Micro dynamics of brain networking for major depression through amplitude modulation–based partial brain functional connectivity analysis

Brain functional networking is complex and dynamical and micro dynamics analysis performs as a good solution to it. In this study, we postulated that micro dynamics of transitioning between states depend on the frequency, enabling the deconstruction of global microstates into a series of partial brain functional connectivities (PBFCs). We propose a novel approach that combines an amplitude modulation (AM)–based algorithm and PBFCs, leveraging a modified similarity measure to assess the differences between two sequences of dynamical state transitions. Moreover, this approach is promising for identifying biomarkers of major depression disorder (MDD). An open data set comprising 128-channel resting-state EEG recordings from both individuals with MDD and healthy controls was used. The results revealed that the characteristic dynamics of the components in the ultra-low-frequency band carrier (0.5–1 Hz) exhibited high accuracy for MDD diagnosis. Moreover, many biomarkers derived from specific PBFCs related to the ultra-low AM of the components of the alpha-band carrier exhibited high sensitivity (area under the receiver operating characteristic curve [AUC] > 0.9). Our novel approach can be used for characterising the intricate brain functional connectivity disparities between individuals with the disorder and their healthy counterparts and thus holds considerable promise for clinical diagnostic applications.