Huaqing Zheng

SU‐E‐T‐752: Energy Spectrum and Fluence Reconstruction of High Energy Photon Beams Based on Analytical Double‐Source Model

Purpose: To improve the accuracy on unfolding photon energy spectra and fluence for dose calculation, by use of measured percentage depth dose and off‐axis ratio, an analytical double‐source model considered both photonbeam and electron contamination was constructed, an improvement of Schiff formula and finite‐size pencil beam (FSPB) model. Methods: For energy spectrum reconstruction, several regression algorithms were used to obtain the optimum solution. It was also compared with the mono‐source model that ignored the electron contamination. For energy fluence reconstruction, the double‐source FSPB was developed by considering the energy spectra of photonbeam and electron contamination, and was also reconstructed by above algorithms. Results: A representative testing sample, the AAPM 55# report as an 18MVs AECL Therac‐20 medical linac, was used. The comparison between the double‐source model and mono‐source model in energy spectrum reconstruction found that there was a great improvement in accuracy; the mean error of double‐source model was about 0.015% while the mono‐source models was about 0.1%. We found that there are two peaks at 2 and 3 MeV in the photon energy spectra and the peak of the electron energy was in the low energy. We also found that if ignoring the electron contamination, the peak of the photon energy spectra could drift to the lower energy, and the mean error became great. We also obtained the accurate energy fluence, only about 0.01% mean error in the field. Conclusions: This study has revealed the great improvement of Schiff formula and FSPB model on energy spectrum and fluence of photonbeam and electron contamination by reconstruction method, an analytical double‐source model. Our finding will help know more about detail of medical linac, and improve the dose accuracy in dose calculation of radiotherapy by use of this reconstructed information. Supported by the National Natural Science Foundation under grant No.30900386 and the Anhui Provincial Natural Science Foundation under grantNo. 090413095 and 11040606Q55.

SU‐E‐T‐864: Beam Orientation Optimization Using Ant Colony Optimization in Intensity Modulated Radiation Therapy

Purpose: In intensity modulated radiation therapy(IMRT)treatment planning, gantry angles are usually preselected on the basis of experience and intuition. Therefore, getting an appropriate beam configuration needs a very long time. Based on the present situation, the paper puts forward beam orientation optimization using ant colony optimization. Methods: Using ant colony optimization to select the optimal beam configurations, after getting the beam configuration using Conjugate Gradient (CG) algorithm to optimize the intensity profiles. Based on the information of the effect of pencil beam‐let, the algorithm could accelerate to find the global optimal solution. Results: In order to verify the feasibility of the presented algorithm, a simulated and clinical case was tested. Comparison of Dose‐Volume histogram and isodose line between target area and organ at risk, the results showed that the plan quality was improved after optimizing beam configurations. Conclusions: The optimization approach can make treatment planning meet clinical requirements more efficiently, so it has extensive application perspective. Supported by the National Natural Science Foundation under grant No.30900386 and the Anhui Provincial Natural Science Foundation under grant No. 090413095 and 11040606Q55.

SU‐E‐T‐711: Monte Carlo Finite‐Size Pencil Beam Dose Calculation Method Based on Energy Spectra and Fluence Reconstruction

Purpose: Dose calculation is one of the core functions in Treatment Planning System (TPS). Finite‐Size Pencil Beam(FSPB) method is used in the clinic TPS widely, with the need of calculated time and inverse plan design. Yet the results of the traditional FSPB method may disagree with the practical situation due to lack of practical information of a given medical accelerator such as energy spectra and fluence. Methods: This study mainly focused on the shortage of traditional FSPB method, developing a new photondose calculation based on the Monte Carlo Finite Size Pencil Beam (MCFSPB) in the Advanced/Accurate Radiotherapy System (ARTS). Based on the MC simulation and the technology of medical accelerator energy spectrum reconstruction, a new pencil beam kernel model was constructed. In the condition of the filters influence, fluence reconstruction was also a part of MCFSPB model. With the convolution of fluence and MCFSPB kernel, energy deposition of body (or phantom) would be known. Results: Based on the above studies, we designed the MCFSPB method and implemented it with the visual c++ development tool. With several tests including the comparison among the AAPM55 Report sample, the results showed that the average error in the field size was less than 0.5% in the homogeneous phantom and less than 2% in the heterogeneous phantom. Conclusions: This method met the clinical criteria, and would be expected to be used as a fast and accurate dose calculation engine for clinic TPS. Supported by the National Natural Science Foundation under grant No.30900386 and the Anhui Provincial Natural Science Foundation under grant No. 090413095 and 11040606Q55.

Reconstrucion and visualization of 3D surface model from serial-sectioned contour points

In three-dimensional accurate radiotherapy treatment planning system, one of the key steps in the whole planning is to reconstruct and visualize the surface model of tumor target and organs at risk (OAR) from a serial of cross-sectioned contour points rapidly and accurately. This study designed a fast 3D-reconstruction and visualization pipeline comprising the following four main steps: ©pre-processing of contour points dataset; ©extraction and simplification of Iso-surface; ©linear transformation of surface model; @ Smoothing the surface model. An open source Visualization Toolkit (VTK) was applied to implement this method and a friendly user interface was developed on the Visual C++ development platform. Several clinic patients CT datasets were chosen for test data. The results show that this method could effectively avoid the “ladder effect” of standard marching cubes (MC) algorithm due to inconsistency between slice spacing and image resolution. The reconstruction surface is simplified to speed up the rendering time. Furthermore, it has been successfully integrated into domestic Accurate/Advanced Radiotherapy Treatment Planning System (ARTS) for clinical use.