A novel clustering approach to positron emission particle tracking

•A new approach to positron emission particle tracking is presented.•Using G-means clustering, multiple particle tracking is enabled.•Multiple and variable number particle tracking is explored.•Parametric dependence of algorithm is studied.

A novel approach to positron emission particle tracking is presented based on determining regions of space with high density of line of response crossing via clustering. The method is shown to be able to accurately track multiple particles in systems where the number of particles is unknown and in which particles can enter and leave the field of view of the scanning system. This method is explored in various environments and its parametric dependence is studied.