Computational protocol: Machine-Learning Classifier for Patients with Major Depressive Disorder: Multifeature Approach Based on a High-Order Minimum Spanning Tree Functional Brain Network

Similar protocols

Protocol publication

[…] The study was carried out in accordance with the recommendations of the medical ethics committee of Shanxi Province (reference number: 2012013). All subjects provided written informed consent in accordance with the Declaration of Helsinki. Twenty-eight healthy subjects and thirty-eight people with MDD underwent rs-fMRI in a 3T scanner (Siemens Trio 3-Tesla scanner, Siemens, Erlangen, Germany). Participant demographic information is shown in .Data collection was completed at the First Hospital of Shanxi Medical University. Radiologists familiar with fMRI performed all scans. During each scan, the participant was asked to relax with their eyes closed and not think about anything in particular but to stay awake and avoid falling asleep. Each scan consisted of 248 contiguous echo-planar imaging (EPI) functional volumes (33 axial slices, repetition time (TR) = 2000 ms, echo time (TE) = 30 ms, thickness/skip = 4/0 mm, field of view (FOV) = 192 × 192 mm, matrix = 64 × 64 mm, and flip angle = 90°). The first 10 volumes in the time series were discarded to account for magnetization stabilization. See Supplemental for detailed scanning parameters.Data preprocessing was performed in SPM8 ( with slice-timing and head-movement corrections. Two samples containing a translation of more than 3.0 mm and rotation of more than 3.0° were excluded from the final analysis of 66 samples. Functional images were normalized using the 12 parameters from the affine transformation and the cosine-based nonlinear transformation from the normalization of the anatomic image to the Montreal Neurological Institute (MNI) space. Additional normalization of the functional data sets to the SPM8 EPI template was then performed, and the data were resampled to a voxel size of 3 × 3 × 3 mm using a sinc interpolation. No smoothing kernel was applied to limit spurious local connectivity between voxels. Finally, we performed linear detrending and band-pass filtering (0.01–0.10 Hz) to reduce the effects of low-frequency drift and high-frequency physiological noise. Then, for each subject, the brain space of the fMRI images was parcellated into 90 regions of interest (ROIs) (45 in each hemisphere) based on the automated anatomical labeling (AAL) atlas [], and each region was defined as a node in the network. Each regional mean time series was regressed against the average cerebral spinal fluid and white-matter signals as well as the six parameters from motion correction. The residuals of these regressions constituted the set of regional mean time series used for undirected graph analysis. […]

Pipeline specifications

Software tools SPM, AAL
Applications Magnetic resonance imaging, Functional magnetic resonance imaging
Organisms Homo sapiens
Diseases Brain Diseases