Mohammad Mohammadi

Evaluation of MLC leaf positioning using a scanning liquid ionization chamber EPID

A method was developed to determine the accuracy of multileaf collimator (MLC) positioning using transmitted dose maps measured by a scanning liquid ionization chamber electronic portal imaging device (SLIC-EPID). Several MLC fields were designed, using the Varian C-series standard MLC-80, as reference fields for open fields. The MLC leaves were then shifted from the reference positions along the direction of MLC leaf movement towards the central axis from 0.1 to 1.6 mm. The electronic portal images (EPIs), acquired for each case, were converted to two-dimensional dose maps using an appropriate calibration method and the relative dose difference maps were then calculated. The experiment was then performed at non-zero gantry angles in the presence of an anthropomorphic phantom for typical prostate and head and neck fields. Several standard edge detection algorithms were also used in order to find the shifted MLC leaf position. In addition, the short-term reproducibility of MLC leaf positioning was evaluated using the above-mentioned methods. It was found that the relationship between the relative dose difference and MLC leaf spatial displacement is linear. A variation of 0.2 mm in leaf position leads to approximately 4% change in the relative dose values for open fields. The variation of the relative dose difference for phantom studies depends on the phantom positioning and the EPI normalization. From the standard edge detection algorithms, used in the current study, the Canny algorithm was found to be the optimum method to identify the minimum detectable MLC leaf displacements with a precision of approximately 0.1 mm for all cases. However, the result of edge detection algorithms generally is binary and there is no additional information compared to the relative dose maps. The reproducibility of MLC positions was found to be within 0.3 mm. In conclusion, a SLIC-EPID can be used for regular quality assurance (QA) of MLC leaf positioning. Despite significant difference in the pixel size of the acquired SLIC-EPIs, it can be concluded that the SLIC-EPID can be used for MLC quality assurance protocols with similar accuracy compared to amorphous silicon (a-Si) EPID results.

Two-dimensional transmitted dose measurements using a scanning liquid ionization chamber EPID

The use of a scanning liquid ionization chamber electronic portal imaging device (SLIC-EPID) for two-dimensional transmitted dosimetry was investigated and a calibration method was developed using extended dose range (EDR2) film. In order to convert pixel value to dose, the acquired SLIC-EPID pixel values were calibrated using an ionization chamber on the central axis. The relationship between pixel values, dose rate and absorbed dose was identified for various linac output repetition rates. To correct EPIs for dosimetric purposes, the off-axis ratio of dose profiles measured by EPIDs and EDR2 film was used to derive correction factor matrices (CFMs) for a range of source-to-EPID distances (SEDs). The corrected relative dose maps acquired for different conditions, including open and wedged fields, measured using a SLIC-EPID were compared with EDR2 film images using a gamma function algorithm with distance to agreement (DTA) = 2.5 mm and dose difference (DeltaDmax) = 1% criteria. The results showed that (a) for two-dimensional dosimetric purposes, EPIDs must be calibrated using appropriate two-dimensional correction factors and (b) SLIC-EPIDs can be used to measure the transmitted dose with good accuracy.

The use of extended dose range film for dosimetric calibration of a scanning liquid-filled ionization chamber electronic portal imaging device

A scanning liquid‐filled ionization chamber electronic portal imaging device (SLIC‐EPID) and extended dose range (EDR2) films were used to evaluate transmitted dose profiles for homogeneous and inhomogeneous phantoms. Calibrated ionization chamber measurements were used to convert the pixel values acquired from the electronic portal images to dose. Because SLIC‐EPID was developed to have a uniform response for all liquid ionization chambers, the off‐axis dose values were reconstructed using a correction factor matrix, defined as the ratio of the relative EDR2 film and the corresponding EPID dose values measured in air. The transmitted dose distributions in the EPID detector layer were also modeled using a Pinnacle3 treatment planning system (TPS: Philips Radiation Oncology Systems, Milpitas, CA). The gamma function algorithm was then used to assess agreement between transmitted dose distributions measured using a SLIC‐EPID and EDR2 film, and those calculated using the TPS. For homogenous and inhomogeneous phantoms, more than 90% agreement was achieved using gamma criteria of 2% and 3 mm and 3% and 2.5 mm respectively. Our results indicate that the calibration procedure proposed in the present study should be performed if SLIC‐EPID is to be used as a reliable two‐dimensional transmitted dosimeter for clinical purposes. PACS numbers: 87.53.Tf, 87.53.Oq

Comparison of two-dimensional transmitted dose maps: evaluation of existing algorithms

Composite analysis and the gamma function are often used to assess the agreement between a reference and an evaluated two-dimensional dose maps. The intent of the study is to compare advantages, disadvantages and limitations of dose evaluation tools reported in the literature. In addition, in order to improve the gamma function output, a “Signed Matrix” was introduced using the ratio of relative dose difference maps. Transmitted dose maps were acquired for a range of homogeneous phantoms using Extended Dose Range (EDR2) films and a Scanning Liquid Ionization Chamber Electronic Portal Imaging Device (SLIC-EPID). For inhomogeneous case, the transmitted dose maps were obtained from EDR2 films measurement and a Treatment Planning System (TPS). The corresponding dose maps were compared based on composite and gamma function algorithms. The results showed that the agreement between reference and evaluated dose maps for the composite analysis were generally greater than those obtained using the gamma function. For homogeneous phantom comparison, the difference between the agreeing fractions calculated using composite analysis and gamma function increases with the increase of phantom thickness for ΔD = 0.5% and 1%. For inhomogeneous cases, a significant difference (≈ 5% for ΔD = 1.5%) was observed between the percentage agreement as calculated by composite and gamma function techniques. The concept of the composite model is closer than gamma function to the idea of the two-dimensional dose verification protocol proposed originally by van Dyk. However, the composite model results only display the passed or failed regions in the dose maps. On the other hand, the gamma function provides continuos information by distinguishing the points within each region. The overdosed/underdosed regions (the ratio of reference and evaluated doses at a given point) and the direction of the misalignment can be recognized with the enhanced gamma map convolved with a “Signed Matrix”.

Investigation of effect of reconstruction filters on cone-beam computed tomography image quality

Cone-beam CT (CBCT) using kV imagers integrated with linear accelerators is now widely used in verifying patient position during radiation therapy. Current CBCT acquisition protocols have lowered tube current to keep the imaging dose to a minimum. This affects the usability of CBCT data sets in treatment planning by reducing the soft tissue contrast and accuracy of CT numbers (Hounsfield values). The purpose of this study is to investigate the effect of reconstruction filters on full-fan and half-fan acquisition modes of CBCT and assess the image quality parameters of contrast- to -noise ratio, spatial resolution, pixel stability and uniformity. The results of this study show the relation between the noise and resolution of a CBCT image by using different reconstruction filters and provide possible estimations of the impact of filters on image quality and subsequent optimization for image-guided radiotherapy purposes.

Optimal light collection in BeO fiber optic dosimetry

The use of BeO ceramic in fiber optic dosimetry is investigated and also new probe setups for optimizing the light collection, where results show that they are more efficient than previously investigated methods.