Predicting Regional Cerebral Blood Flow Using Voxel-Wise Resting-State Functional MRI.

Regional cerebral blood flow (rCBF) is a putative biomarker for neuropsychiatric disorders, including major depressive disorder (MDD). Here, we show that rCBF can be predicted from resting-state functional MRI (rsfMRI) at the voxel level while correcting for partial volume averaging (PVA) artifacts. Cortical patterns of MDD-related CBF differences decoded from rsfMRI using a PVA-corrected approach showed excellent agreement with CBF measured using single-photon emission computed tomography (SPECT) and arterial spin labeling (ASL). A support vector machine algorithm was trained to decode cortical voxel-wise CBF from temporal and power-spectral features of voxel-level rsfMRI time series while accounting for PVA. Three datasets, Amish Connectome Project ( = 300; 179 M/121 F, both rsfMRI and ASL data), UK Biobank ( = 8396; 3097 M/5319 F, rsfMRI data), and Amen Clinics Inc. datasets ( = 372: = 183 M/189 F, SPECT data), were used. PVA-corrected CBF values predicted from rsfMRI showed significant correlation with the whole-brain ( = 0.54, = 2 × 10) and 31 out of 34 regional ( = 0.33 to 0.59, < 1.1 × 10) rCBF measures from 3D ASL. PVA-corrected rCBF values showed significant regional deficits in the UKBB MDD group (Cohen's = -0.30 to -0.56, < 10), with the strongest effect sizes observed in the frontal and cingulate areas. The regional deficit pattern of MDD-related hypoperfusion showed excellent agreement with CBF deficits observed in the SPECT data ( = 0.74, = 4.9 × 10). Consistent with previous findings, this new method suggests that perfusion signals can be predicted using voxel-wise rsfMRI signals. CBF values computed from widely available rsfMRI can be used to study the impact of neuropsychiatric disorders such as MDD on cerebral neurophysiology.