Because positron emission tomography (PET) provides biochemical information in vivo with the sensitivity at the sub-pico-molar level, pre-clinical research using PET plays an important role in biological and pharmaceutical sciences. However, small animal imaging by PET has been challenging with respect to spatial resolution and sensitivity due to the small volume of the imaging objects. A DOI-encoding technique allows for pre-clinical PET to simultaneously achieve high spatial resolution and high sensitivity. Thus many DOI-encoding methods have been proposed. In this paper we describe why DOI measurements are important, what is required in DOI-encoding designs, and how to extract DOI information in scintillator-based DOI detectors. Recently, there has been a growing interest in DOI measurements for TOF PET detectors to correct time walk as a function of DOI position. Thus, the DOI-encoding method with a high time performance suitable for TOF detectors is now required. The requirements to improve the time resolution in DOI detectors are discussed as well.Keywords Positron emission tomography (PET), Depth of interaction (DOI), Multi-layer detector, Dual-ended readout, Single-ended readout
THE IMPORTANCE OF DOI MEASUREMENT
Simultaneous improvement in spatial resolution and sensitivityThe PET is an important pre-clinical imaging technique because PET provides biochemical information down to the sub-pico-molar level in vivo [1]. Thus, the development of pre-clinical PET scanners has been actively promoted. The major focus of the development is to obtain a similar quality in small animal images as human images, while the size of the imaging subject decreases and the amount of radiopharmaceutical injected into the small animals is limitedTo obtain a similar level of detail and SNR in mouse images as human images, spatial resolution and sensitivity should be increased. That is why many PET instrument researchers have been attempting for the pre-clinical PET to simultaneously achieve high spatial resolution and high sensitivity. The pre-clinical PET systems therefore have been designed by using very long and narrow crystals with small diameter ring geometry. However, such system structures cause the parallax error to become lager when providing no depth-of-interaction (DOI) information within crystals (general PETs can measure no DOI information), and bring the degradation of the radial resolution in the peripheral field of view (FOV). That is because the reconstruction algorithm, when drawing a line of response (LOR) without DOI information, usually assigns the interaction positions over all depths within a crystal to a single position (i.e. the center position on the front of the interacted crystals).