B De Smedt

Monte Carlo evaluation of the AAA treatment planning algorithm in a heterogeneous multilayer phantom and IMRT clinical treatments for an Elekta SL25 linear accelerator

The Anisotropic Analytical Algorithm (AAA) is a new pencil beam convolution/superposition algorithm proposed by Varian for photon dose calculations. The configuration of AAA depends on linear accelerator design and specifications. The purpose of this study was to investigate the accuracy of AAA for an Elekta SL25 linear accelerator for small fields and intensity modulated radiation therapy (IMRT) treatments in inhomogeneous media. The accuracy of AAA was evaluated in two studies. First, AAA was compared both with Monte Carlo (MC) and the measurements in an inhomogeneous phantom simulating lung equivalent tissues and bone ribs. The algorithm was tested under lateral electronic disequilibrium conditions, using small fields (2 x 2 cm(2)). Good agreement was generally achieved for depth dose and profiles, with deviations generally below 3% in lung inhomogeneities and below 5% at interfaces. However, the effects of attenuation and scattering close to the bone ribs were not fully taken into account by AAA, and small inhomogeneities may lead to planning errors. Second, AAA and MC were compared for IMRT plans in clinical conditions, i.e., dose calculations in a computed tomography scan of a patient. One ethmoid tumor, one orophaxynx and two lung tumors are presented in this paper. Small differences were found between the dose volume histograms. For instance, a 1.7% difference for the mean planning target volume dose was obtained for the ethmoid case. Since better agreement was achieved for the same plans but in homogeneous conditions, these differences must be attributed to the handling of inhomogeneities by AAA. Therefore, inherent assumptions of the algorithm, principally the assumption of independent depth and lateral directions in the scaling of the kernels, were slightly influencing AAAs validity in inhomogeneities. However, AAA showed a good accuracy overall and a great ability to handle small fields in inhomogeneous media compared to other pencil beam convolution algorithms.

MCDE: a new Monte Carlo dose engine for IMRT

A new accurate Monte Carlo code for IMRT dose computations, MCDE (Monte Carlo dose engine), is introduced. MCDE is based on BEAMnrc/DOSXYZnrc and consequently the accurate EGSnrc electron transport. DOSXYZnrc is reprogrammed as a component module for BEAMnrc. In this way both codes are interconnected elegantly, while maintaining the BEAM structure and only minimal changes to BEAMnrc.mortran are necessary. The treatment head of the Elekta SLiplus linear accelerator is modelled in detail. CT grids consisting of up to 200 slices of 512 x 512 voxels can be introduced and up to 100 beams can be handled simultaneously. The beams and CT data are imported from the treatment planning system GRATIS via a DICOM interface. To enable the handling of up to 50 x 10(6) voxels the system was programmed in Fortran95 to enable dynamic memory management. All region-dependent arrays (dose, statistics, transport arrays) were redefined. A scoring grid was introduced and superimposed on the geometry grid, to be able to limit the number of scoring voxels. The whole system uses approximately 200 MB of RAM and runs on a PC cluster consisting of 38 1.0 GHz processors. A set of in-house made scripts handle the parallellization and the centralization of the Monte Carlo calculations on a server. As an illustration of MCDE, a clinical example is discussed and compared with collapsed cone convolution calculations. At present, the system is still rather slow and is intended to be a tool for reliable verification of IMRT treatment planning in the case of the presence of tissue inhomogeneities such as air cavities.

The importance of accurate linear accelerator head modelling for IMRT Monte Carlo calculations

Two Monte Carlo dose engines for radiotherapy treatment planning, namely a beta release of Peregrine and MCDE (Monte Carlo dose engine), were compared with Helax-TMS (collapsed cone superposition convolution) for a head and neck patient for the Elekta SLi plus linear accelerator. Deviations between the beta release of Peregrine and MCDE up to 10% were obtained in the dose volume histogram of the optical chiasm. It was illustrated that the differences are not caused by the particle transport in the patient, but by the modelling of the Elekta SLi plus accelerator head and more specifically the multileaf collimator (MLC). In MCDE two MLC modules (MLCQ and MLCE) were introduced to study the influence of the tongue-and-groove geometry, leaf bank tilt and leakage on the actual dose volume histograms. Differences in integral dose in the optical chiasm up to 3% between the two modules have been obtained. For single small offset beams though the FWHM of lateral profiles obtained with MLCE can differ by more than 1.5 mm from profiles obtained with MLCQ. Therefore, and because the recent version of MLCE is as fast as MLCQ, we advise to use MLCE for modelling the Elekta MLC. Nevertheless there still remains a large difference (up to 10%) between Peregrine and MCDE. By studying small offset beams we have shown that the profiles obtained with Peregrine are shifted, too wide and too flat compared with MCDE and phantom measurements. The overestimated integral doses for small beam segments explain the deviations observed in the dose volume histograms. The Helax-TMS results are in better agreement with MCDE, although deviations exceeding 5% have been observed in the optical chiasm. Monte Carlo dose deviations of more than 10% as found with Peregrine are unacceptable as an influence on the clinical outcome is possible and as the purpose of Monte Carlo treatment planning is to obtain an accuracy of 2%. We would like to emphasize that only the Elekta MLC has been tested in this work, so it is certainly possible that alpha releases of Peregrine provide more accurate results for other accelerators.

Investigation of geometrical and scoring grid resolution for Monte Carlo dose calculations for IMRT

Monte Carlo based treatment planning of two different patient groups treated with step-and-shoot IMRT (head-and-neck and lung treatments) with different CT resolutions and scoring methods is performed to determine the effect of geometrical and scoring voxel sizes on DVHs and calculation times. Dose scoring is performed in two different ways: directly into geometrical voxels (or in a number of grouped geometrical voxels) or into scoring voxels defined by a separate scoring grid superimposed on the geometrical grid. For the head-and-neck cancer patients, more than 2% difference is noted in the right optical nerve when using voxel dimensions of 4 x 4 x 4 mm3 compared to the reference calculation with 1 x 1 x 2 mm3 voxel dimensions. For the lung cancer patients, 2% difference is noted in the spinal cord when using voxel dimensions of 4 x 4 x 10 mm3 compared to the 1 x 1 x 5 mm3 calculation. An independent scoring grid introduces several advantages. In cases where a relatively high geometrical resolution is required and where the scoring resolution is less important, the number of scoring voxels can be limited while maintaining a high geometrical resolution. This can be achieved either by grouping several geometrical voxels together into scoring voxels or by superimposing a separate scoring grid of spherical voxels with a user-defined radius on the geometrical grid. For the studied lung cancer cases, both methods produce accurate results and introduce a speed increase by a factor of 10-36. In cases where a low geometrical resolution is allowed, but where a high scoring resolution is required, superimposing a separate scoring grid on the geometrical grid allows a reduction in geometrical voxels while maintaining a high scoring resolution. For the studied head-and-neck cancer cases, calculations performed with a geometrical resolution of 2 x 2 x 2 mm3 and a separate scoring grid containing spherical scoring voxels with a radius of 2 mm produce accurate results and introduce a speed increase by a factor of 13. The scoring grid provides an additional degree of freedom for limiting calculation time and memory requirements by selecting optimized scoring and geometrical voxel dimensions in an independent way.

DOSSCORE: an accelerated DOSXYZnrc code with an efficient stepping algorithm and scoring grid

DOSSCORE is an accelerated version of DOSXYZnrc that allows photons to cross voxel boundaries of the same medium and utilizes a separate scoring grid superimposed on the geometrical grid. Two different stepping algorithms, the hownear method and the scaling method are implemented in DOSSCORE. The hownear method allows particles to travel larger distances in homogeneous regions where there is no interest in the dose deposition of these particles, whilst the scaling method utilizes a stepping algorithm in which particles are only slowed down by the boundaries of the geometrical voxels and not by the boundaries of the scoring voxels. For CT-based phantoms, only photon ray tracing is applied, which results in a rather modest speed gain of factor 1.2 compared to DOSXYZnrc. The hownear method and scaling method do not increase the speed for CT-based phantoms, but only for homogeneous phantoms and phantoms with a limited number of small heterogeneities. In cases where a small number of scoring voxels are needed, the hownear method performs better than the scaling method, whilst the opposite is true for cases when many scoring voxels are needed. The photon transport is accelerated by almost a factor of 2 for all phantoms (homogeneous, heterogeneous with much homogeneity and CT-based phantoms) compared to DOSXYZnrc. For a small number of scoring voxels, the hownear method is up to a factor of 2.6 and 1.9 faster than DOSXYZnrc for homogeneous and heterogeneous phantoms in the case of photon beams. For an electron beam, a speed gain of factor 2.4 is obtained. For a full scoring grid like the one used in DOSXYZnrc, the scaling method is up to a factor of 2.2 and 1.7 faster than DOSXYZnrc for homogeneous and heterogeneous phantoms in the case of photon beams. For an electron beam, a speed gain of factor 2 is obtained. A speed increase without biasing the results is very relevant. The use of two separate grids, the more efficient stepping algorithms and the accelerated photon transport can be applied to every EGS-based or other Monte Carlo code.

Denoising of Monte Carlo dose calculations: smoothing capabilities versus introduction of systematic bias.

In order to evaluate the performance of denoising algorithms applied to Monte Carlo calculated dose distributions, conventional evaluation methods (rms difference, 1% and 2% difference) can be used. However, it is illustrated that these evaluation methods sometimes underestimate the introduction of bias, since possible bias effects are averaged out over the complete dose distribution. In the present work, a new evaluation method is introduced based on a sliding window superimposed on a difference dose distribution (reference dose-noisy/denoised dose). To illustrate its importance, a new denoising technique (ANRT) is presented based upon a combination of the principles of bilateral filtering and Savitzky-Golay filters. This technique is very conservative in order to limit the introduction of bias in high dose gradient regions. ANRT is compared with IRON for three challenging cases, namely an electron and photon beam impinging on heterogeneous phantoms and two IMRT treatment plans of head-and-neck cancer patients to determine the clinical relevance of the obtained results. For the electron beam case, IRON outperforms ANRT concerning the smoothing capabilities, while no differences in systematic bias are observed. However, for the photon beam case, although ANRT and IRON perform equally well on the conventional evaluation tests (rms difference, 1% and 2% difference), IRON clearly introduces much more bias in the penumbral regions while ANRT seems to introduce no bias at all. When applied to the IMRT patient cases, both denoising methods perform equally well regarding smoothing and bias introduction. This is probably caused by the summation of a large set of different beam segments, decreasing dose gradients compared to a single beam. A reduction in calculation time without introducing large systematic bias can shorten a Monte Carlo treatment planning process considerably and is therefore very useful for the initial trial and error phase of the treatment planning process.