Background: Positron emission tomography (PET) attenuation correction (AC) using computed tomography (CT) can be affected by respiratory motion: hi-speed CT captures 1 point of the respiratory cycle while PET emission data averages many cycles. We quantified the changes in apparent myocardial uptake due to this respiratory-induced CT attenuation mismatch.
Methods: Twenty-two patients undergoing fluorine-18 fluorodeoxyglucose (FDG) PET/CT received 3 sequential CT scans at normal resting end-inspiration (CT(INSPIR)), ending expiration (CT(EXPIR)), and at midvolume between end-expiration and end-inspiration (CT(MIDVOL)). A pneumotachometer measured absolute changes in lung volume. Seven subjects also underwent a 3-minute transmission scan with a 68Ge rotating rod source (RRS). The PET emission data set was reconstructed up to 4 times using CT(EXPIR), CT(INSPIR), CT(MIDVOL), and RRS AC maps. Relative heart position and cardiac uptake was measured for each CT attenuation correction.
Results: Respiratory motion produced marked changes in global and regional myocardial uptake. Changes were large in the lateral and anterior regions at the lung-soft tissue interface (up to 30% using CT(INSPIR) compared to CT(EXPIR) for AC) and smaller in the septal region (10% or less). Data corrected with CT(EXPIR) agreed best with the RRS.
Conclusion: Respiratory effects can introduce large inhomogeneities in apparent myocardial uptake when CT is used for attenuation correction.