Maximum likelihood based positioning and energy correction for pixelated solid state PET detectors

As part of a preclinical MR compatible small animal PET scanner, a compact block detector for gamma-ray detection has been developed. The block detector consists of a LYSO crystal array with 22 × 22 pixels and an array of 8 × 8 Silicon photomultiplier. An intermediate light guide enables positioning by light-sharing. Due to dark noise and variations of the photomultiplier gain and the light collection efficiency, the crystal identification and energy computation using the center of gravity method is erroneous. We propose an alternative positioning scheme that is based on maximizing the likelihood of the scintillation events. The method directly gives the index of the active crystal pixel and allows also to correct the registered gamma-ray energy. First studies show that the all-over energy resolution for one block detector is enhanced from about 28% to 15%. For the used data set, energy correction with the presented method clearly outperforms energy correction based on the center of gravity method.