Combining phase analysis and causal stimulation enables longitudinal mesoscale connectivity mapping

Uncovering the circuit mechanisms of flexible behavior requires characterizing how neural signals propagate across large-scale networks and tracking how those networks evolve over time. Signal propagation through networks can be characterized by injecting activity and analyzing signals propagation, thereby measuring "causal connectivity". However, current techniques cannot measure causal connectivity at centimeter scales over weeklong periods of time, especially in non-human primates. Here, we present a stimulation and recording method to longitudinally measure causal connectivity across centimeter-scale cortical networks in macaques. We combined multi-site optogenetic stimulation with micro-electrocorticography using a semi-chronic implant approach that enables stable daily measurements over weeks-long periods spanning years. Stimulation evoked activity reflected a mixture of local, stimulation-driven responses and secondary, delayed activity presumably driven through connections. We applied phase-based spectral analysis to disentangle stimulation-driven and secondary activity related to causal connections. We validated this approach by tracking connectivity across weeks. We demonstrated improved stability over long durations while preserving sensitivity to eyes-open vs. closed behavioral states. This platform provides a robust means to investigate the large-scale circuit dynamics underlying complex behavior in the primate brain.