Respiration artifacts in whole-body (18)F-FDG PET/CT studies with combined PET/CT tomographs employing spiral CT technology with 1 to 16 detector rows.

Abstract

Co-registration accuracy in combined whole-body (WB) PET/CT imaging is impaired by respiration-induced mismatches between the CT and the PET. Furthermore, PET/CT misregistration may bias the PET tracer distribution following CT-based attenuation correction (CT-AC). With the introduction of multi-row CT technology of up to 16 detector rows into PET/CT designs, we investigated the incidence of respiration artifacts in WB PET/CT examinations of patients who were unable to follow any breath-hold instructions. A total of 80 WB studies from six international sites operating PET/CT tomographs with 1-, 2-, 4-, 6-, 8-, and 16-row spiral CT were included. PET/CT examinations were acquired with the patients breathing normally during both the CT and the PET scan, and CT-AC was performed routinely. All studies were anonymized and reviewed independently by three radiologists and three nuclear medicine specialists. We report the frequency and severity of artifacts on CT and PET for the thorax and the abdomen. In WB PET/CT imaging of normally breathing patients, significant gains in diagnostic image quality can be expected from employing CT technology with six or more detector rows. In our study, fewer PET images appear distorted than corresponding CT images, which is due to the limited propagation of only mild CT image artifacts after the resampling of the CT-based attenuation map during CT-AC. In whole-body PET/CT imaging of normally breathing patients, respiration-induced artifacts are reduced in both magnitude and prominence for PET/CT systems employing CT components of six or more detector rows.