3D dose distribution in gamma knife treatment near tissue inhomogeneities

The treatment planning system, GammaPlan, uses CT, MR or angiographic images to calculate and simulate the dose distribution in a matrix volume of interest assuming that tissues in human head are homogeneous and water equivalent. The absence of electronic equilibrium in the vicinity of air-tissue inhomogeneity in the head will misrepresent the deposited dose under the above assumption. Polymer gel dosimetry has already been used in different scenarios of radiotherapy dosimetry; however, little work has been reported for polymer gel phantoms with air cavities irradiated in Gamma Knife surgery. Increasing dose levels are reflected into lower MR relaxation time constants T1 and T2, in the neighbouring water protons. The MAGIC Gel was manufactured under normal atmospheric conditions using the formulation proposed by Fong, et al.: 8% Gelatine Type A from porcine skin Sigma Bloom 300; 10mmol/l Hydroquinone, 99%; 2 mmol/l Ascorbic Acid, 99%; 0,02 mmol/l CuSO4*5 H2O; 9% Methacrylic acid, and 83% distilled water. For the paranosal sinuses cavity experiment (a lesion in the head near the paranosal sinuses is simulated), two spherical glass balloons with a volume of 2 liter each were the phantom containers. Both glass balloons were filled with the MAGIC gel. The inhomogeneous phantom was prepared by placing a cylindrical cork to represent the air cavity: the diameter was 2,5 cm and the length 8 cm (3). The homogeneous phantom simulates the physical structure considered in the GammaPlan. Seven plastic vials of 100 ml were filled with the gel and were irradiated with doses of 0, 3, 5, 10, 15, 20 and 25 Gy with the Cobalt-60 TeleTherapy machine to obtain the calibration curve in order to derive the equivalent dose values from GammaPLan. The simulated tumour was given one shot with a dose of 20 Gray in the Gamma Knife using the 18 mm Helmet. A week following the irradiation, the phantoms and vials were scanned in a clinical Siemens 1.5 Tesla MR unit. For calculating the dose distributions in the phantoms, T2 mapping of the exposed normoxic gel is performed by using the Spin Echo MR sequences with TR=2000 ms and TE=10, 100, 110, 120, 170 and 269 ms, respectively. The phantoms were scanned in 9 slices of 3 mm each. MR images were reconstructed solely using T2 relaxation time constants, with each pixel value calculated by exponential nonlinear least squares fitting for different TE values. Both phantoms were irradiated with the same dose and the same coordinates. The Gammaplan dose distributions were compared by selecting the slices where the maximum dose has the largest area. In terms of the Leksell frame coordinates, this slice was located at Z=105 in the axial plane and Y=105 in the coronal plane. By using the fudicial marker system and Leksell frame coordinate conversion, identical slice coordinates were used in the GammaPlan. The 30%, 50% and the 70% isodose distances in the GammaPlan slices were measured using the pointer system on the monitor. In the case of the homogeneous phantom, the dose measured distribution is in good agreement with the GammaPlan calculated dose distribution. The dose distribution in the inhomogeneous phantom shows significant differences with that of the homogeneous phantom. The dose distribution in the inhomogeneous phantom is perturbed by the air cavity. The 30%, 50% and 70% isodoses in the inhomogeneous phantom are all asymmetric with respect to the shot center. We can conclude that there is a significant dose perturbation in the air cavity regions under Gamma Knife surgery, which is normally ignored by the GammaPlan