Supervisors info:
Παντελής Καραΐσκος, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Μιχάλης Μαζωνάκης, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, Πανεπιστήμιο Κρήτης
Δημήτριος Θανασάς, Ε.Τ.Ε.Π., Ιατρική Σχολή, ΕΚΠΑ
Summary:
Purpose: The purpose of this study is twofold: firstly, to calibrate the thermoluminescence dosimeter (TLD) system available at the Medical Physics Laboratory of the Medical School of Athens, and secondly, to use the TLDs (LiF-100) to verify a stereotactic radiotherapy plan.
Materials and Methods: For the calibration, we used a batch of LiF-100 dosimeters, each with cubic geometry and dimensions of 1 mm³. The equipment included a TLD reader (Victoreen 2800) and a specialized furnace from PTW. Initially, five irradiations of the batch were performed on a solid water phantom using a linear accelerator to calculate the Element Correction Sensitivity Factor (ks,i) and the reproducibility of each TLD. The TLDs were then grouped and irradiated with a range of nominal doses from 0.25 to 12 Gy, constructing dose-charge curves. In the second part of the study, the TLDs were placed in a humanoid phantom simulating the head, using a specially designed insert. The phantom included three spherical targets, which were irradiated with an SRS technique using 4 non-coplanar arcs with a single isocenter, executed by a Varian linear accelerator. The target under evaluation was PTV1, with a diameter of less than 2 cm, and CI_Paddick and GI_Paddick indices were found 0.88 and 2.83, respectively. Prescription dose was 27Gy in 3 fractions, while one fraction was delivered. The method applied to verify the plan was the gamma index, calculated both locally and globally with various dose-distance criteria: 7% / 2mm, 5% / 2mm, and 7% / 1.5mm.
Results: Analyzing the calibration results, the mean ks,i value for the batch was found to be 1.004, while the mean reproducibility of the system, quantified via the coefficient of variation (CV), was 3.27%. Two dose-charge graphs (linear and quadratic fit) were constructed, both with R² > 0.99. The results of the comparison of the two distributions using the gamma index showed very high passing rates, with the linear model outperforming the quadratic model for all dose-difference distance agreement criteria used, both locally and globally. The highest passing rate (96%) was observed in the case of 7% / 2mm globally, while the passing rates decreased as the criteria became stricter.
Conclusions: The mean reproducibility of the system contrasts with the literature, which reports values < 1%, thus imposing a limitation on the accuracy of our results. However, the results were satisfactory, especially in the case of the global gamma index, where the passing rate was almost always above 90%. Thus, LiF-100 dosimeters are deemed ideal for end-to-end verification SRS/SBRT techniques.
Keywords:
Radiotherapy, TLD, LiF-100, SRS, SBRT, Gamma index
