Transparent implant can read neural activity deep in the brain

Transparent implants can read deep neural activity from the surface of the brain. Image source: UC San Diego Jacobs School of Engineering

Researchers at the University of California, San Diego, have developed a neural implant. When placed on the surface of the brain, it can read information about activity deep inside the brain. Testing the technology on genetically modified mice shows researchers are one step closer to creating a minimally invasive brain-computer interface. This research was published in the journal Nature Nanotechnology on the 11th.

Newly developed neural implant overcomes limitations of current technology. It consists of a thin, transparent strip of flexible polymer that conforms to the surface of the brain. Embedded within are high-density arrays of tiny, circular graphene electrodes. Each electrode is 20 microns in diameter and is connected to the circuit board by a micron-thin graphene wire.

In tests on genetically modified mice, the implant allowed researchers to simultaneously capture high-resolution information on two types of neural activity: electrical activity and calcium activity. When placed on the surface of the brain, the implant records electrical signals from the outer layer of neurons.

At the same time, the researchers used a two-photon microscope to fire lasers at the implant to image spikes in neuronal calcium fluorescence signals located 250 micrometers below the brain's surface. They found a correlation between surface electrical signals and deeper calcium fluorescence signal spikes. This correlation allowed the researchers to use surface electrical signals to train neural networks to predict calcium activity at different depths, including the activity of multiple and single neurons.

Previously, when imaging spikes in calcium fluorescence signals, the subject's head had to be held still under the microscope. And these experiments can only last an hour or two at a time. Being able to predict calcium activity from electrical signals now overcomes this limitation, allowing longer experiments to be performed.

The researchers say the technology's success is due to transparency and high electrode density combined with machine learning methods. A new generation of transparent graphene electrodes embedded at high density can sample neural activity at higher spatial resolution, which, combined with machine learning, makes it possible to predict deep neural activity from surface signals.