Planning System for Proton Therapy
Modernization of a Treatment Planning System for Proton Therapy (Development of Monte Carlo Codes, Databases, Treatment Planning System Verification, Verification of Complicated Treatment Plans)
Tech Area / Field
- INF-SOF/Software/Information and Communications
- PHY-PFA/Particles, Fields and Accelerator Physics/Physics
8 Project completed
Senior Project Manager
Tocheny L V
VNIITF, Russia, Chelyabinsk reg., Snezhinsk
- ITEF (ITEP), Russia, Moscow
- Lawrence Livermore National Laboratory, USA, CA, Livermore
Project summaryDevelopment of effective methods and means for cancer detection and treatment is one of the priorities for the World Health Organization and national health departments of all high-developed countries. Every year more than 18 million people throughout the world contract cancer and more than 4 million people die of it. The number of cancer patients registered in the world exceeds 40 million.
Radiotherapy plays a great role in cancer treatment. About 40-70% of cancer patients need radiotherapy often in combination with other treatment methods (surgery, chemotherapy, or hormonotherapy). The main task of radiotherapy is to provide the local control of the tumor by delivering a dose that would be such as to allow tumor destruction and/or resorption at the least impact on healthy organs and tissues near the tumor. The task can be accomplished subject to the following basic factors:
- the possibility of early detection and identification of tumors;
- the possibility of generating dose fields conformal to tumors of even very complicated configurations;
- the possibility of delivering an adequate dose within optimal terms and at the least percentage of radioreactions and complications in the skin, mucous coats and other critical organs;
- the possibility of reducing radiation damage to healthy tissues and organs that helps increase the dose so as to increase the probability local control; and
- the possibility of assessing the tumor response to the dose and the rate of tumor regression to adjust further treatment and allow for the inpidual peculiarities of patients.
Almost all of the above factors are taken into account and actively optimized in the use of treatment planning systems (TPS) for radiotherapy at the stage of preparation to and conduct of proton treatment. As a rule, all up-to-date TPSs use topometric data from computer tomography (CT). It is also possible to use images from magnetic resonance tomography (MRT) and other diagnostic methods. Modern TPSs are capable of a detailed analysis of computer tomograms and other digitized images (contrasting, identification of target zones, target visualization as a volume structure and other). Thus, besides the use of TPSs for their designed purpose, these systems help exact diagnosis.
Since 1967 ITEP has used the proton beam generated by its synchrotron for medical purposes. Now ITEP has three booths to which stationary horizontal beams of protons of energies between 80 and 220 MeV are delivered. The booths are equipped with four positioners for
- multistage convergent irradiation of head targets through the target and with the stop at a given depth;
- wide-field irradiation of body targets in lying;
- irradiation of head and neck targets in the seating; and
- irradiation of chest targets in the seating position.
Several years ago scientists from ITEP began to develop a hardware and software suite to be used for planning proton and combined irradiation. The suite was called ProCom. The developed code ProGam is a computational kernel of a system which models the dose field of interest.
In order that a TPS provide correct information, it is necessary that its database contain information on the proton beams to be used. The computational model of ProGam uses experimental data on beam parameters (track length, cross sectional and through-depth dose depositions etc.) measured in a homogeneous water phantom. On the one hand, this helps avoid some errors which are inevitable in the use of a solely computational method to predict dose depositions because any analytical methods used to speedup calculations are based on approximations and simplifications. On the other hand, experiments aimed to get data on beam parameters are labor-intensive and expensive. The more so as the quality of TPS end product is fully dependent on measurement accuracy, and measurements can be deemed reliable only after certain statistics have been collected.
Modern computational methods which use Monte Carlo (MC) method for tracking protons are much more accurate than the analytical methods used in many TPS kernels primarily to speedup calculations. Indeed, a planner which uses a TPS to plan the proton therapy séance must operate practically in real time, and analytical methods help implement this. Moreover, the large number of patients and the time and cost of treatment planning require fast calculations. On the other hand, the time is not so critical when the TPS database is being filled with information. The time required for full-scale experimental measurements is much more expensive than the computer time required for dose deposition calculations if even it makes hours. At the same time, accuracy of Monte Carlo calculations is comparable with measurement accuracy and sometimes exceeds it. That is why the use of Monte Carlo simulations to fill TPS databases seems much promising.
The first and main objective of this research is to develop a state-of-the-art Monte Carlo code to predict dose depositions which will be used as initial information entered into the database of the ProGam kernel of the ProCom treatment planning system.
The next important task is to develop simulation codes making calculations with the use of tomography data entered into a CT-cube. This will help obtain the exact dose field for each particular patient. The use of codes based on the Monte Carlo method will provide reliable information which helps adequately implement advantages offered by proton therapy. It is assumed here that these calculations may become an essential addition to ProGam kernel calculations when, for instance, the presence of compound heterogeneities or critical organs near the target poses stricter requirements for the accuracy of dose deposition calculations. In these cases, requirements for the time of calculation critical to MC methods are not very strict. With the modular structure of the ProCom TPS, it is relatively easy to change the computational kernel.
At the same time it is desirable to minimize the time needed to prepare patients for treatment. This is especially important in complicated medical cases. High requirements for computational accuracy also results in much longer times of calculation. Despite the rapid progress of high performance computers, it is impossible to minimize the time of calculation by building up only computer resources. This is especially true for real time calculations. Specialized solvers offer an alternative solution to the problem.
The project aims to develop and prototype a specialized Monte Carlo solver based on large programmable gate arrays. It is expected that with this solver the time needed to calculate a “standard” problem which now makes ~1 hour on Pentium-4 @ 3GGz will reduce to 1-2 minutes, i.e., the calculations will become ~30-60 times faster. There exists an alternative based on parallel processing on a PC cluster, but it is much more expensive.
The above goal will be achieved due to parallelization of basic operations (generation of new particle coordinates, determination of boundary crossings, collection of statistics etc.) implemented in specialized conveyer blocks of the solver.
Scientists from RFNC-VNIITF, who will deal with the problem, are experienced in the development of specialized solvers for neural networks (ISTC Project 31) and for tomography problems where the desired effect has been achieved (from 20 to 60 times faster calculations).
So, the objective of this project is to develop
- a code package to predict dose deposition from accelerated protons in homogeneous media to fill the database of ProCom treatment planning system;
- a code package to predict dose deposition from accelerated protons in heterogeneous media to verify ProCom treatment plans in cases where complicated heterogeneities or critical structures are present near the exposure site; and
- a specialized Monte Carlo solver.
Unique experience gained by RFNC-VNIITF scientists in the development of MC codes for radiation transport will contribute to the success of the project. For more than 30 years, a universal code called PRIZMA has been developed and used to model the separate and coupled transport of neutrons, photons, electrons, positrons and ions. The code HANDRA was developed to track high energy hadrons; it was used to predict radiation fields around high-energy facilities, specifically activation of and radioactivity evolution in HF CMS detector. What will also be used under the proposed project is the experience gained in the adaptation of SERA treatment planning system to Snezhinsk Neutron Therapy Center implemented under ISTC Project #2145 in 2002-2004. PRIZMA code was used under that project to develop a computer model of the NG-12I neutron generator, which allowed rather accurate dose deposition calculations by SERA within a time acceptable for clinical practice.
Scientists from ITEP have also gained vast experience in the use of protons for treating some cancers and complicated diseases other than cancer. Today, despite certain progress of proton therapy which in recent 15 years has more than doubled the number of proton therapy centers throughout the world, the contribution of ITEP to the total amount of patents treated by protons makes about 10%. The proton treatment department is part of a large research center, namely ITEP; it unites high-skilled scientists and engineers, who rely upon the experience gained during about 40 years of ITEP existence. The synchrotron used at the department allows parallel operations for several customers, making it possible to perform required experiments in evenings and nights. The experimental equipment available will allow measurements which are needed to provide Monte Carlo codes with initial data on proton beam parameters and to verify calculations against the actual conditions of use. ITEP measures the absolute absorbed dose with accuracy required by international standards (within 5%). Dose deposition in the fields of accelerated protons can be measured within 2-3%, and geometrical parameters of these fields can be determined to fractions of millimeter, as well as the positioning of an exposed object on the irradiation bench. The quality of the user interface to be developed will be evaluated by physicians-radiologists. The interface will be developed using experience gained by scientists from SSC RF–ITEP, who developed ProCom treatment planning system.
The project falls into the applied research category; its possible achievement includes a treatment planning system as a transferable product.
After the project is completed, its authors plan to continue research with the purpose to introduce the treatment planning system developed into clinical practice.
Cooperation with foreign collaborators will include:
- information exchange,
- consultations and discussions,
- comments to technical reports (quarterly, annual and final), to be submitted to the ISTC, and
- verification of results obtained under the project.
The International Science and Technology Center (ISTC) is an intergovernmental organization connecting scientists from Kazakhstan, Armenia, Tajikistan, Kyrgyzstan, and Georgia with their peers and research organizations in the EU, Japan, Republic of Korea, Norway and the United States.
ISTC facilitates international science projects and assists the global scientific and business community to source and engage with CIS and Georgian institutes that develop or possess an excellence of scientific know-how.