Dynamic Causal Modelling for EEG and MEG

Transcription

1 Dynamic Causal Modelling for EEG and MEG Stefan Kiebel Ma Planck Institute for Human Cognitive and Brain Sciences Leipzig, Germany

2 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

3 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

4 Mismatch negativity (MMN) Paradigm standards deviants pseudo-random auditory sequence 80% standard tones 500 Hz 20% deviant tones 550 Hz time Raw data (e.g., 128 sensors) Preprocessing (SPM8) Evoked responses (here: single sensor) μv time (ms) Garrido et al., NeuroImage, 2007

5 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

6 Electroencephalography (EEG) amplitude (μv) time (ms) Modelling aim: Eplain all data with few parameters How to: Assume data are caused by few interacting brain sources

7 Probabilistic inference Forward problem p( y, m) Likelihood Posterior distribution p( y, m) Inverse problem

8 Connectivity models Conventional analysis: Which regions are involved in task? DCM analysis: How do regions communicate? STG STG STG STG A1 A1 A1 A1 Input (stimulus) Input (stimulus)

9 Model for auditory evoked response Garrido et al., PNAS, 2007

10 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

11 Inference at meso-scale macro-scale meso-scale micro-scale eternal granular layer eternal pyramidal layer internal granular layer internal pyramidal layer AP generation zone synapses

12 Neural mass equations and connectivity Etrinsic forward connections spiny stellate cells inhibitory interneurons pyramidal cells ) ) ( ) (( e e L F e e Cu S I A A H 1 2 ( 0 ) A F S ( 0 ) A L S ( 0 ) A B S Etrinsic backward connections Intrinsic connections Etrinsic lateral connections,u, f )) ( ) (( e e L B e e S I A A H ) ( 2 )) ( ) ( ) (( i i i i e e L B e e S H S S A A H Jansen et Rit, Biol Cybern,1995 David et al., NeuroImage, 2006

13 Source activity over time Source dynamics f Forward Backward Lateral Input u f (, u, ) states parameters θ

14 Spatial forward model Kiebel et al., NeuroImage, 2006 Daunizeau et al., NeuroImage, 2009

15 The generative model Source dynamics f Spatial forward model g f (, u, ) states parameters θ y g(, ) Evoked responses Input u David et al., NeuroImage, 2006 Kiebel et al., Human Brain Mapping, 2009

16 Probabilistic inference Forward problem p( y, m) Likelihood Posterior distribution p( y, m) Inverse problem

17 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

18 Bayesian inference Evoked responses Specify generative forward model (with prior distributions of parameters) Variational (Bayesian) algorithm Iterative procedure: 1. Compute model response using current set of parameters 2. Compare model response with data 3. Improve parameters, if possible 1. Posterior distributions of parameters p( y, m) 2. Model evidence p( y m)

19 best? Model selection: Which model is the best? data y Model 1 Model 2 p( y m1 ) p y, m ) ( 1 p( y m2 ) p y, m ( 2 ) Model selection: Select model with highest model evidence... p( y mi ) best? Model n p( y mn ) p y, m ( n ) Stephan et al., NeuroImage, 2009 Penny et al., PLoS Comp Biol, 2010

20 Overview 1 M/EEG analysis 2 Dynamic Causal Modelling Motivation 3 Dynamic Causal Modelling Generative model 4 Bayesian inference 5 Applications

21 Auditory evoked potential Garrido et al., PNAS, 2007

22 Auditory evoked potential time (ms) time (ms) Garrido et al., PNAS, 2007

23 Mismatch negativity: EEG IFG IFG IFG Forward - F Backward - B Forward and Backward - FB STG STG STG STG STG STG STG A1 A1 A1 A1 A1 A1 input input input Forward Backward Lateral Forward Backward Lateral Forward Backward Lateral modulation of effective connectivity Garrido et al., NeuroImage, 2007

24 log-evidence MMN: Group model comparison Bayesian Model Comparison Group level Forward (F) subjects Backward (B) Forward and Backward (FB) Garrido et al., NeuroImage, 2007

25 Mismatch negativity: MEG Standard D1 D2 D3 CVC Bart Bart Burt beat Tone Matched to formants of vowel Schofield et al., PNAS, 2009

26 Mismatch negativity: MEG Schofield et al., PNAS, 2009

27 Summary DCM enables testing of hypotheses about how brain sources communicate. Differences between conditions or groups are modelled as modulation of connectivity. Bayesian inference is used to take into account the variability over models and parameters.

28 Thanks to: Marta Garrido Jean Daunizeau Karl Friston Jeremie Mattout Christophe Phillips Thank you!