After the initial arrest of a tumour treated with radiation therapy, regrowth can occur due to insufficient doses of radiation or incomplete coverage of the tumour. Newly developed irradiation techniques have made it possible to produce conformal dose patterns, which can elevate the dose to the tumour without increasing the complications in nearby normal tissue. Advances in conformal radiation require the development of new dosimeters capable of directly verifying the complex dose patterns in 3-D. The overall goal of this project is the optimization of a gel dosimeter that can be incorporated into phantoms that simulate patients, and the comparison of two imaging techniques used to read the 3-D dose patterns.

The specially developed dosimeter gels contain ferrous ions (Fe2+) and xylenol orange (FX gels). When the Fe2+ ions in the gel are irradiated, they oxidize to Fe3+ ions. The xylenol orange complexes with the Fe3+ ions, causing a colour change from translucent orange to translucent brown, depending on the level of radiation dose. The varying optical densities of the resultant dose pattern can be read using optical computed tomography (optical CT). Because the magnetic properties of Fe2+ and Fe3+ ions are different, the dose pattern can also be read with magnetic resonance imaging (MRI).

At present, the gel has been optimized for optical CT work, and an accurate map of the 3-D dose distribution within the gel can be obtained. The effect of temperature gradients on the gel dose response has also been investigated. Although optical CT has shown great promise as a fast, inexpensive, accurate method of determining dose, it has an inherent limitation: only optically transparent mediums can be scanned with this method. Also, reflection and refraction at medium interfaces can lead to inaccurate dose values. Therefore, MR is more suitable modality for imaging within inhomogeneous or opaque phantoms.

One concern associated with MRI is its long data acquisition times, which may allow significant diffusion of the Fe3+ ions, resulting in degraded spatial resolution. For MRI to be used as a detection method in dosimetry, pulse sequences must be developed to acquire images quickly, before significant "blurring" of the dose pattern has occurred. A fast T1 mapping sequence based on the Look-Locker sequence has shown promise as a 3-D mapping technique. The next phase of this project is to improve the signal-to-noise ratio of 3D MRI dosimetry, allowing accurate dose mapping in gels irradiated and read while still within heterogeneous phantoms.