Implementace metody rekonstrukce obrazu ve výpočetní tomografii na platformě FPGA

Abstract

The aim of this thesis is to develop and implement a system for online image reconstruction from a computed tomography (CT) demonstration model on an FPGA platform as a replacement for the current offline data processing, in which the measured data is stored on an SD card and subsequently reconstructed in the MATLAB environment. As part of the thesis, literature review was conducted on computed tomography, the potential use of FPGAs for image reconstruction, and CT data processing. Based on this review, solution was proposed utilizing a modular architecture that ensures data reception from the CT model, data storage, reconstruction using the back projection and filtered backprojection methods, and subsequent display of the results via a VGA output on the ZYBO board. The solution includes a description of the developed architecture and the method of controlling the system. The functionality of the proposed system was verified through simulation and via an experimental setup with a CT model using digital and real phantoms. The results of the simulations and experimental verification demonstrated the correct operation of the proposed solution, image consistency with the reference reconstruction in the MATLAB environment, and, at the same time, accelerated reconstruction when using an FPGA. Based on the obtained results, it can be concluded that the proposed and implemented system represents a viable alternative for the existing off-line reconstruction.