Fig. 5.

Spatial specificity of the positive relationship between myelin estimates and dipole moments. Variational Bayes GLMs with myelin estimates based on A the MT, B the R1 or C the R2* maps were used as independent variables to explain dipole moments across ROIs with different spatial extent. Differences in log Bayes factors for ROIs with increasing spatial extent, TE1.2, auditory koniocortex TE1 (i.e. TE1.0, TE1.1 and TE1.2 combined) and myelin estimates averaged across the whole hemisphere (global), are displayed. Since the log model evidence is a relative metric, all plots are normalised by the worst model of the corresponding quantitative MPM (set to zero). Log Bayes factor differences larger than 3 indicate that the data support the model with the larger Bayes factor.