A dosimetric comparison between the LGP dose algorithms: TMR Classic and TMR 10Hamza Benmakhlouf1, Thomas Kraepelien2, Ernest Dodoo31Norsborg, Sweden 2Department of Hospital Physics, Karolinska University Hospital 3Department of Neurosurgery, Karolinska University Hospital Keywords: dose planning, gamma knife, physics, radiosurgery, Dose distribution
Recently, the new dose algorithm, TMR 10, for Leksell Gamma Plan was released; the differences in the two algorithms will be reflected in the absolute dose profiles, which will be used in the present work to evaluate the difference between the algorithms.
Our purpose was to compare these 2 calculation methods.
The dose profiles for x-, (in the x-y plane the dose profiles are symmetrical) and z-axis were extracted from LGP with both algorithms and compared to each other.
Prescribing 20 Gy to the 50 % isodose in TMR Classic gives following difference in TMR 10: In the z-axis (cranio-caudal direction) the dose difference is +4.4 Gy (+10.9 %) and -1.6 Gy (-4 %) in the cranial and caudal direction, respectively, for the 16 mm collimator. In the case of 8 mm collimator, the difference is -2.5 Gy (-6.2 %) and +4.4 Gy (+11 %), and for 4 mm collimators the difference is +1.8 Gy (+4.4 %) and +0.4 Gy (+1.1 %) in the cranial and caudal directions, respectively. For the x-axis (which is symmetrical with y-axis) the dose difference is +0.7 Gy (+1.8 %), +0.7 Gy (+1.8 %) and +1 Gy (+2.6 %) for the 16, 8 and 4 mm collimators, respectively.
Not all dose prescription concepts were evaluated.
A factor that converts the dose from one algorithm to the other cannot be given because the changes in the algorithm are more complex than that; the change in the dose profile and output factor does not go in the same direction for all the cases and has different orders of magnitude. Furthermore, the treatment planner must be aware of the, in some cases, considerable difference in the periphery dose between the two algorithms.
The impact the new algorithm has on the treatment plan could in some cases be critical, especially when a critical structure is located where the differences between the two algorithms are most pronounced. Project Roles:
H. Benmakhlouf (), T. Kraepelien (), E. Dodoo ()