Abstract | ||
---|---|---|
Based on the structural connectomes constructed from diffusion tensor imaging (DTI) data, we present a novel framework to discover functional connectomics signatures from resting-state fMRI (R-fMRI) data for the characterization of brain conditions. First, by applying a sliding time window approach, the brain states represented by functional connectomes were automatically divided into temporal quasi-stable segments. These quasi-stable functional connectome segments were then integrated and pooled from populations as input to an effective dictionary learning and sparse coding algorithm, in order to identify common functional connectomes (CFC) and signature patterns, as well as their dynamic transition patterns. The computational framework was validated by benchmark stimulation data, and highly accurate results were obtained. By applying the framework on the datasets of 44 post-traumatic stress disorder (PTSD) patients and 51 healthy controls, it was found that there are 16 CFC patterns reproducible across healthy controls/PTSD patients, and two additional CFCs with altered connectivity patterns exist solely in PTSD subjects. These two signature CFCs can successfully differentiate 85% of PTSD patients, suggesting their potential use as biomarkers. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/ISBI.2013.6556551 | ISBI |
Keywords | Field | DocType |
connectome,biomarker,medical disorders,benchmark stimulation data,image coding,functional connectomic signature pattern identification,biodiffusion,computational framework,learning (artificial intelligence),dynamic transition pattern,sliding time window approach,dti data,post-traumatic stress disorder,dictionary learning algorithm,diffusion tensor imaging,common functional connectome identification,ptsd patient,fmri,temporal quasistable functional connectome segment,functional magnetic resonance imaging,sparse coding algorithm,biomedical mri,cfc identification,resting-state fmri data,dti,brain,brain state,medical image processing,brain condition characterization,dictionaries,learning artificial intelligence,encoding,data models,vectors | Diffusion MRI,Connectomics,Dictionary learning,Pattern recognition,Neural coding,Computer science,Connectome,Image coding,Artificial intelligence,Machine learning | Conference |
Volume | Issue | ISSN |
null | null | 1945-7928 |
ISBN | Citations | PageRank |
978-1-4673-6456-0 | 0 | 0.34 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xiang Li | 1 | 4 | 1.55 |
Dajiang Zhu | 2 | 320 | 36.72 |
Xi Jiang | 3 | 311 | 37.88 |
Changfeng Jin | 4 | 12 | 2.19 |
Lei Guo | 5 | 1661 | 142.63 |
Lingjiang Li | 6 | 12 | 2.53 |
Tianming Liu | 7 | 1033 | 112.95 |