Abstract
Purpose/Introduction: A very promising tool for 3D dose verification in radiotherapy is represented by polymer gel dosimetry (PGD) utilizing MRI. Nonetheless, due to very long acquisition times (Tacq.) using the standard multi spin-echo (multi-SE) evaluation, this method is not used for standard dose verification yet. In this study, we prove
the feasibility of an evaluation employing on-resonant R1q, the socalled relaxation rate in the rotating frame, which leads to significantly reduced Tacq.
Subjects and Methods: The used PAGAT dosimetry gel consists of monomers, embedded within a gelatin matrix that starts to polymerize after irradiation1. This polymerization process results in a dose-dependent increase of the relaxation rate R2, which was evaluated by MRI (Biograph mMR, Siemens, Germany) using a 2D multi-SE sequence (32 echoes, TE = 40 ms, TR = 6 s, resolution 0.5 9 0.5 mm2, Tacq. = 25:36 min). The spin-lock sequence used for
determination of R1q consists of a tip-down pulse, followed by a locking pulse cluster (max. locking time 255 ms), and a subsequent tip-up pulse. Tipping pulses are realized using adiabatic half-passage pulses for improved B1-insensitivity. Image readout was realized using a HASTE2 sequence (32 locking times, TR = 12 s, 0.5x0.5 mm2, Tacq. = 6:36 min). For irradiation with 6 MV photons from a linear accelerator (Artiste, Siemens, Germany), a small spherical dose distribution (3 coplanar and equally–spaced 1x1 cm2-
fields) located completely within the gel container was administered. For dose calibration eight gel containers were irradiated with known doses of 0–7 Gy.
Results: The calibration curves can be described by a mono-exponential function (Fig. 1). Converting the R2- and R1q-images by means of this calibration curve into a dose-image revealed a clearly visible focal dose distribution with steep gradients as exemplarily shown for R2 in Fig. 2. Quantitative comparison of measured and calculated dose distributions showed a good agreement of 92.1 for R2-based and 98.1 for R1q-based evaluation using the 2D-gamma index criteria for 3 mm/3 3. The dose profiles through a sagittal slice (red line in Fig. 2) are shown in Figure 3.
Discussion/Conclusion: We could show that R1q-mapping is a feasible tool for accelerated quantitative PGD dose measurement. As the global spin-lock preparation can be combined with every conventional 3D-MRI readout (e.g., EPI) it is possible to acquire a 3Dvolume within the same acquisition time. This leads to an additional
acceleration factor as the multi-SE measurement time increases linearly with the number of slices. To further improve R1q-based PGD measurement, both dosimetry gels recipe and spin-lock preparation need to be perfectly tuned which is currently under investigation.
