Radiation Dose Estimation Preliminary Study for Sparse-View CT: Monte-Carlo Simulation

Computed Tomography (CT) is one of the most used image-guided diagnosis modalities in clinical practice. However, the ionization produced by the x-rays poses a potential risk to patients exposed to this technique. Therefore, a methodology to reduce x-ray exposition without compromising the image quality will minimize the health negative impact of CT imaging. To analyse the influence of the patients x-ray exposure and the received dose, an accurate simulation of the radiation transport through matter is needed. Following this aim, first, this work presents a methodology to acquire sinogram data via detailed Monte Carlo simulations of radiation transport using PenRed. This approach produces more detailed and realistic data than sinograms obtained with mathematical models, which comprises a very important simulation tool for researchers who do not have access to a real CT scanner. Secondly, the simulated data is used to compare two image reconstruction algorithms using few projections, an analytical method (FBP) and an algebraic iterative technique (LSQR). This study has been done on both a mathematical phantom and a real COVID-19 patient data. These ones show that the LSQR maintains a high-quality image even when few projections are used, in contrast to the FBP algorithm, which would allow a significant dose reduction.