Head motion evaluation and correction for PET scans with 18F-FDG in the Japanese Alzheimer’s disease neuroimaging initiative (J-ADNI) multi-center study

• Published in: Annals of nuclear medicine 2012-08, Vol.26 (7), p.535-544
• Main Authors: Ikari, Yasuhiko, Nishio, Tomoyuki, Makishi, Yoko, Miya, Yukari, Ito, Kengo, Koeppe, Robert A., Senda, Michio
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Alzheimer Disease - diagnostic imaging; Alzheimer Disease - physiopathology; Alzheimer's disease; Cameras; Cognitive ability; Cortex (cingulate); Female; Fluorodeoxyglucose F18; Head; Head - diagnostic imaging; Humans; Imaging; Imaging, Three-Dimensional; Japan; Male; Medicine; Medicine & Public Health; Middle Aged; Movement; Multimodal Imaging - methods; Neurodegenerative diseases; Neuroimaging; Neuroimaging - methods; Nuclear Medicine; Original Article; Positron emission tomography; Radiology; Tomography, X-Ray Computed
• ISSN: 0914-7187; 1864-6433
• DOI: 10.1007/s12149-012-0605-4

Objective Head motion during 30-min (six 5-min frames) brain PET scans starting 30 min post-injection of FDG was evaluated together with the effect of post hoc motion correction between frames in J-ADNI multicenter study carried out in 24 PET centers on a total of 172 subjects consisting of 81 normal subjects, 55 mild cognitive impairment (MCI) and 36 mild Alzheimer’s disease (AD) patients. Methods Based on the magnitude of the between-frame co-registration parameters, the scans were classified into six levels (A–F) of motion degree. The effect of motion and its correction was evaluated using between-frame variation of the regional FDG uptake values on ROIs placed over cerebral cortical areas. Result Although AD patients tended to present larger motion (motion level E or F in 22 % of the subjects) than MCI (3 %) and normal (4 %) subjects, unignorable motion was observed in a small number of subjects in the latter groups as well. The between-frame coefficient of variation (SD/mean) was 0.5 % in the frontal, 0.6 % in the parietal and 1.8 % in the posterior cingulate ROI for the scans of motion level 1. The respective values were 1.5, 1.4, and 3.6 % for the scans of motion level F, but reduced by the motion correction to 0.5, 0.4 and 0.8 %, respectively. The motion correction changed the ROI value for the posterior cingulate cortex by 11.6 % in the case of severest motion. Conclusion Substantial head motion occurs in a fraction of subjects in a multicenter setup which includes PET centers lacking sufficient experience in imaging demented patients. A simple frame-by-frame co-registration technique that can be applied to any PET camera model is effective in correcting for motion and improving quantitative capability.