Simulação Monte Carlo do processo de aquisição de imagens de um tomógrafo de dupla energia

Abstract

Dual Energy (DE) Computed Tomography (CT) is one of the fields of tomographic images that has evolved rapidly during the last years. DECT uses two Xray spectra to irradiate patients It is capable to differentiate materials based on its elementary composition. Despite being similar to standard CT devices, DECT devices require the development of specific tools that allow the study of their image properties. The objective of this work was to build a modelled CT system capable of producing images similar to those obtained in real DECT devices. The modelled CT would also allow exploring the image properties of test materials before their physical construction. This work presents the simulation of the acquisition process of a DECT device that works with rapid kV switching, the GE Discovery CT 750 HD. The simulated geometry was based on a device currently available at the InRad (Institute of Radiology of the Faculty of Medicine of the University of São Paulo). The simulations were carried out using the PENELOPE/penEasy Monte Carlo code, which simulates radiation transport through the materials and detectors. A comparison between the images obtained in the real device and from simulations is also presented. To do so, a real phantom was prepared to be imaged and an equivalent system was simulated. The phantom contained inserts with concentrations of iodine and calcium. The images were acquired and reconstructed according to the possibilities of the real CT device. Standard, material concentration and virtual monoenergetic images were acquired from both, the real CT device and simulations. The Projection{Based BMD method was implemented using the mass attenuation coefficients of water and iodine. Then, material concentration images of water and iodine were obtained. The iodine concentrations estimated from the images agreed with the expected values in both real device and simulated images. Beam hardening artefacts were observed in the simulated material concentration images. Monoenergetic images were obtained for different energies. Such images were obtained as a superposition of the concentration images of water and iodine, weighed by their respective mass attenuation coefficient. It was verified that in the simulated and real device images, at high energies, the water concentration image predominated in the monoenergetic images, producing images that presented the iodine cavities as less attenuating than water. In contrast, at low energies, the predominant component of the monoenergetic images was the iodine concentration image. Contrast Noise Ratio (CNR) was analysed in the monochromatic images as a function of energy. Simulated and real device CNR curves exhibited similarities in their shape but with a different scale due to their difference in noise. It was possible to conclude that the simulated DECT model presented qualitative results similar to the obtained in the real device. The modelled CT system permits exploring the image features with different materials and compositions. It could also be used as a didactic t...

Aplicações de métodos para validação de materiais radiologicamente equivalentes a tecido humano e à água

“…A partir das densidades dos materiais base 1 e 2 , a imagem virtual monoenergética de energia base pode ser gerada a partir da equação: (Puerto, 2018). Ao longo do arco de rotação de 360°, os dados de projeção são adquiridos duas vezes (uma para baixa e outra para alta tensão) na mesma posição, aproximadamente, da fonte de raios X. O cintilador de oxissulfeto de gadolíneo utilizado nos equipamentos da GE permite que a aquisição dos dados de projeção seja quase simultânea (com intervalo de troca menor que 0,5 milissegundos) (Russo, 2017).…”