo radiation transport code, MCNPX. Organ dose factors, called S values, were then derived from the SAFs for over 300 radionuclides commonly used in nuclear medicine procedures (ICRP Publication 107). A graphical user interface-based computer program, named National Cancer Institute dosimetry system for Nuclear Medicine (NCINM) was finally developed using MATLAB (Figure 1). We compared our photon SAFs with those from OLINDA (Stabin, 2005) – a widely used tool for nuclear medicine dosimetry based on the old human models.

The insertion of the immobilizer as an attenuator in the calculation of doses is necessary in SRS treatments, since it can cause deviations around 2.5% of doses. The inclusion of the immobilizer in the body structure is the simplest (it does not involve contouring new structures manually) and it is considered the most accurate option. EP-1834 EPID calibration assessments and implications for transit dosimetry J. Valera Lopez 1 , S. Savva 1 , N. Whilde 1 , S. Duggleby 1 1 Northampton General Hospital NHS Trust, Medical Physics, Northampton, United Kingdom Purpose or Objective In radiotherapy, the use of amorphous silicon (aSi) electronic portal imaging device (EPID) has been extended to obtain dosimetric information specifically for pre-treatment and in-vivo dosimetry verification. In our department, verifications performed during treatment are implemented using transit dosimetry with EPIgray (DOSIsoft), in which the predicted dose from Eclipse (Varian) treatment planning system (TPS) is compared with the reconstructed dose from the acquired integrated EPID images. In order for EPIgray to be effective clinically, it is important to maintain an accurate dosimetry and imager calibration of the detector. This study is focused on checking the uniformity of response of the EPID (DMI panel on a TrueBeam) across the entire panel; investigating stability, consistency of the performance, and understanding the implications for the The uniformity across the detector was investigated by comparing its response at specific positions relative to the central position. Integrated images were acquired over the entire panel and measured EPID doses were back-projected into solid water using EPIgray. Calibrated unit (CU) values were also obtained using Portal Dosimetry (Varian). The parameters studied include reproducibility, linearity, response to reference beam, signal uniformity across the panel and image profiles as a function of CU. The EPID measurements were also compared with ionisation chamber measurements to account for stability of the TrueBeam (Varian) outputs, which are necessary to performed the dosimetry calibration of the detector. Results High intensity CU values have been observed in the central region of the panel which gradually decrease toward the corners. Images profiles acquired at 100MU at the edge of the detector had gradients in measured values of up to -13% relative to the central pixel value obtained from the reference image. Therefore, EPIgray verifications are failing for those off-centre integrated images, and the relative dose deviation is affected for more than 5% over the tolerance values (+/-7%) configured in the software from the commissioning. Conclusion We have seen, in a clinical context, out-of-tolerance measurements for some off-centre images acquired during treatment of a patient. The results obtained imply that it is no practicable to utilise the entire area of the panel for acquiring integrated images necessary for transit dosimetry. As it is not always possible in a clinical setting to offset the imager such that the image is centrally placed on the imager, limits have been set on the maximum distance from the centre of the imager that useful information can be acquired. EP-1835 A novel dosimetry tool for epidemiological studies of nuclear medicine patients D. Villoing 1 , C. Lee 1 clinical use of EPIgray. Material and Methods

Figure 1. Flowchart of the organ dose calculation algorithm in the NCINM program Results Ratios of photon SAFs from NCINM to the values from OLINDA were calculated for 5 phantoms (1-y-o male, 5-y- o female, 10-y-o male, 15-y-o female and adult-male), 8 photon energies, and 7 different source/target couples including cases of self-absorption (brain and thyroid). The results varied from 0.27 to 2.97 depending on the photon energy and the source/target combinations. Figure 2 illustrates average ratios between NCINM and OLINDA over these five different models, in four different source/target combinations.

1 National Cancer Institute, Division of Cancer Epidemiology and Genetics, Rockville, USA