Performance of a PET detector module utilizing an array of silicon photodiodes to identify the crystal of interaction

Summary form only given, as follows. A multilayer PET (positron emission tomography) detector module consisting of an array of 3 mm square by 30 mm deep BGO (bismuth germanate) crystals coupled on one end to a single photomultiplier tube and on the opposite end to an array of 3 mm square silicon photodiodes has been investigated. The photomultiplier tube provides an accurate timing pulse and energy discrimination for all the crystals in the module, while the silicon photodiodes identify the crystal of interaction. When a single BGO crystal is excited with 511 keV photons, a photodiode signal centered at 500 electrons (e^-) with a width of 375 e^- FWHM at an operating temperature of +25°C is measured. When a four-crystal/photodiode module is excited with a collimated beam of 511 keV photons, the photodiode array correctly identifies the crystal of interaction roughly 50% of the time. This misidentification rate is higher than predicted by signal to noise arguments, and is thought to be due to insufficient beam collimation and Compton interactions in the module. The misidentification rate can be greatly reduced, and an 8×8 crystal/photodiode module constructed, by using thicker depletion layer photodiodes or cooling to 0°C