FRED: a fast-MC tool for treatment planning and dose verification in proton therapy

The advent of general-purpose programmable GPUs has prompted the development of MC tools that can deliver a plan calculation within minutes [1]. The new hardware requires to substantially rewrite the core kernels of a MC calculation adopting the parallel execution model of a typical GPU. A fast-MC tool called Fred has been developed at University of Rome "La Sapienza". The code can track proton pencil beams through a voxel 3D patient reconstruction grid based on high resolution CT scans. The kernel implements energy loss based on tabulated stopping power [2].
Different models (single Gaussian, double Gaussian, Gauss-Rutherford) can be used to reproduce Multiple Coulomb Scattering.  Secondary protons and deuterons are produced in the beam nuclear interaction following the multiplicity, energy and angular distributions provided by ICRU63 [3], and interpolated for other materials.

The voxel HU of the CT is converted in elemental composition following the method by Schneider[4]. Alpha and heavier fragments produced are treated as local dose deposition and (in this version) neutral particles are not produced. The proton RBE value is kept constant for direct comparison with commercial TPS results, but it can be also loaded from an external library of RBE values. The dose matrix can be processed by an optimizer based on a least-squares optimization algorithm as in [5].
The output/input system has been adapted to the CNAO environment[6]. All these solutions have been checked for efficacy and accuracy against a full-MC code[7], and the resulting kernel can deliver more than 1 million complete histories per second on a single GPU. The achieved tracking rate allows to perform a typical plan recalculation within minutes. 


IMPLEMENTATION OF CARBON IONS FRAGMENTATION

For the exceptional speed of the proton tracking algorithms implemented in FRED and for the excellent results achieved, the door to several applications within the Particle Therapy field has been opened. In particular it determined the interest to develop FRED also for Carbon therapy applications, in order to recalculate treatment plans with Carbon ions. The main difference between proton and Carbon beams is the nuclear fragmentation of the projectile in a ¹²C treatment which doesn’t occur with protons. The simulation of the ion beam fragmentation gives an important contribution to the dose deposition. Indeed, the total dose release is due not only to the primary beam, but also to the secondary and tertiary particles. The fragments, having on average the same energy per nucleon of the primary beam and a lower mass, can release dose also after the Bragg Peak causing the well known fragmentation tail (see figure). The code has been upgraded with a model for the Carbon fragmentation, developing a complete modelling of the interaction between beam and patient body, as it has been done for proton beams. 

Presently, the implementation of the 12C model is satisfactory, with an excellent agreement with the full-MC FLUKA (as shown in the picture). Beside the successful implementation of the nuclear model, capable of clinical precision when computing the absorbed dose in particle therapy conditions, FRED also achieved an impressive improvement in computing time, with respect to conventional full MC software solutions. Exploiting the parallel programming power of GPU architectures, FRED is capable of tracking millions of primary particles per second on a single GPU card. The observed gain in processing time, when comparing to the FLUKA full MC, was nearly a factor ~2000, depending on the energy of the primary beam.

IMPLEMENTATION OF ELECTROMAGNETIC INTERACTION

Collaboration network:

• A. Schiavi, V. Patera, A. Sarti, M. De Simoni, M. Fischetti, G. Franciosini, F. Salvati, M. Pacitti, G. Acciaro, G. Traini, G. Battistoni (Italy, University La Sapienza Rome and INFN)
• N. Krah (France, CREATIS, CNRS/University Lyon)
• A. Rucinski, J. Gajewski, M. Garbacz, A. Skrzypek, J. Baran (Poland, PAN, Krakow)
• I. Rinaldi (Netherlands, Maastro clinic, Maastricht)

Papers

• M. De Simoni et al, A Data-Driven Fragmentation Model for Carbon Therapy GPU-Accelerated Monte-Carlo Dose Recalculation 2022 Frontiers in Oncology 12
• M. Senzacqua et al, A fast - Monte Carlo toolkit on GPU for treatment plan dose recalculation in proton therapy 2017 J. Phys.: Conf. Ser. 905 012027
• A. Schiavi et al, Fred: a GPU-accelerated fast-Monte Carlo code for rapid treatment plan recalculation in ion beam therapy 2017 Phys. Med. Biol. 62 7482
• A. Rucinski et al, GPU-accelerated Monte Carlo code for fast dose recalculation in proton beam therapy Acta Physica Polonica B 48(10), pp. 1625-1630

Bibliography

[1] X. Jia et al., Phys. Med. Biol. (2012) 57:7783–7797; X. Jia et al., Phys. Med. Biol. (2014) 59:R151–R182; D. Giantsoudi et al., Phys. Med. Biol. (2015) 60:2257–2269 


[2] M.J.Berger et al, PSTAR, version 1.2.3 (2005). Available online: http://physics.nist.gov/Star


[3] Schneider U et al., Phys. Med. Biol. (1996) 41:111–24


[4] ICRU Report 63 (2000)


[5] A. Mairani et al., 2013 Phys Med Bio . 58, 2471-2490; Lomax et al 1999  Phys. Med. Bio. 44 185-205


[6] S. Molinelli et al., Phys. Med. Biol. (2013) 58:3837-47


[7] A. Ferrari et al., CERN-2005-10 (2005), INFN/TC_05/11, SLAC-R-773

[8] RADIANCE- A planning software for intra-operative radiation therapy 2015- Manilo, F. et al. In: Trans- lational Cancer Research, Vol 4;                                                                                                                

[9] V. Favaudon et al., Ultrahigh dose-rate flash irradiation increases the differential response be- tween normal and tumor tissue in mice Science Translational Medicine 6 (2014) 245ra93. doi:10.1126/scitranslmed.3008973.