A user-friendly Monte Carlo system for the accurate simulation of clinical beams and dose distributions in radiotherapy patients

• Applicant: Professor Dr. Lorenzo Brualla Barbera
• Subject Area: Medical Physics, Biomedical Technology
• Term: from 2014 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 267890656

The Monte Carlo (MC) method is considered the state-of-the-art technique for the simulation of radiation transport problems in radiotherapy, in particular for modeling medical linear accelerators (linacs). However, most treatment planning systems are fully or partly based on analytical methods to determine absorbed dose distributions in patients. These analytical methods are faster to compute but, in principle, less accurate than MC algorithms. Simulation speed is not the only reason that has hindered the wide use of the MC method in the daily clinical practice. The skill and effort required for using a MC code also play a decisive role.To address these difficulties we developed the computer program PRIMO, which is based on the general-purpose MC radiation transport code PENELOPE. PRIMO simulates radiation beams produced by most Varian and Elekta linacs, and estimates the subsequent absorbed dose distribution in water phantoms or computerized tomographies of patients. It prepares, executes and analyzes the simulation of a linac and the dose distribution without requiring knowledge about the MC method or the geometrical details of the linac. PRIMO is free software and can be downloaded from http://www.primoproject.net.The project will quest for solutions to the following open research problems: i) graphics processing unit (GPU) parallelization of a general-purpose MC radiation transport code that exploits the full computing capacity of the GPU while not curtailing the accuracy of the physics and transport models; ii) optimization of multiple simultaneous variance-reduction parameters; iii) development of an algorithm for determining the primary beam parameters of a MC simulation of a linac based on experimental dose profiles; iv) development of a fast denoising algorithm that calculates an upper bound of the statistical uncertainty; v) generation of a method for determining experimentally-based geometrical descriptions of undisclosed linac components; vi) derivation of an accurate Monte Carlo estimator of the statistical uncertainty for multiple irradiation fields simulation in a parallel architecture.The solutions obtained for these research problems will allow to implement the simulation of intensity-modulated radiotherapy in PRIMO and to significantly speed up the code. The main expected result of the project is the first fully MC-based treatment planning and verification code which can be used for scientific research in medical physics and radiation oncology to solve problems that cannot be assessed by analytical methods with satisfactory accuracy. A more accurate dose calculation translates into a better treatment, a more effective use of radiotherapy and a better understanding of the effects of radiation in tumors and normal tissues.

• DFG Programme: Research Grants
• International Connection: Panama, Spain
• Co-Investigators: Privatdozent Dr. Lutz Lüdemann; Professor Dr. Wolfgang Sauerwein
• Cooperation Partners: Dr. Miguel Rodriguez; Dr. Josep Sempau