Performance of a DOI-encoding small animal PET system with monolithic scintillators M. Carles a,n , Ch.W. Lerche b , F. Sa ´ nchez a , A. Orero a , L. Moliner a , A. Soriano a , J.M. Benlloch a a Instituto de Instrumentacio ´n para Imagen Molecular(I3M), Centro mixto CSIC-Universitat Polit  ecnica de Val  encia- IEMAT, Camino de Vera s/n, 46020 Valencia, Spain b Department X-Ray Imaging Systems, Philips Research Europe, Weisshausstrasse 2, D-52066 Aachen, Germany article info Available online 15 November 2011 Keywords: DOI PET Spatial resolution Continuous scintillators abstract PET systems designed for specific applications require high resolution and sensitivity instrumentation. In dedicated system design smaller ring diameters and deeper crystals are widely used in order to increase the system sensitivity. However, this design increases the parallax error, which degrades the spatial image resolution gradually from the center to the edge of the field-of-view (FOV). Our group has designed a depth of interaction(DOI)-encoding small animal PET system based on monolithic crystals. In this work we investigate the restoration of radial resolution for transaxially off-center sources using the DOI information provided by our system. For this purpose we have designed a support for point like sources adapted to our system geometry that allows a spatial compression and resolution response study. For different point source radial positions along vertical and horizontal axes of a FOV transaxial plane we compare the results obtained by three methods: without DOI information, with the DOI provided by our system and with the assumption that all the g-rays interact at half depth of the crystal thickness. Results show an improvement of the mean resolution of 10% with the half thickness assumption and a 16% achieved using the DOI provided by the system. Furthermore, a 10% restoration of the resolution uniformity is obtained using the half depth assumption and an 18% restoration using measured DOI. & 2011 Elsevier B.V. All rights reserved. 1. Introduction In recent years there has been an increasing development of PET systems designed for specific applications, such as small animal imaging which has a significant role in the context of biomedical research. The relative small size of the objects under study in small animal imaging requires instruments with a finer spatial resolution and also higher sensitivity. In current small animal PET scanners a compromise between sensitivity and spatial resolution is made due to the parallax error caused by the lack of DOI. Small animal imaging employs smaller ring diameter that means higher sensitivity, lower cost and smaller photon non- collinearity effect. Moreover, longer crystals are chosen because of the higher sensitivity. However, these two features of the system design increase the degree of the DOI problem. For this reason there has been an increasing focus on detector designs with depth-encoding ability. Several methods have been pro- posed to avoid the so-called parallax error which degrades the spatial resolution from the center to the edge [1]. Some of them are the use of a phoswich design [2], the connection of two ends of the crystal to separate photosensor [3], or the use of light absorbing material around each crystal [4]. However, all methods require either extensive work with the detector crystal or addi- tional electronics, so it increases significantly the costs and the spatial resolution is still not satisfactory. By contrast, with the method of DOI measurement developed by our group [5], detector modifications do not substantially increase its cost. This method takes advantage of the strong correlation between the width of the undisturbed light distribu- tion in continuous crystals and the g-ray’s DOI. It consists of an inexpensive modification of the position-sensitive proportional resistor chain (DPC) readout [6,7] that allows the simultaneous measurement of the overall released amount of scintillation light, the centroids along the x and y spatial directions of the light distribution, and its bare second moment. The standard deviation, s light , can be computed from the digitized centroids and the bare second moment without any energy or spatial resolution degrada- tion. Furthermore, since nearly all computation steps are performed by analogical electronics, the computation of the DOI is very fast and their implementation inexpensive. In our previous work [8], an improvement in spatial resolution and decreasing compression was reported. That work presented the results obtained with and without our DOI correction method, Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/nima Nuclear Instruments and Methods in Physics Research A 0168-9002/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2011.11.021 n Corresponding author. E-mail address: montcar@ific.uv.es (M. Carles). Nuclear Instruments and Methods in Physics Research A 695 (2012) 317–321