Performance of an image-based motion compensation algorithm for the hrrt: A striatum-phantom study with true motion

High Resolution Research Tomograph (HRRT, Siemens) is a dedicated human brain PET scanner with nearly isotropic and shift invariant reconstructed resolution of 2.5 mm (FWHM) as measured using a point source and reconstruction settings of NEMA standard [1]. Spatial resolution can be however significantly degraded if the imaged object is moving during PET scanning resembling impact of partial volume effect (PVE). Head motion is particularly detrimental in brain studies investigating e.g. tracer receptor binding in the small brain nuclei, and their displacement thereof. In fact, decreased tracer uptake as depicted by time-activity curve (TAC) data may often be due to head motion rather than inhibition of tracer binding (c.f. [2]).