Exercise8 - TLE [VRT]
Go into the folder
tle. This simulation
Gate mac/main.mac computes the dose deposition of a low energy x-ray irradiation (for example a during a radiography).
Dose convergence is very slow with conventional method. An additional actor is used named TLE for Track Length Estimator which is a VRT method proposed in the 80' (Williamson 87) allowing a much faster convergence (see for example the analysis in the article [Baldacci2014]). Compare the outputs
dose-tle-Dose.mhd, comment the differences.
Gate output the Dose in Gy unit. Because very small number of particle is used, dose pixel values are around 10e-14 and so not easy to visualise. We provide a python script to rescale the pixel values between 0 and 100. You can use the script with
./py/normalize.py output/output-Dose.mhd output/output-Dose-norm.mhd (you need to provide the input and output filenames). Note: you need to install the Python module named simpleITK in order to use the script (type
pip install SimpleITK).
Then, the VV software can be used to visualize the dose distribution. Type
vv data/phantom.mhd --fusion output/dose-tle-Dose.mhd data/phantom.mhd --fusion output/dose-Dose.mhd --linkall & to compare the dose maps. You can obtain help in VV by pressing F1.
Replace the 3D dose actor with a 1D dose depth profile (and associated uncertainty). Plot the depth dose with gnuplot for both conventional dose and TLE dose.
An improvement of the TLE method has been recently proposed. This method is named seTLE and include splitting. Execute and comment the macro
main-setle.max that uses this seTLE method (described in [Smekens2014]). Compare the number of simulated particle per second (PPS in the stat file), compare the dose convergence and the uncertainty.