Stimulation – control à 30 % differences
n Spontaneous correlations = functional connectivity ?
ICA (independent component analysis)
Review : Calhoun NeuroImage 2008/ Cole Frontiers Sys Neur 2010
ICA : Comon Signal Processing 1994/ Bell Neural Computation 1995
ICA for fMRI: McKeown Human Brain Mapping 1998
ICA for resting fMRI: Kiviniemi NeuroImage 2003
ICA for FMRI software: MELODIC in FSL/ GIFT / Brain voyager
Pattern:
Beckmann phil trans toy soc b 2005 / Smith PNAS 2009
Which area is activate with the task
Temporal characteristics:
n Dynamic Causal Modelling friston Neuroimage 2003
1. fit to FMRI data using Bayesian methods
2. not yet proven for high-dimensional resting FMRI
n Lag-based, e.g. Granger causality Gobel MRI 2003
1. If timeseries B lags behind A then A causes B
2. Haemodynamic variability a serious problem
Paper: Smith Neuroimage 2010 / rubinov neuroimage 2010 (graph theory)
Craddock MRM 2009 (applications of network matrix)
HCP – pushing spatial & temporal resolusion Feinberge PLoS 2010
-----------------------------------------------------------------------------------------------------
n Functional Connectivity
Region 1 / seed à correlated ?
Region 2 / seed à correlated ?
n Potential roles of resting-state FMRI in neuropsychiatric disorders
n Disease mechanism
n Diagnostics
n Prognostics
n Treatment efficacy ( seizure/ epilepsy)
è Paper : Soddu et al., Hum Brain Mapp 2011
n Alzheimer biomarker
Greicius et al PNAS 2004
n FTD pateint vs AD
Zhou et al brain. 2010