Abstract: Whole-body Positron Emission Tomography (PET) scanners are required in order to span large Fields of View (FOV). Therefore, reaching the sensitivity and spatial resolution required for early stage breast tumor detection is not straightforward. MAMMI is a dedicated breast PET scanner with a ring geometry designed to provide PET images with a spatial resolution as high as 1.5 mm, being able to detect small breast tumors ( < 1 cm). The patient lays down in prone position during the scan, thus making possible to image the whole breast, up to regions close to the base of the pectoral without the requirement of breast compression. Attenuation correction (AC) for PET data improves the image quality and the quantitative accuracy of radioactivity distribution determination. In dedicated, high resolution breast cancer scanners, this correction would enhance the proper diagnosis in early disease stages. In whole-body PET scanners, AC is usually taken into account with the use of transmission scans, either by external radioactive rod sources or by Computed Tomography (CT). This considerably increases the radiation dose administered to the patient and time needed for the exploration. In this work we propose a method for breast shape identification by means of PET image segmentation. The breast shape identification will be used for the determination of the AC. For the case of a specific breast PET scanner the procedure we propose should provide AC similar to that obtained by transmission scans as we take advantage of the breast anatomical simplicity. Experimental validation of the proposed approach with a dedicated breast PET prototype is also presented. The main advantage of this method is an important dose reduction since the transmission scan is not required.

Abstract: A method to improve light collection efficiency of gamma-ray imaging detectors by using retroreflector arrays has been tested, simulations of the behaviour of the scintillation light illuminating the retroreflector surface have been made. Measurements including retroreflector arrays in the setup have also been taken. For the measurements, positron emission tomography (PET) detectors with continuous scintillation crystals have been used. Each detector module consists of a continuous LSO-scintillator of dimensions 49x49x10 mm(3) and a H8500 position-sensitive photo-multiplier (PSPMT) from Hamamatsu. By using a continuous scintillation crystal, the scintillation light distribution has not been destroyed and the energy, the centroids along the x- and y-direction and the depth of interaction (DOI) can be estimated. Simulations have also been run taking into account the use of continuous scintillation crystals. Due to the geometry of the continuous scintillation crystals in comparison with pixelated crystals, a good light collection efficiency is necessary to correctly reconstruct the impact point of the gamma-ray. The aim of this study is to investigate whether micro-machine retro-reflectors improve light yield without misestimation of the impact point. The results shows an improvement on the energy and centroid resolutions without worsening the depth of interaction resolution. Therefore it can be concluded that using retroreflector arrays at the entrance side of the scintillation crystal improves light collection efficiency without worsening the impact point estimation.

Abstract: In medicine, computed tomographic images are reconstructed from a large number of measurements of X-ray transmission through the patient (projection data). The mathematical model used to describe a computed tomography device is a large system of linear equations of the form AX = B. In this paper we propose the QR decomposition as a direct method to solve the linear system. QR decomposition can be a large computational procedure. However, once it has been calculated for a specific system, matrices Q and R are stored and used for any acquired projection on that system. Implementation of the QR decomposition in order to take more advantage of the sparsity of the system matrix is discussed.