Neural interfaces that detect or influence nerve cell signals help treat some nervous system conditions, yet many current versions have limits. Most can only send signals to the brain and cannot accurately record brain activity at the same time. Even those that record signals often miss very weak activity at extremely low frequencies.
Researchers have developed a new graphene-based neural interface that overcomes these limits. The device records brain signals, including ultra-low frequency ones, decodes the information, and then sends electrical pulses back to adjust brain activity in real time. This two-way ability may support more precise treatments for neurological disorders that adapt to each patient’s changing brain patterns.
Combining graphene sensors and electrodes in one device
The interface brings together two graphene technologies. One part uses monolayer graphene transistors that detect brain electrical changes with high sensitivity to very low frequencies. The other part uses microelectrodes made from nanoporous reduced graphene oxide, a graphene material with tiny pores, to deliver controlled electrical pulses that influence nerve cell activity. Earlier attempts to combine similar functions often produced signal interference - the stimulation pulses distorted the recorded brain signals. The new design avoids this problem. Tests showed that recording continued clearly even while stimulation occurred.
The device was built in clean-room facilities and tested in living mice. In these experiments it detected real-time brain signals and responded with adjusted stimulation. This shows it can identify important brain markers and react with precise, changing pulses rather than fixed settings.
The work builds on earlier studies by the same researchers, including a 2018 demonstration of graphene implants that recorded very low-frequency brain activity and later work on nanoporous graphene. Their long collaboration has also led to the creation of a spin-off company that licenses the technology. The company has already completed an initial human trial to check safety and performance of graphene neural interfaces for future clinical use. This advance brings closer the possibility of adaptive, patient-specific therapies that both listen to and speak to the brain.
This research is published in Nature Communications.