- Email: [email protected]
461 Analytical model of electronic portal imaging device response
Posters To asses the reliability of the EPID for constancy checks, portal images were taken during a three months period for a fixed setup: 20x20cm field size and a source-detectordistance of 140cm.The registered variations of flatness and symmetry were analysed and compared with the obtained ones using an ion chamber in an automatic waterphantom and reference conditions. Results: The pixel value-dose rate relationship was of the type square-root for both energies. The EPID stability is proven to be better than 2% (1 standard deviation, 1SD) during a three months interval. The 1SD variations over a quarter of flatness and symmetry measured with EPID for the 6MV (18MV) photon quality were of 0.33% (0.26%) and 0.30% (0.25%), respectively. The corresponding variations found using a waterphantom were 0.21% (0.26%) and 0.20% (0.11%). Conclusions: The daily variations detected using the EPID for the parameters flatness, symmetry and output and theirs comparisons with the values obtained from a waterphantom, show that the portal device is suitable for constancy measurements of such parameters for a 6 and 18MV photon beams.
460 Relative electron dosimetry using the Scanditronix Beam I m a g i n g System 2G K. Nyqaard, O. Odland, L.P. Muren Haukeland University Hospital, Dept of Oncology and Medical Physics, Bergen, Norway Purpose: The Beam Imaging System 2G (BIS-2G) from Scanditronix-WellhSfer is a two-dimensional COD-camera which measures the scintillation light produced by incident radiation in a plane exposed to the beam. We examined the performance of the BIS-2G as a tool for investigating dose variations caused by bolus edges during electron irradiation. This was done by comparison with corresponding measurements using a diode in a water phantom. Next, the BIS-2G was used to measure the dose distribution below patient boluses used in the clinic. In an attempt to simplify the production of the patient boluses, the dose distributions below bolus edges built as staircases were measured and compared to the dose profiles below corresponding sloped bolus edges. Methods and Materials: Perspex plates covering half the irradiated fields were used as bolus material for the sloped and the staircase bolus edges. All BIS-2G measurements were performed using build-up of solid water while the diode measured the dose profiles in two depths, 2 and 5 cm, in a water phantom. The patient boluses were measured with 5 mm build-up and regions with doses <95% and >107% of the prescribed dose (PD) were defined as cold and hot spots, respectively. Results: In the region below the edge, the relative doses measured by the BIS-2G were within 3% in dose and 3 mm in position compared to the diode measurements. Close to the penumbra region below the bolus, the BIS°2G measurements were in some cases as much as 7% lower in dose than the diode measurements. The BIS-2G revealed hotspots below the patient boluses covering 1-16% of the total irradiated area. The highest point dose measured below the patient boluses ranged from 105% to 125% of the PD, and the cold spots were in one case as low as 84% of the PD. The cold spots were typically found near the field edge, probably due to the fact that the BIS-2G underestimates the dose at the field edges due to lack of lateral and back scatter. Each edge in the staircase bolus edges caused a fluctuation in dose and increased the maximum dose compared to the sloped edge for all bolus thicknesses. For several cases, the maximum dose increased with 13% in relative dose, e.g. from 103% to 116%. Conclusions: The dose variations below the bolus edges were measured with acceptable accuracy for clinical purposes using the BIS-2G. The BIS-2G was found to be a useful tool
S199 in quality control of patient boluses, revealing large hotspots below the edges in several patients. Bolus edges built as staircases instead of sloped edges cause considerable dose fluctuations and thus increase the maximum dose, and therefore cannot be recommended.
461 Analytical model of electronic portal imaging device response E. Harris, P. Evans, J. Seco Institute of Cancer Research, Joint Department of Physics, Sutton, UK This work is concerned with variation of the output of flat panel electronic portal imaging devices (EPIDs) as a function of incident x-ray energy and the thickness of the copper build-up and Gd202S layers. The aim is to develop a model of the EPID in order to understand the transfer of signal within the imaging system and hence predict system response. The model is derived from a combination of analytical techniques based on fundamental physics principles and electron ranges calculated using Monte Carlo (MC) modelling. The first part of the model of the EPID looks at the transport of electrons through the Cu build-up layer and the deposition of energy in the Gd202S layer. The model uses published physical data to predict the number of secondary electrons generated through photoelectric (PE), Compton and pair production (PP) x-ray interactions in both the copper and the scintillator. Electron transport for Compton electrons is then modelled on a macroscopic scale by sampling their energy and angular probability distributions. To find the energy deposited as a function of depth in the copper and Gd202S layers Compton electrons are then transported by their range. PE and PP electrons are treated in a similar manner, however for PE electrons their kinetic energy is equal to the incoming x-ray energy (binding energy losses are ignored) and for PP electrons their angular distributions are sampled isotropically. To calculate the amount of energy deposited at a given depth within the Gd202S it is assumed that electrons lose energy at a constant rate along their path. Two different electron ranges were used: the continuous slowing down approximation (CSDA) range and electron ranges that were calculated from MC simulation of depth dose curves for energetic electrons in both Cu and Gd202S. To verify the model, the EPID response was modelled using MC. Results calculated using the CSDA range showed large overestimations of energy deposited in the Gd202S. The CSDA range does not take into account multiple scattering of electrons and therefore more secondary electrons are transported through the copper into the Gd202S layer where they deposit energy. Using the electron ranges calculated from MC simulation the analytical model has been used successfully to predict (within + / - 5%) the energy deposited as a function of incident x-ray energy (0 to 6MeV) in a single layer of Gd202S and an EPID comprising 1ram of Cu and 0.345mm of Gd202S. Work is now being undertaken to model the EPID response as a function of Cu and Gd202S thickness. This model will be used to predict the response of the EPID to spectral changes and to optimise EPID design through an understanding of the signal transfer at different stages in the imaging system. Future work will extend the model to include transport of optical photons through the Gd202S layer and their detection in the flat panel detector.
462 Gafchromic EBT film dosimetry outside a multi-leaf collimated beam C. Fiandra, R. Ragona, U. Ricardi Radioterapia, Universit~ di Torino, Turfn, Italy Purpose: Multileaf collimator is widely used in clinical practice to conform isodose distribution to the tumor, avoiding organ at risk. The purpose of this study was to investigate dose outside a multileaf collimated beam in a
460 Relative electron dosimetry using the Scanditronix Beam I m a g i n g System 2G K. Nyqaard, O. Odland, L.P. Muren Haukeland University Hospital, Dept of Oncology and Medical Physics, Bergen, Norway Purpose: The Beam Imaging System 2G (BIS-2G) from Scanditronix-WellhSfer is a two-dimensional COD-camera which measures the scintillation light produced by incident radiation in a plane exposed to the beam. We examined the performance of the BIS-2G as a tool for investigating dose variations caused by bolus edges during electron irradiation. This was done by comparison with corresponding measurements using a diode in a water phantom. Next, the BIS-2G was used to measure the dose distribution below patient boluses used in the clinic. In an attempt to simplify the production of the patient boluses, the dose distributions below bolus edges built as staircases were measured and compared to the dose profiles below corresponding sloped bolus edges. Methods and Materials: Perspex plates covering half the irradiated fields were used as bolus material for the sloped and the staircase bolus edges. All BIS-2G measurements were performed using build-up of solid water while the diode measured the dose profiles in two depths, 2 and 5 cm, in a water phantom. The patient boluses were measured with 5 mm build-up and regions with doses <95% and >107% of the prescribed dose (PD) were defined as cold and hot spots, respectively. Results: In the region below the edge, the relative doses measured by the BIS-2G were within 3% in dose and 3 mm in position compared to the diode measurements. Close to the penumbra region below the bolus, the BIS°2G measurements were in some cases as much as 7% lower in dose than the diode measurements. The BIS-2G revealed hotspots below the patient boluses covering 1-16% of the total irradiated area. The highest point dose measured below the patient boluses ranged from 105% to 125% of the PD, and the cold spots were in one case as low as 84% of the PD. The cold spots were typically found near the field edge, probably due to the fact that the BIS-2G underestimates the dose at the field edges due to lack of lateral and back scatter. Each edge in the staircase bolus edges caused a fluctuation in dose and increased the maximum dose compared to the sloped edge for all bolus thicknesses. For several cases, the maximum dose increased with 13% in relative dose, e.g. from 103% to 116%. Conclusions: The dose variations below the bolus edges were measured with acceptable accuracy for clinical purposes using the BIS-2G. The BIS-2G was found to be a useful tool
S199 in quality control of patient boluses, revealing large hotspots below the edges in several patients. Bolus edges built as staircases instead of sloped edges cause considerable dose fluctuations and thus increase the maximum dose, and therefore cannot be recommended.
461 Analytical model of electronic portal imaging device response E. Harris, P. Evans, J. Seco Institute of Cancer Research, Joint Department of Physics, Sutton, UK This work is concerned with variation of the output of flat panel electronic portal imaging devices (EPIDs) as a function of incident x-ray energy and the thickness of the copper build-up and Gd202S layers. The aim is to develop a model of the EPID in order to understand the transfer of signal within the imaging system and hence predict system response. The model is derived from a combination of analytical techniques based on fundamental physics principles and electron ranges calculated using Monte Carlo (MC) modelling. The first part of the model of the EPID looks at the transport of electrons through the Cu build-up layer and the deposition of energy in the Gd202S layer. The model uses published physical data to predict the number of secondary electrons generated through photoelectric (PE), Compton and pair production (PP) x-ray interactions in both the copper and the scintillator. Electron transport for Compton electrons is then modelled on a macroscopic scale by sampling their energy and angular probability distributions. To find the energy deposited as a function of depth in the copper and Gd202S layers Compton electrons are then transported by their range. PE and PP electrons are treated in a similar manner, however for PE electrons their kinetic energy is equal to the incoming x-ray energy (binding energy losses are ignored) and for PP electrons their angular distributions are sampled isotropically. To calculate the amount of energy deposited at a given depth within the Gd202S it is assumed that electrons lose energy at a constant rate along their path. Two different electron ranges were used: the continuous slowing down approximation (CSDA) range and electron ranges that were calculated from MC simulation of depth dose curves for energetic electrons in both Cu and Gd202S. To verify the model, the EPID response was modelled using MC. Results calculated using the CSDA range showed large overestimations of energy deposited in the Gd202S. The CSDA range does not take into account multiple scattering of electrons and therefore more secondary electrons are transported through the copper into the Gd202S layer where they deposit energy. Using the electron ranges calculated from MC simulation the analytical model has been used successfully to predict (within + / - 5%) the energy deposited as a function of incident x-ray energy (0 to 6MeV) in a single layer of Gd202S and an EPID comprising 1ram of Cu and 0.345mm of Gd202S. Work is now being undertaken to model the EPID response as a function of Cu and Gd202S thickness. This model will be used to predict the response of the EPID to spectral changes and to optimise EPID design through an understanding of the signal transfer at different stages in the imaging system. Future work will extend the model to include transport of optical photons through the Gd202S layer and their detection in the flat panel detector.
462 Gafchromic EBT film dosimetry outside a multi-leaf collimated beam C. Fiandra, R. Ragona, U. Ricardi Radioterapia, Universit~ di Torino, Turfn, Italy Purpose: Multileaf collimator is widely used in clinical practice to conform isodose distribution to the tumor, avoiding organ at risk. The purpose of this study was to investigate dose outside a multileaf collimated beam in a