Figure 2

Experimental data. A Normalized intensity (i) for the first 128 time points for the first three channels of the EEG data. (B) Normalized intensity (i) for the first 128 time points for the first three channels of the fNIRS data. (C) Normalized intensity (i) for the first 128 time points for the first three channels of the ECoG data. (D) Estimated timeseries for the EEG data using the original model. (E) Estimated timeseries for the fNIRS data using the original model. (F) Estimated timeseries for the ECoG data using the original model. (G) Estimated timeseries for the EEG data using the modified model. (H) Estimated timeseries for the fNIRS data using the modified model. (I) Estimated timeseries for the ECoG data using the modified model. (J) Posterior estimates of the intrinsic connectivity in the EEG data using the original model. (K) Posterior estimates of the intrinsic connectivity in the fNIRS data using the original model. (L) Posterior estimates of the intrinsic connectivity in the ECoG data using the original model. (M) Posterior estimates of the intrinsic connectivity in the EEG data using the modified model. (N) Posterior estimates of the intrinsic connectivity in the fNIRS data using the modified model. (O) Posterior estimates of the intrinsic connectivity in the ECoG data using the modified model. (P) Approximate lower bound log model evidence given by the free energy (F) following Bayesian model inversion for the original (o.) and modified (m.) models using the EEG data Probabilities (p) derived from the log evidence are shown in the inset left. (Q) Approximate lower bound log model evidence given by the free energy (F) following Bayesian model inversion for the original (o.) and modified (m.) models using the fNIRS data Probabilities (p) derived from the log evidence are shown in the inset left. (R) Approximate lower bound log model evidence given by the free energy (F) following Bayesian model inversion for the original (o.) and modified (m.) models using the ECoG data Probabilities (p) derived from the log evidence are shown in the inset left. (S) The intensity (int.) at every point in time for the non-dissipative form of the modified state equation, i.e., excluding external driving inputs and noise, using the posteriors from the modified EEG model in (G) and the Hamiltonian as the observer equation. (T) The intensity (inten.) at every point in time for the non-dissipative form of the modified state equation, i.e., excluding external driving inputs and noise, using the posteriors from the modified fNIRS model in (H) and the Hamiltonian (Hamil.) as the observer equation. (U) The intensity (inten.) at every point in time for the non-dissipative form of the modified state equation, i.e., excluding external driving inputs and noise, using the posteriors from the modified ECoG model in (I) and the Hamiltonian (Hamil.) as the observer equation.