ABSTRACT
The aim of this study is to validate the electron Monte Carlo module implemented in XiO, a treatment planning system commercialized by Elekta CMS inc. Two types of phantoms were investigated: homogeneous water phantoms with irregular surfaces and phantoms containing slab and 3D heterogeneities. The phantoms were CT scanned, and dose to water calculations were performed in the eMC module using 2 ×2 × 2 mm2 voxels and a mean relative statistical uncertainty of 0.5%. Concurrently, Gafchromic EBT3 film measurements were performed in the same phantoms. To obtain reliable absolute dose readings from the films, a new method using triple channel dosimetry in the Film QA Pro software was developed. The accuracy of the proposed method was determined empirically and an uncertainty of ±1.5% was found over the range [75, 800] cGy. Dose comparisons between film and simulations were done using an in-house MATLAB program. XiO's eMC module provides accurate dose distributions in the presence of surface irregularities and slab heterogeneities for 12 MeV beams. In the presence of 3D heterogeneities, the percent dose difference comparisons highlighted the need to perform 3D gamma comparisons. In conclusion, the electron Monte Carlo module offered in the XiO treatment planning system is promising and could greatly improve the accuracy of clinical dose calculations. The validation of the software is ongoing, notably concerning more complex phantom geometries. Small field calculations, oblique incidences and cutout factors will also be investigated.
ABSTRACT
A newly acquired nanoDot In-Light system was compared with TLD-100 dosimeters to confirm the treatment dose in the multiple cases: an electron eye treatment, H&N IMRT and VMAT validation for small targets. Eye tumour treatment with 9 MeV electrons A dose of 1.8 Gy per fraction was prescribed to the 85% isodose. The average dose measured by three TLDs and three Dots was 1.90 and 1.97 Gy. Both detectors overestimated dose, by 2.9% and 6.7% respectively. H&N IMRT treatment of skin cancer with 6 MV photons Dose per fraction is 2.5 Gy. The average doses measured by two TLDs and two Dots were 2.48 and 2.56 Gy, which represent errors of -0.8% and 2.2%, respectively. VMAT validation for small targets using an Agarose phantom, dose 15 Gy A single-tumour brain treatment was delivered using two coplanar arcs to an Agarise phantom containing a large plastic insert holding 3 nanoDots and 4 TLDs. The difference between the average Pinnacle dose and the average dose of the corresponding detectors was -0.6% for Dots and -1.7% for TLDs. A two-tumour brain treatment was delivered using three non-coplanar arcs. Small and large plastic inserts separated by 5 cm were used to validate the dose. The difference between the average Pinnacle dose and the average dose of the corresponding detectors was the following; small phantom 0.7% for Dots and 0.3% for TLDs, large phantom-1.9% for Dots and -0.6% for TLDs. In conclusion, nanoDot detectors are suitable for in-vivo dosimetry with photon and electron beams.
ABSTRACT
To develop a tomosynthesis-based dose assessment procedure that can be performed after an I-125 prostate seed implantation, while the patient is still under anaesthesia on the treatment table. Our seed detection procedure involves the reconstruction of a volume of interest based on the backprojection of 7 seed-only binary images acquired over an angle of 60° with an isocentric imaging system. A binary seed-only volume is generated by a simple thresholding of the volume of interest. Seeds positions are extracted from this volume with a 3D connected component analysis and a statistical classifier that determines the number of seeds in each cluster of connected voxels. A graphical user interface (GUI) allows to visualize the result and to introduce corrections, if needed. A phantom and a clinical study (24 patients) were carried out to validate the technique. A phantom study demonstrated a very good localization accuracy of (0.4+/-0.4) mm when compared to CT-based reconstruction. This leads to dosimetric error on D90 and V100 of respectively 0.5% and 0.1%. In a patient study with an average of 56 seeds per implant, the automatic tomosynthesis-based reconstruction yields a detection rate of 96% of the seeds and less than 1.5% of false-positives. With the help of the GUI, the user can achieve a 100% detection rate in an average of 3 minutes. This technique would allow to identify possible underdosage and to correct it by potentially reimplanting additional seeds. A more uniform dose coverage could then be achieved in LDR prostate brachytherapy.
