Test-retest reliability of task-based and resting-state blood oxygen level dependence and cerebral blood flow measures

• Published in: PloS one 2018-11, Vol.13 (11), p.e0206583-e0206583
• Main Authors: Holiga, Štefan, Sambataro, Fabio, Luzy, Cécile, Greig, Gérard, Sarkar, Neena, Renken, Remco J, Marsman, Jan-Bernard C, Schobel, Scott A, Bertolino, Alessandro, Dukart, Juergen
• Format: Article
• Language: English
• Subjects: Activation; Adult; Biology and Life Sciences; Blood; Blood flow; Brain - anatomy & histology; Brain - blood supply; Brain - physiology; Cerebral blood flow; Cerebrovascular Circulation; Computer and Information Sciences; Connectivity; Correlation analysis; Dependence; Disease; Domains; Female; Functional magnetic resonance imaging; Functional Neuroimaging - statistics & numerical data; Humans; Imaging, Three-Dimensional; Information dissemination; Longitudinal studies; Magnetic resonance; Magnetic resonance imaging; Magnetic Resonance Imaging - statistics & numerical data; Male; Medical imaging; Medicine and Health Sciences; Methods; Models, Neurological; Neurosciences; NMR; Nuclear magnetic resonance; Oxygen; Oxygen - blood; Reliability analysis; Reproducibility; Reproducibility of Results; Research and Analysis Methods; Rest - physiology; Social Sciences; Spin labeling; Spin Labels; Studies; Task Performance and Analysis; Young Adult
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0206583

Despite their wide-spread use, only limited information is available on the comparative test-retest reliability of task-based functional and resting state magnetic resonance imaging measures of blood oxygen level dependence (tb-fMRI and rs-fMRI) and cerebral blood flow (CBF) using arterial spin labeling. This information is critical to designing properly powered longitudinal studies. Here we comprehensively quantified and compared the test-retest reliability and reproducibility performance of 8 commonly applied fMRI tasks, 6 rs-fMRI metrics and CBF in 30 healthy volunteers. We find large variability in test-retest reliability performance across the different tb-fMRI paradigms and rs-fMRI metrics, ranging from poor to excellent. A larger extent of activation in tb-fMRI is linked to higher between-subject reliability of the respective task suggesting that differences in the amount of activation may be used as a first reliability estimate of novel tb-fMRI paradigms. For rs-fMRI, a good reliability of local activity estimates is paralleled by poor performance of global connectivity metrics. Evaluated CBF measures provide in general a good to excellent test-reliability matching or surpassing the best performing tb-fMRI and rs-fMRI metrics. This comprehensive effort allows for direct comparisons of test-retest reliability between the evaluated MRI domains and measures to aid the design of future tb-fMRI, rs-fMRI and CBF studies.