Optical CT and MR imaging of radiation dose distributions using the FBX -gel dosimeter.

摘要

In recent years, magnetic resonance imaging of gelatin doped with the Fricke solution has been applied to the direct measurement of three-dimensional (3D) dose distributions. However, the 3D-dose distribution can also be imaged more economically and efficiently using the method of optical absorption computed tomography. This is accomplished by first preparing a gelatin matrix containing a radiochromic dye and mapping the radiation-induced local change in the optical absorption coefficient. Ferrous Sulphate-Benzoic Acid-Xylenol Orange (FBX) was the dye of choice for this investigation. The complex formed by Fe 3+ and xylenol orange exhibits a linear change in optical attenuation (cm−1) with radiation dose in the range between 0 and 1000 cGy, and the local concentration of this complex can be probed using a green laser light (λ = 543.5 nm). An optical computed tomography (CT) scanner was constructed analogous to a first-generation x-ray CT scanner, using a He-Ne laser, photodiodes, and rotation-translation stages controlled by a personal computer. The optical CT scanner itself can reconstruct attenuation coefficients to a baseline accuracy of <2% while yielding dose images accurate to within 5% when other uncertainties are taken into account.;The radiation-induced conversion of ferrous ion (Fe2+) to ferric ion (Fe3+) in the FBX Gelatin dosimeter can also be measured using magnetic resonance imaging, similar to the standard Fricke-gelatin system. The oxidation process causes a shortening of the spin-spin (T 2), and spin-lattice (T1) relaxation times, each of which can be measured, with varying accuracy and precision, using different MR pulse sequences. In this investigation, the spin-lattice relaxation times of FBX gelatin were determined using both a fast inversion recovery pulse-sequence, and a three-dimensional Look-Locker (3D-LL) pulse-sequence. The inverse spin-lattice relaxation time (R1 = 1/T1) is shown to vary linearly with absorbed dose in the range 500–2000 cGy. The 3D-LL sequence however, can collect data for a full three-dimensional determination of T1 (12 slices, and 256 x 256 pixels per slice) in less than 10 minutes while the FIR sequence is limited to one slice every 35 minutes. The FIR sequence is shown to be much more accurate and precise at measuring T1 than the 3D-LL sequence. The accuracy of the 3D-LL sequence is shown to be dependent upon the interrogation pulse flip angle, the pulse repetition time, and the total number of pulses in the position of the images within the pulse train. Simulations were performed to determine the optimum parameters in which the 3D-LL sequence can be used to measure three-dimensional dose distributions in FBX gelatin.;By incorporating both the optical CT and MRI techniques we were able to develop a dual modality 3D dosimeter that can be used to validate modern conformal radiation techniques.