A flexible acquisition system for modular dual head Positron Emission Mammography Giancarlo Sportelli, Francesca Attanasi, Nicola Belcari, Giovanni Franchi, Pedro Guerra, Member IEEE, Sascha Moehrs, Valeria Rosso, Andrés Santos, Senior Member IEEE, Franco Spinella, Sara Vecchio and Alberto Del Guerra, Senior Member IEEE Abstract– We have developed a dedicated scanner for Positron Emission Mammography, equipped with a new detection architecture that enhances its flexibility and reduces dead time. The scanner is going to use Luthetium based scintillators, which offer good detection efficiency, and a novel modular acquisition system, capable of sustaining the high scintillation rate and being less sensitive to background radiation. The final goal is the construction of an instrument able to provide an early diagnosis and to improve the effectiveness of follow-up studies for smaller tumours with respect to those studied with present clinical equipment (e.g. PET, SPECT o scintigraphy) so as to be able to visualize and characterize breast lesions with diameters < 5 mm. I. INTRODUCTION FTER 15 years since the introduction of Positron Emission Mammography (PEM) by Thompson et al [1], the interest on the application of the PET technique for breast cancer imaging is still high. A PEM system is usually made of a pair of planar detectors that can compress the breast. A number of PEM prototypes has been proposed with this geometry. A system utilizing two sets of scanning planar detectors has also been developed [2]. The advent of multi anode flat panel photomultiplier tubes (such as Hamamatsu H8500) has helped the development of large area planar detectors. Breast tomography is also possible with detectors that are large enough (e.g. 15 cm × 15 cm) [3,4]. With this geometry the system has strong count rate requirements. With the dual head planar geometry each detector head subtends a large solid angle for detecting annihilation γ-rays. Hence it is exposed to a large γ-ray flux both coming from the breast fraction within the FOV and from Manuscript received November 14, 2009. This work was partially supported by the MIUR, in the framework of the PRIN2006 program, “Development of a dedicated PET tomograph for breast cancer imaging”, the “Consejería de Educación de la Comunidad de Madrid” and the European Social Fund. G. Sportelli and A. Santos are with the Biomedical Image Technologies group, E.T.S.I.T., Universidad Politécnica de Madrid, 28040 Madrid, Spain, (email: [gsportelli, andres]@die.upm.es) and the Research Center in Bioengineering, Biomaterials and Nanomedicine, 50018 Zaragoza, Spain. N. Belcari, F.Attanasi, S. Moehrs, V. Rosso, S. Vecchio and A. Del Guerra are with the Department of Physics “E. Fermi”, University of Pisa, 56127 Pisa, Italy, (email:) and INFN – Sezione di Pisa, 56127 Pisa, Italy, (email: [belcari, sascha.moehrs, alberto.delguerra]@df.unipi.it, [francesca.attanasi, valeria.rosso, franco.spinella, sara.vecchio]@pi.infn.it). P. Guerra is with the Research Center in Bioengineering, Biomaterials and Nanomedicine, 50018 Zaragoza, Spain, (email: pguerra@ciber-bbn.es). F. Spinella is with INFN – Sezione di Pisa, 56127 Pisa, Italy, (email: franco.spinella@pi.infn.it). G. Franchi is with AGE Scientific, Capezzano Pianore, 55041 Lucca, Italy, (email: gfranchi@agescientific.com). regions outside the FOV. In particular a strong background is expected to come from the tracer uptake in the thorax region, for instance in the heart myocardium. In this way a high single count rate can be expected in each detector. For this reason the electronic pile-up could be a strong limitation if the detector head is read out as a single detector. In addition, the minimization of the electronic dead time is critical for the maximization of the actual system efficiency. II. SCANNER CONCEPT The developed system is a dual head Positron Emission Mammograph with planar detectors, whose active area in the current version is about 10 cm × 10 cm. Each head is made up of a matrix with 2 × 2 independent detector modules. The modules are comprised of a square 64 anodes photomultiplier tube (Hamamatsu H8500) coupled to a matrix of 23 × 23 LYSO scintillating crystals (1.9 mm × 1.9 mm × 16 mm pixel dimensions, with a 2.0 mm pitch). Figure 1. The PEM motherboard with four plugged DAQs (left) and the coincidence board (right). The division of the scintillating matrix into submatrices implies a loss of active area due to the dead space between the modules. With our system the dead space is about 6 mm. In this way we have a geometrical efficiency loss wich is about 12% with respect to a solution based on a large scintillating matrix, read out by a four (2 × 2) tubes assembly. However, this geometrical efficiency loss can be largely compensated by a gain in count rate characteristics. For this reason we have developed a flexible and expandable acquisition system specifically designed to work with modular detectors. A first advantage of this modular approach consists in the spreading of the coincidence events, and then the data flow, among the modules, thus reducing both system dead time and the probability of electronic pile-up. In fact, if the A 2009 IEEE Nuclear Science Symposium Conference Record M09-332 9781-4244-3962-1/09/$25.00 ©2009 IEEE 3395