DCM%20for%20evoked%20responses

1
DCM for evoked responses

• Ryszard Auksztulewicz
• SPM for M/EEG course, 2014

2
Does network XYZ explain my data better than
network XY?
?
input
input
3
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?
?
input
input
4
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent
fashion?
?
input
input
5
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent fashion?Is my
effect driven by extrinsic or intrinsic
connections?
?
context
input
6
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent fashion?Is my
effect driven by extrinsic or intrinsic
connections?Which neural populations are
affected by contextual factors?
?
context
input
7
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent fashion?Is my
effect driven by extrinsic or intrinsic
connections?Which neural populations are
affected by contextual factors?Which
connections determine observed frequency
coupling?
?
context
input
8
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent fashion?Is my
effect driven by extrinsic or intrinsic
connections?Which neural populations are
affected by contextual factors?Which
connections determine observed frequency
coupling?How changing a connection/parameter
would influence data?
?
context
input
9
The DCM analysis pathway
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
10
The DCM analysis pathway
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
11
Data for DCM for ERPs / ERFs

• Downsample
• Filter (e.g. 1-40Hz)
• Epoch
• Remove artefacts
• Average
• Per subject
• Grand average
• Plausible sources
• Literature / a priori
• Dipole fitting / 3D source reconstruction

12
The DCM analysis pathway
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
13
The DCM analysis pathway
Hardwired model features
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
14
Models
15
Neuronal (source) model
 
Kiebel et al., 2008
16
NEURAL MASS MODEL
CANONICAL MICROCIRCUIT
Pyr
Inhib Inter
L2/3
xv
Spiny Stell
Spiny Stell
mv
L4
Inhib Inter
Pyr
Pyr
L5/6
spm_fx_cmc
spm_fx_erp
17
Canonical Microcircuit Model (CMC)
Output equation
Supra-granular Layer
Granular Layer
Infra-granular Layer
Pinotsis et al., 2012
18
Canonical Microcircuit Model (CMC)
Bastos et al. (2012) Pinotsis et al.
(2012)
19
Canonical Microcircuit Model (CMC)
Supra- granular Layer
Granular Layer
Infra-granular Layer
20
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
21
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
22
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
23
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
24
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
25
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
26
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
27
Canonical Microcircuit Model (CMC)
Inhibitory Interneurons
Superficial Pyramidal Cells
Supra-granular Layer
Spiny Stellate Cells
Granular Layer
Infra-granular Layer
Deep Pyramidal Cells
Pinotsis et al., 2012
28
Canonical Microcircuit Model (CMC)
Voltage change rate f(current) Current change
rate f(voltage,current)
Pinotsis et al., 2012
29
Canonical Microcircuit Model (CMC)
Voltage change rate f(current) Current change
rate f(voltage,current)
H, t Kernels pre-synaptic inputs -gt
post-synaptic membrane potentials H max PSP
t rate constant S Sigmoid operator PSP -gt
firing rate
David et al., 2006 Pinotsis et al., 2012
30
Canonical Microcircuit Model (CMC)
Supra-granular Layer
Granular Layer
Infra-granular Layer
Pinotsis et al., 2012
31
The DCM analysis pathway
Hardwired model features
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
32
5
4
3
2
1
33
5
4
3
2
1
Input
34
5
4
3
2
1
Input
35
5
4
3
2
1
Input
36
5
4
3
2
1
Input
37
Factor 1
5
4
3
2
1
Input
38
Factor 1
Factor 2
5
4
3
2
1
Input
39
The DCM analysis pathway
Fixed parameters
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
40
Fitting DCMs to data
41
Fitting DCMs to data
H. Brown
42
Fitting DCMs to data
H. Brown
43
Fitting DCMs to data

• Check your data

H. Brown
44
Fitting DCMs to data

• Check your data
• Check your sources

H. Brown
45
Fitting DCMs to data

• Check your data
• Check your sources
• Check your model

IPL
IPL
V4
V4
Model 2
H. Brown
46
Fitting DCMs to data

• Check your data
• Check your sources
• Check your model
• Re-run model fitting

H. Brown
47
The DCM analysis pathway
Fixed parameters
Build model(s)
Fit your model parameters to the data
Pick the best model
Make an inference (conclusion)
Collect data
48
Does network XYZ explain my data better than
network XY?Which XYZ connectivity structure
best explains my data?Are X Y linked in a
bottom-up, top-down or recurrent fashion?Is my
effect driven by extrinsic or intrinsic
connections?Which connections/populations are
affected by contextual factors?
?
context
input
49
Example 1 Architecture of MMN
Garrido et al., 2008
50
Example 2 Role of feedback connections
Garrido et al., 2007
51
Example 3 Group differences
Boly et al., 2011
52
Example 4 Parametric effects
Auksztulewicz Blankenburg, 2013
53
Auksztulewicz Blankenburg, 2013
54
Auksztulewicz Blankenburg, 2013
55
Example 5 Specific CMC populations
Moran et al., 2013
56
Motivate your assumptions!
Rubbish data
Perfect model
Rubbish results
Perfect data
Rubbish model
Rubbish results
57
Thank you!

• Karl Friston
• Gareth Barnes
• Andre Bastos
• Harriet Brown
• Jean Daunizeau
• Marta Garrido
• Stefan Kiebel
• Vladimir Litvak
• Rosalyn Moran
• Will Penny
• Dimitris Pinotsis
• Bernadette van Wijk