Time-delay structure predicts clinical scores for patients with disorders of consciousness using resting-state fMRI

• Published in: NeuroImage clinical 2021-01, Vol.32, p.102797-102797, Article 102797
• Main Authors: Cao, Bolin, Guo, Yu, Guo, Yequn, Xie, Qiuyou, Chen, Lixiang, Huang, Huiyuan, Yu, Ronghao, Huang, Ruiwang
• Format: Article
• Language: English
• Subjects: Brain - diagnostic imaging; Connectome; Connectome-based predictive modeling (CPM); Consciousness; Consciousness Disorders - diagnostic imaging; Default mode network (DMN); Humans; Magnetic Resonance Imaging; Parietal Lobe; Probabilistic flow estimation (PFE); Propagation structure; Regular; Time delay estimation (TDE)
• ISSN: 2213-1582; 2213-1582
• DOI: 10.1016/j.nicl.2021.102797

•TDE and PFE can characterize the temporal organization of iBA in DOC patients.•DOC patients showed abnormal TDP and PFP mainly in DMN and SMN.•DOC patients had a shortened range of TD value relative to the healthy controls.•The temporal propagation structure of iBA can be used to predict clinical scores for DOC patients. The detection of intrinsic brain activity (iBA) could assist clinical assessment for disorder of consciousness (DOC) patients. Previous studies have revealed the altered iBA in thalamocortical, frontoparietal, and default mode network in DOC patients using functional connectivity (FC) analysis. However, due to the assumption of synchronized iBA in FC, these studied may be inadequate for understanding the effect of severe brain injury on the temporal organization of iBA and the relationship between temporal organization and clinical feature in DOC patients. Recently, the time delay estimation (TDE) and probabilistic flow estimation (PFE) were proposed to analyze temporal organization, which could provide propagation structure and propagation probability at whole-brain level. We applied voxel-wise TDE and PFE to assess propagation structure and propagation probability for the DOC patients and then applied the connectome-based predictive modeling (CPM) to predict clinical scores for patients based on the ROI-wise TDE and PFE. We found that: 1) the DOC patients showed abnormal voxel-wise time delay (TD) and probabilistic flow (PF) in the precentral gyrus, precuneus, middle cingulate cortex, and postcentral gyrus, 2) the range of TD value in the patients was shorter than that in the controls, and 3) the ROI-wise TD had a better predictive performance for clinical scores of the patients compared with that based on ROI-wise PF. Our findings may suggest that the propagation structure of iBA could be used to predict clinical scores in DOC patients.