Publicação
GATE Model of a SPECT-CT equipment for breast cancer diagnosis
| Resumo: | Dedicated imaging systems for breast cancer imaging have been comprehensively studied over the past decade. However, since they comprise only one imaging modality these systems are only able to provide either anatomical or functional information of the object of interest. The aim of this work was to simulate and evaluate the implementation of a dedicated system that would be able to extract information regarding both systems in order to provide a complementary diagnostic tool which could be used in inconclusive diagnosis cases. The proposed solution was to develop two dedicated systems. A dedicated breast computed tomography (DBCT) system that would provide anatomical information and a dedicated single photon emission mammography (SPEM), using convergent collimators, that would retrieve functional information. To create a computer model of this multimodality system, Monte Carlo (MC) simulations were conducted with Geant4 Application for Emission Tomography (GATE) using a simple breast cylindrical phantom with 35 mm radius and 150 mm height, which had 5 mm radius spherical masses composed of aluminum inside it. For DBCT, the MC simulations were acquired with a PaxScan A® 2520D/CL Amorphous Silicon Digital X-Ray Imager with total dimensions of 192x242x4 mm and 0.508x0.508x4 mm pixels, over 16 projections covering 180º of the phantom, extended to 360º exploiting its cylindrical symmetry. Inside the phantom were placed 5 tumour masses equidistantly along two axis with a tumour mass at the center. For SPECT, the MC simulations were performed using a dual-head SPECT scanner designed by Dr. Ricardo Capote with 64.0x151.2x188.5 mm as dimensions with pixelated LYSO crystals of 20x2x2 mm and convergent collimators with the same phantom, but with the same 5 mm radius tumour masses placed solely in the x axis. The MC simulations were conducted in a computer cluster with 4 executions machines. The projections resultant of the simulations were reconstructed using different algorithms. For DBCT it was used an analytical method of filtered backprojection (FBP) and for SPEM it was used an iterative maximum likelihood expectation maximization (MLEM) algorithm. To validate the results two non-absolut metrics were calculated to make a relative evaluation of the image quality results. These metrics were only applied to DBCT, since the results obtained for SPEM were not as expected. Contrast and contrast to noise ratio demonstrated that the image quality degrades from the center to the periphery of the DBCT detector. In conclusion, the acquired results demonstrated the feasibility of breast dedicated systems to, especially for the DBCT system, which yielded the best results, but further development need to be pursued in order to take the most potential of the developed systems which have potential for being used in future studies with more complex and realistic conditions and voxelised phantoms. |
|---|---|
| Autores principais: | Monteiro, André Miguel Martins Costa Correia |
| Assunto: | Cancro da mama Tomografia computorizada Mamogafia por emissão de Fotão Único Reconstrução de imagem Simulações de Monte Carlo Geant4 application for emission tomography Filtered back projection Maximum likelihood expectation maximization Contraste Rácio entre o contraste e o ruído Teses de mestrado - 2016 |
| Ano: | 2016 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | Dedicated imaging systems for breast cancer imaging have been comprehensively studied over the past decade. However, since they comprise only one imaging modality these systems are only able to provide either anatomical or functional information of the object of interest. The aim of this work was to simulate and evaluate the implementation of a dedicated system that would be able to extract information regarding both systems in order to provide a complementary diagnostic tool which could be used in inconclusive diagnosis cases. The proposed solution was to develop two dedicated systems. A dedicated breast computed tomography (DBCT) system that would provide anatomical information and a dedicated single photon emission mammography (SPEM), using convergent collimators, that would retrieve functional information. To create a computer model of this multimodality system, Monte Carlo (MC) simulations were conducted with Geant4 Application for Emission Tomography (GATE) using a simple breast cylindrical phantom with 35 mm radius and 150 mm height, which had 5 mm radius spherical masses composed of aluminum inside it. For DBCT, the MC simulations were acquired with a PaxScan A® 2520D/CL Amorphous Silicon Digital X-Ray Imager with total dimensions of 192x242x4 mm and 0.508x0.508x4 mm pixels, over 16 projections covering 180º of the phantom, extended to 360º exploiting its cylindrical symmetry. Inside the phantom were placed 5 tumour masses equidistantly along two axis with a tumour mass at the center. For SPECT, the MC simulations were performed using a dual-head SPECT scanner designed by Dr. Ricardo Capote with 64.0x151.2x188.5 mm as dimensions with pixelated LYSO crystals of 20x2x2 mm and convergent collimators with the same phantom, but with the same 5 mm radius tumour masses placed solely in the x axis. The MC simulations were conducted in a computer cluster with 4 executions machines. The projections resultant of the simulations were reconstructed using different algorithms. For DBCT it was used an analytical method of filtered backprojection (FBP) and for SPEM it was used an iterative maximum likelihood expectation maximization (MLEM) algorithm. To validate the results two non-absolut metrics were calculated to make a relative evaluation of the image quality results. These metrics were only applied to DBCT, since the results obtained for SPEM were not as expected. Contrast and contrast to noise ratio demonstrated that the image quality degrades from the center to the periphery of the DBCT detector. In conclusion, the acquired results demonstrated the feasibility of breast dedicated systems to, especially for the DBCT system, which yielded the best results, but further development need to be pursued in order to take the most potential of the developed systems which have potential for being used in future studies with more complex and realistic conditions and voxelised phantoms. |
|---|