1. DCM for evoked responses Ryszard Auksztulewicz SPM for M/EEG course, 2015

2. ? 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? input context

3. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway

4. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway

5. Data for DCM for ERPs / ERFs 1.Downsample 2.Filter (e.g. 1-40Hz) 3.Epoch 4.Remove artefacts 5.Average Per subject Grand average 6.Plausible sources Literature / a priori Dipole fitting / 3D source reconstruction

6. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway

7. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway ‘Hardwired’ model features

8. Models

9. Kiebel et al., 2008 Neuronal (source) model

10. spm_fx_erp Inhib Inter Spiny Stell Pyr L2/3 L4 L5/6 NEURAL MASS MODEL

11. Canonical Microcircuit Model (‘CMC’) Bastos et al. (2012) Pinotsis et al. (2012)

12. vv vv spm_fx_cmcspm_fx_erp Inhib Inter Spiny Stell Pyr L2/3 L4 L5/6 NEURAL MASS MODELCANONICAL MICROCIRCUIT Pyr Spiny Stell Inhib Inter Pyr

13. Canonical Microcircuit Model (‘CMC’) Supra- granular Layer Infra- granular Layer Granular Layer

14. Superficial Pyramidal Cells Superficial Pyramidal Cells Granular Layer Supra- granular Layer Infra- granular Layer Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012 Canonical Microcircuit Model (‘CMC’)

15. Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012 Canonical Microcircuit Model (‘CMC’)

16. Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012 Canonical Microcircuit Model (‘CMC’)

17. Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012

18. Canonical Microcircuit Model (‘CMC’) Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012

19. Canonical Microcircuit Model (‘CMC’) Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012

20. Canonical Microcircuit Model (‘CMC’) Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012

21. Canonical Microcircuit Model (‘CMC’) Granular Layer Supra- granular Layer Infra- granular Layer Superficial Pyramidal Cells Superficial Pyramidal Cells Spiny Stellate Cells Deep Pyramidal Cells Deep Pyramidal Cells Inhibitory Interneurons Pinotsis et al., 2012

22. Voltage change rate: f(current) Current change rate: f(voltage,current) Pinotsis et al., 2012 Canonical Microcircuit Model (‘CMC’)

23. David et al., 2006; Pinotsis et al., 2012 Voltage change rate: f(current) Current change rate: f(voltage,current) H, τ Kernels: pre-synaptic inputs -> post-synaptic membrane potentials [ H: max PSP; τ: rate constant ] SSigmoid operator: PSP -> firing rate Canonical Microcircuit Model (‘CMC’)

24. Granular Layer Supra- granular Layer Infra- granular Layer Pinotsis et al., 2012

25. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway ‘Hardwired’ model features

26. 2 1 4 3 5

27. 2 1 4 3 5 Input

28. 2 1 4 3 5

29. 2 1 4 3 5

30. 2 1 4 3 5

31. 2 1 4 3 5 Factor 1

32. 2 1 4 3 5 Input Factor 1Factor 2

33. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway Fixed parameters

34. Fitting DCMs to data

35. H. Brown

36. Fitting DCMs to data H. Brown

37. Fitting DCMs to data 1.Check your data H. Brown

38. Fitting DCMs to data 1.Check your data 2.Check your sources H. Brown

39. 1.Check your data 2.Check your sources 3.Check your model Model 1 V4 IPL A19 OFC V4 IPL A19 OFC V4 IPL Model 2 V4 IPL H. Brown Fitting DCMs to data

40. 1.Check your data 2.Check your sources 3.Check your model 4.Re-run model fitting H. Brown

41. Collect data Build model(s) Fit your model parameters to the data Pick the best model Make an inference (conclusion) The DCM analysis pathway Fixed parameters

42. ? 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? input context

43. Garrido et al., 2008 Example #1: Architecture of MMN

44. Garrido et al., 2007 Example #2: Role of feedback connections

45. Boly et al., 2011 Example #3: Group differences

46. Auksztulewicz & Friston, 2015 Example #4: Factorial design & CMC Bastos et al., Neuron 2012 Attention cf. Feldman & Friston, 2010 L2/3 L4 L5/6 s xx  xx FORWARD PREDICTION ERROR BACKWARD PREDICTIONS A1 ST G   

47. 2x2 design: Attended vs unattended Standard vs deviant (Only trials with 2 tones) N=20 Auksztulewicz & Friston, 2015

48. Flexible factorial design Thresholded at p<.005 peak-level Corrected at a cluster-level pFWE<.05 Contrast estimate A1E1 A1E0 A0E1 A0E0 Attention Expectation Auksztulewicz & Friston, 2015

49. Flexible factorial design Thresholded at p<.005 peak-level Corrected at a cluster-level pFWE<.05 Contrast estimate A1E1 A1E0 A0E1 A0E0 Auksztulewicz & Friston, 2015

50. input Inh Int input SP input Connectivity structure Extrinsic modulation Intrinsic modulation

51. Superficial pyramidal cells: signal prediction errors to higher areas Brown & Friston, 2010 Feldman & Friston, 2013 Inhibitory interneurons: prediction errors of hidden states Bastos et al., 2012 attentional effects of ACh -> depression of I.I. activity Xiang et al., 1998 Buia and Tiesinga, 2006 attentional gamma spike-LFP phase locking Vinck et al., 2013 vv vv xx xx SPM CMC

53. Auksztulewicz & Friston, 2015

54. Rubbish data Perfect model Rubbish results Perfect data Rubbish model Rubbish results Motivate your assumptions!

55. 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