Optimization of multipinhole arrangements for quantitative mouse brain SPECT by Monte Carlo simulation

This study investigated an overlapping multipinhole SPECT system and sought to optimize the arrangement of pin holes for quantitative mouse brain imaging. An analytical Monte Carlo simulation method took into account the pinhole aperture function, edge penetration and intrinsic detector resolution. A model-based 3D OSEM algorithm was developed and used to reconstruct images. Three pinhole models (ideal pinhole, geometric pinhole and real pinhole) were implemented in the reconstruction algorithm. Results showed that the compensation for pinhole aperture and edge penetration had improved resolution. Study using a resolution (cold-rod) phantom with various pinhole diameters (0.3, 0.6, 0.9 and 1.2 mm) and various pinhole numbers (1, 4, 9 and 16) showed that the multipinhole had improved the trade-off between sensitivity and resolution. Quantitative study of cerebral Dopamine Transporter (DAT) binding using a mouse brain phantom and 1.0-mm pin holes showed that the multipinhole had outperformed single pinholes by lowering the variance of quantification. Results also suggest that approximately 10 pinholes is a good choice for the mouse brain study.