Scintillation event position determination in a radiation particle detector

Abstract

A method for determining the position of a scintillation event in a radiation particle detector with multiple scintillator element locations which are configured to emit a burst of photons responsive to a radiation particle being absorbed at the scintillator element location and with a plurality of photosensors (5.1, 5.2, 5.3, 5.4) optically coupled to said scintillator element locations, comprising the steps of determining, for each of the photosensors (5.1, 5.2, 5.3, 5.4), a triggering probability indicative of the probability of said photosensor (5.1, 5.2, 5.3, 5.4) measuring a number of photons that exceeds a predetermined triggering threshold; measuring a photon distribution with the photosensors (5.1, 5.2, 5.3, 5.4) indicative of the number of photons incident on the individual photosensors (5.1, 5.2, 5.3, 5.4); calculating, for each of the scintillator element locations, a likelihood that a scintillation event with a predetermined energy value took place in said scintillator element location based on the measured photon distribution and the triggering probability of each of the photosensors (5.1, 5.2, 5.3, 5.4); and identifying the scintillator element location having the maximum likelihood.

Description

FIELD OF THE INVENTION[0001]The present invention relates to the field of radiation particle detectors. It finds particular application in nuclear imaging systems such as, for example positron emission tomography (PET) scanners for clinical or research studies as well as single photon emission computed tomography (SPECT) scanners.BACKGROUND OF THE INVENTION[0002]In PET scanners pixellated scintillator elements are typically used to convert incident radiation particles to bursts of photons with a wavelength in the UV or visible spectrum. The scintillator elements are typically arranged in a matrix wherein each scintillator element has a base area in the order of 1×1 mm2 to 4×4 mm2. The scintillation events are detected by photosensors coupled to the scintillator elements. State of the art PET scanners use solid-state photosensors, e.g. silicon photomultipliers (SiPMs), typically comprising an array of single photon avalanche diodes (SPADs) being configured to break down responsive to...

AI Technical Summary

Benefits of technology

[0029]According to the present invention, for each of the photosensors, a triggering probability is determined which indicates the probability of said photosensor measuring an amount of light, i.e. a number of photons, that exceeds a predetermined threshold. Photosensors may have a triggering threshold in order to distinguish between measurements that result from scintillation events and measurements that result from dark noise. Typically, the photosensor outputs measurement data when the number of detected photons exceeds the predetermined triggering threshold. The distribution of light on the photosensors is measured. For each of the scintillator element locations a likelihood that a scintillation event with a predetermined energy took place in said scintillator element location is calculated based on the measured light distribution. When calculating the likelihood of an individual scintillator element location the triggering probability of each of the photosensors is additionally taken into account. By calculating the likelihood based on the triggering probability, measured information from photosensors having a relatively low probability of triggering (e.g. due to being deactivated or having a large dead time) is given less importance when calculating the likelihood. Photosensors that show a relatively high probability of triggering are given higher weight when calculating the likelihood. It has been found that the inventive method improves the localization of the scintillation event even when information from some of the photosensors is missing.

[0041]It is further preferred that the step of determining the triggering probability includes determining, in particular measuring, a dependence of the triggering probability on the scintillation event rate and that the likelihoods are additionally calculated based on a measured scintillation event rate. As a consequence, the dependence of the triggering probability on the scintillation event rate can be accounted for when calculating the likelihoods, so as scintillation event localization is further improved. The scintillation event rate can be a mean event rate or an actual event rate.

Problems solved by technology

Since Anger logic relies on information from neighboring photosensors to identify the scintillator element location, Anger logic becomes inaccurate when information of some of the photosensors is missing, e.g. due to the scintillation event happening during a dead time period of a photosensor or due to individual photosensors being inactive for other reasons.

However, known maximum-likelihood methods fail if information of multiple photosensors is missing resulting in deteriorated scintillation event localization.

Images