Abstract
<jats:p> Purpose: Attenuation correction (AC) of PET images is essential for accurate quantification. Brain PET studies comprising simultaneous EEG (PET <jats:sub>EEG</jats:sub> ) may suffer from metal artifacts in CT images (CT <jats:sub>EEG</jats:sub> ), or improper correction when electrodes are not present in the CT (CT <jats:sub>0</jats:sub> ). As these influences are not well-characterized, we aim to compare metal artifact reduction (MAR) techniques for CT <jats:sub>EEG</jats:sub> images, and evaluate differences between attenuated-corrected PET <jats:sub>EEG</jats:sub> using CT <jats:sub>0</jats:sub> and CT <jats:sub>EEG</jats:sub> with MAR, synthetically placed electrodes (CT <jats:sub>EEG-synth</jats:sub> ) and extended Hounsfield unit (HU) range. Methods: 19 healthy participants underwent two total-body PET/CT scans with [ <jats:sup>18</jats:sup> F]FDG, with and without 32 EEG scalp electrodes, respectively. We evaluated five MARs to reduce streaks caused by the EEG electrodes in the CT <jats:sub>EEG</jats:sub> . Finally, CT <jats:sub>0</jats:sub> , CT <jats:sub>EEG</jats:sub> with (CT <jats:sub>EEG-iMAR-Ext</jats:sub> ) and without extended HU range (CT <jats:sub>EEG-iMAR</jats:sub> ) and CT <jats:sub>EEG-synth</jats:sub> were used to perform attenuation correction of PET <jats:sub>EEG</jats:sub> . We compared our results to PET <jats:sub>0</jats:sub> /CT <jats:sub>0</jats:sub> scan using relative differences. Results: CT <jats:sub>EEG</jats:sub> and CT <jats:sub>EEG-iMAR</jats:sub> showed the smallest differences to CT <jats:sub>0</jats:sub> . PET <jats:sub>EEG/CTEEG-iMAR-Ext</jats:sub> exhibited the lowest differences to PET <jats:sub>0</jats:sub> /CT <jats:sub>0</jats:sub> (average bias across all regions of -0.46%), followed by similar performance of PET <jats:sub>EEG</jats:sub> /CT <jats:sub>EEG-iMAR</jats:sub> (-0.73%) and PET <jats:sub>EEG</jats:sub> /CT <jats:sub>EEG</jats:sub> (-0.76%). Conversely, PET <jats:sub>EEG</jats:sub> /CT <jats:sub>0</jats:sub> demonstrated the largest average differences (-1.81%), with values reaching -2.71% in the parietal lobe. These differences were consistent across subjects, yielding significant effects in most of the brain (p <jats:sub>FWE</jats:sub> < 0.05). CT <jats:sub>EEG-synth</jats:sub> performed not as good as CT <jats:sub>EEG</jats:sub> (-1.21%). Conclusions: CT <jats:sub>EEG</jats:sub> with extended HU range is most suitable for attenuation correction of PET <jats:sub>EEG</jats:sub> images, with MAR correction offering little additional improvement. </jats:p>