#TITLE_ALTERNATIVE#

• Main Author: DWI SETIANI (NIM: 20911002) PEMBIMBING :Dr. Suprijadi ;Dr. Rer.nat Freddy Haryanto, TIA
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/19554
• Institution: Institut Teknologi Bandung

Monte Carlo (MC) is a powerful technique for simulation in x-ray imaging. Monte Carlo method can simulate the transport of radiation within matter with high accuracy and provides a natural way to simulate radiation transport in complex systems. MC-GPU is one of the codes based on Monte Carlo Algorithm developed by Andreu Badal that are used widely for radiographic images simulation and dose calculation in diagnostic radiology. The aim of this thesis was to investigate if MC simulation of an x-ray imaging system can be significantly accelerated in a GPU (Graphics Processing Unit) instead of in a standard CPU (Central Processing Unit). This thesis will also show the effect of physical paremeters to the quality of radiographic images, and evaluates if the images resulted from the simulation in the GPU has the same quality with the images from the CPU. Two different computation were applied, one of simulation run at CPU which simulated in serial condition, each particle/photon was calculated and after finished, the next particle will calculated and this calculation were repeatedly untill all particle, while in second methods, the GPU were used. Two GPU with 384 cores and 2304 cores were used, in this method, each cores can calculate one photon, so it can simulates many photon simultaneously. Three main components of the imaging system in this simulation are the radiation source, the detector, and the simulated material placed between the source and the detector. Each photon hit a simulated material to replace bone and water in three different density with dimension 30 x 30 x 10 cm2. Different enegy photon were used in range 10 keV – 120 keV. Simulation on the 2304 core of GPU were performed about 64 -114 times faster than on CPU, while the simulation on the 384 core of GPU were performed about 20 – 30 times faster than in the single core of CPU. Maximal speed of the presented simulation in CPU was 3,2 x 105 x rays/second and the maximal speed of the same simulation executed in the GPU was 4,4 x 107 x-rays/second. Optimal quality of images from the simulation was gained at the history start from 108 and the energy from 60 keV to 90 keV. Analyzed by statistical approach, The GPU and CPU images are relatively the same.