Premium
Sci‐Sat AM(1): Imaging‐04: Respiratory errors in cardiac PET/CT with manual alignment of the CT image
Author(s) -
Wells RG
Publication year - 2008
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.2965988
Subject(s) - nuclear medicine , expiration , image registration , breathing , cardiac pet , medicine , positron emission tomography , correction for attenuation , medical imaging , pet ct , respiratory system , radiology , computer science , computer vision , image (mathematics) , anatomy
Respiratory motion can produce misregistration errors between CT and PET images in cardiac PET/CT imaging. The objective of this study was to determine if manual registration of a single‐phase end‐expiration CT scan to the PET image would eliminate respiratory‐induced artifacts. Listmode data from 71 cardiac PET patient scans were rebinned into a 8‐frame respiratory‐gated image series based on a respiratory trigger signal obtained with an optical tracking system. CT‐based attenuation correction (AC) was performed after registering the CT image with the mean position of the PET images. The 8 phases of the gated PET study were coregistered and the breathing motion was measured. Images from end‐inspiration and end‐expiration were compared to assess the effect of motion. Studies in which the motion was >8mm were reconstructed again, with the CT scan aligned to end‐expiration or end‐inspiration, to determine if phase‐specific registration could reduce the residual errors. The motion was found to be greatest in the axial direction (mean 4.1mm +\− 1.8mm) and 4 Rb stress studies (17%) had motion >8mm. The maximum displacement during breathing was greater for Rb‐stress imaging (<15mm) than for resting (<7.5mm) or NH3‐stress (<5.4mm) imaging. No significant differences were noted between the respiratory phases of the rest studies. Errors in myocardial radiotracer uptake of up to 35% were noted between end‐inspiration and end‐expiration for studies with >8mm of motion. Phase‐specific registration of the CT reduced the extent of the errors but did not fully resolve them, suggesting that more sophisticated AC is required.