Carbon-based neural interfaces to probe retinal and cortical circuits with functional ultrasound imaging in vivo

Graphene could address some challenges which current neural interfaces are facing. For instance, graphene-based field effect transistors (gFETs) can record brain signals at infraslow frequencies while reduced graphene oxide (rGO) microelectrodes display high charge injection enabling focal electrical stimulation. Here, we assessed the potential of different graphene-based electrodes for recording and stimulation in rodents, with functional ultrasound imaging (fUS) to quantify cerebral neural activity as read-out on visual cortex activity. In our first project we investigated how fast and slow brain signals couple to blood flow under pathophysiological conditions. Using a model of pharmaco-induced epileptic rat, we recorded seizures simultaneously with fUS and gFETs placed on the cortex, thereby validating their functional compatibility. This innovative concurrent platform revealed oscillations in cerebral blood volume (CBV) and infraslow signal before the seizure onset and a tight positive correlation during seizures, suggesting a common shared mechanism in the generation of CBV oscillations and infraslow oscillations. In a second project, we checked the biocompatibility and functionality of rGO subretinal, electrodes for vision restoration, implanted in the subretinal space of rats. Post-mortem immunolabelling of microglia revealed a non-significant inflammation of the implanted retina compared to control eyes. Cortical CBV changes were measured with fUS in response to monopolar electrical stimulation of 100-µm and 25-µm-diameter rGO electrodes. Significant fitting of CBV with stimulation pattern in the superior colliculus suggest successful information transmission to the visual areas. These results provide great confidence in graphene technologies for the elaboration of future prosthetics devices for neural recording or stimulation.