A BRACHYTHERAPY TREATMENT PLANNING SOFTWARE BASED ON MONTE

A BRACHYTHERAPY TREATMENT PLANNING SOFTWARE BASED ON MONTE CARLO SIMULATIONS AND ARTIFICIAL NEURAL NETWORK ALGORITHM Amir Moghadam 1

Prostate Brachy. Therapy 2

Treatment Planning System(TPS) �A Treatment Planing System essentially consists of three parts: � a system for dose prescription. � a set of rules to distribute the sources inside a defined volume to achieve a clinically acceptable dose distribution. � a method to calculate patient dose. 3

Dose Calculation Methods � Analytical models (not used at low energies) � Deterministic solution to the transport equation (almost fast, also accurate) � Monte Carlo (most accurate but too slow) � TG-43 Formalism (fast but not accurate) assumes human body as homogeneous water 4

What are we going to do? � We want to create a method which is as fast as TG-43 Formalism but has a similar accuracy to Monte Carlo method. � To do this, first we have to create a 3 D model of the patient’s body in our Monte Carlo code MCNP, based on his CT Scan image. 5

How to define the material and densities CT Number -1000 to -150 -151 to 90 91 to 2200 Material Air Soft tissue Bone Cortical Density range (g/cm^3 ) 0. 001 to 0. 9 0. 85 to 1. 16 to 3. 1 6

Dicom to MCNP input conversion 7

Two methods of MCNP calculation Total phantom Modeling Supper position Method 8

The reason for Dose calculation by ANN � Dose calculation using MCNP takes at least 4 hours to reach an acceptable error on a Cori-7 computer. � To improve the speed of Monte Carlo dose calculation method we trained a set of Artificial Neural Networks to perform the dose calculation in a shorter time but with the accuracy similar to MCNP calculations. 9

How ANN works � Inputs : attenuation coefficients of the voxels of the cube � Output: 3 D Dose distribution inside the cube because of the seed at the center of the cube 10

The ANN Training Flow cchart 11

Last step (using data compression) Outer region 0. 6 cm 10*10 mesh Inner region 0. 3 cm 8*8 mesh Model details are reduced as we get far from the seed 12

Time and number of sample cubes used for training We modeled 44832 model cubes inside MCNP and we used 15000 of these model cubes for training the ANN and the remaining were saved for validation of ANN for unknown cases. � MCNP calculations took 40 minutes for each of the model cubes. (total MCNP simulation for 44832 cubes will be 103 days on 12 Nodes!!? ) � Training of each of ANNs took about 20 to 60 minutes. � 13

Results of inner region for single seeds 14

Results of inner region for single seeds 15

Results of inner region for single seeds 16

Results of outer region for single seeds 17

Results of outer region for single seeds 18

Results of outer region for single seeds 19

Validation of ANNs 20

Realistic Treatment Plan 21

Non-realistic treatment plan 22

Speed assessment of the New Method Dose Calculation Method Accuracy Time needed for 1 seed MCNP 40 minutes for 1 12 days on a coriseed 7 computer Highest Time needed for 467 seed positions Varian’s Acuros High (Linear Boltzman (dependant Transport Equation Solver) on the mesh size) 3 -8 minutes for one change in position 23 hours Artificial Neural Network 8 minutes for 1 seed 17 minutes on a rusty computer Similar to MCNP 23

Future work This study can be extended to the cases of High Dose Rate and Low Dose Rate Brachy. Therapy and any other field in which effects of heterogeneity are much more of interest and the speed of calculation is critical. � Other methods of compression can be tested for improving the ANNs accuracy. � Other Architectures of ANN can be used for decreasing the relative errors of ANNs for Farther voxels. � 24

Thanks for your attention 25