Study on Dosimetric Impacts of Spatial–Temporal Uncertainity of Organ Motion against a Wobbling Proton Beam Using 3D/4D Voxelized Phantoms

CGMH proton center will be the first few facilities in the world equipped with wobbling proton beam system, and will have first beam delivered in late 2013. The wobbling system utilizes rotating dipoles to produce a uniform dose distribution, which can save beam shaping absorbers in the treatment head (increase the efficiency), reduce secondary particle contamination, and increase the flexibility of delivering variable modulation of the SOBP. Unlike in US where prostate cancer is the most treated cancer using proton therapy, in Taiwan, however, the candidate cancer to be treated would be mostly liver or lung cancer. When treating a moving target, target and proton beam are both dynamic, the mis-synchronization of beam delivery and organ motion may cause dosimetric impacts owing to spatial-temporal uncertainty. This project proposes to study these dosimetric impacts of spatial-temporal uncertainty against a wobbling proton beam by (1) modeling proton motion in a magnetic field, (2) constructing 3D/4D voxelized phantoms, and (3) Monte Carlo simulation. Proton motion in a magnetic field can be sampled by its kinetic energy, E, maximum spreading radius r, and applied magnetic field B. 3D medical imaging techniques, such as CT and MRI, allow us to construct 3D voxelized models, which contain a huge number of tiny cubes grouped together to represent each anatomical structure (tissues or organs) and are inherently realistic since they contain a large amount of anatomical information. For modeling a respirative human, it is important to model his dynamic behavior in different time frames (4D). A 4D breath-simulating anatomical model is a series of 3D voxelized models, whose shape, size and location change according to specific respiratory motion patterns. Organs can be converted into polygon surface models to extract the anatomical features. Polygon models were then translated into NURBS surfaces, and then be deformed by changing control points according to clinical breathing models. MCNPX and GEANT4 will be utilized in this study to model both the wobbling nozzle and 3D/4D voxelized phantoms. After a standardized 4D human model is constructed, extra spatial-temporal uncertainty will be added into this model to simulate mis-synchronization of beam delivery and organ motion.

Project ID:PC10207-0436
External Project ID:NSC102-2314-B182-054