ABSTRACT
PET imaging is regarded as a quantitative imaging modality that provides an accurate measure of physiological function within a patient. However, several physical effects, as well as data collection procedures, can lead to bias and artifacts in the reconstructed image. Corrections to PET data are therefore needed for an accurate clinical interpretation of the patient image, as well as to obtain physiological information from the static (e.g., standardized uptake value [SUV]) or dynamic (kinetic modeling) image. Corrections to PET data can be broadly split into three categories depending on where in the data acquisition and image generation chain the correction is applied. The three categories are (1) calibrations at the detector level, which affect the quality of collected events; (2) corrections as part of image reconstruction, which remove or reduce bias or nonuniformities in the image; and (3) calibrations after image reconstruction, which allow for measuring activity levels in the patient.
