Fig. 3: Groups are distinguished and characterised by their latent space embedding in an ordered manner. | Communications Biology

Fig. 3: Groups are distinguished and characterised by their latent space embedding in an ordered manner.

From: Whole brain modelling for simulating pharmacological interventions on patients with disorders of consciousness

Fig. 3

A Scatter plot of FC matrices obtained from the three baseline models for the three groups of participants. The groups are clearly separated within the embedded regions. Embedded FC matrices from Disorders of Consciousness (DoC) conditions are distinguished by (B) their distance from the centroid of the FC matrices from the healthy control (CNT) condition, which is the largest for unresponsive wakefulness syndrome (UWS) and smallest for minimally conscious state (MCS). We sampled FC matrices from a grid to calculate different metrics over the latent space: (C) Entropy, (D) mean FC value and (E) correlation to structural connectivity. Each square corresponds to a point from that grid and has therefore an associated value. Highlighted squares indicate the position of the centroids of each group in this space. The values of these metrics in the regions corresponding to each condition are consistent with their interpretation concerning consciousness. F–H Distribution of the previous measures when sampling the embedded points from (A).

Back to article page