Artificial neural network based positioning algorithm for PEM imaging

The objective of this work is to improve the resolution and linearity of a positron emission mammography (PEM) detector using a continuous LSO scintillation crystal, by using an algorithm based on artificial neural networks. Specifically, the effect of crystal thickness on the spatial resolution and bias of the detector is investigated. As a result, spatial resolution and bias is improved for all crystal thicknesses by using an artificial neural network based positioning algorithm compared to Anger algorithm and it is seen that without sacrificing resolution, thick scintillation crystals can be used to improve sensitivity of the PEM detector.