Connectome replacement

Connectome replacement was performed using the initial balanced DMF model (i.e., with optimised G parameter, but without additional inhibitory gain modulation), based on the consensus connectome from diffusion imaging of healthy controls (referred to as the healthy connectome).

Three perturbed connectomes were used. Firstly, the consensus connectome obtained from diffusion imaging of n = 21 DOC patients, referred to as the DOC connectome. Secondly, the original healthy connectome was randomised according to the weight-preserving procedure of¹⁰² to generate a random connectome that differs from the original in terms of topology, but preserves the weight distribution. Thirdly, we used the procedure described in¹⁰² to turn the healthy connectome into a lattice network with the same weight distribution - providing a different and opposite perturbation of the network’s topology.

For each perturbed connectome, the DMF model was used to generate n = 100 simulations, using the optimal global coupling G, but with inter-regional connectivity given by the perturbed connectome rather than the original connectome. This was repeated for each dataset (propofol and DOC) and the resulting simulations were compared with each condition (awake/propofol and control/DOC) in terms of KS-distance.