Impact of motion correction on parametric images in PET neuroreceptor studies

The calculation of parametric images in PET studies of neuroreceptors is based on dynamic data which have been recorded over many minutes. It is essential that the subject´s head remains unmoved during the PET scan, otherwise the data may become useless in the worst case. This work studies the degrading of parametric images caused by head movements and improvements which is achieved by an appropriate motion correction. The head movements present in PET neuroreceptor studies cause artifacts in the calculation of parametric images. Whereas the activity images look blurred, the parametric images contain discontinuities especially at the cortex. It is concluded that the linear regression is sensitive to the specific errors present in the dynamic images because of the movements