A 20-year neural interface must preserve the brain tissue around the electrode
Takashi Kozai: a 20-year neural interface must preserve the living tissue around the electrode
On July 12, University of Pittsburgh bioengineer Takashi Kozai described his vision of a neural interface that could function for ten, twenty, or thirty years. An intracortical neural interface is inserted into the cortex, where it records neuronal activity or delivers electrical impulses to neurons. Twenty years is Kozai's stated goal, not a service life that has already been demonstrated.
When an electrode enters the cortex, it can damage small blood vessels. This damage disrupts the boundary between the blood and brain tissue and triggers an inflammatory response. Microglia and astrocytes then migrate to the site. These cells clear the damaged area and alter its local environment. A prolonged tissue response pushes active neurons away from the contact and makes recordings less stable.
Kozai argues that implant design must account for this entire sequence. The vascular network must be preserved during implantation, and the foreign body response must then be controlled so that neurons remain functional. Oligodendrocytes, the cells that maintain the myelin insulation around nerve fibers, must remain viable near the electrode. Lysosomes must also continue to process cellular waste. Stimulation parameters should be selected to match the natural activity of the neural network.
There is already evidence supporting this model. In a 2017 review coauthored by Kozai, glia were described as active participants in implant function. These supporting cells affect both the state of the neural network and the fate of the electrode.
In April, a soft hydrogel mesh placed on the surface of the animal cortex recorded signals for up to 550 days. The soft material reduces the mechanical mismatch between the electrode and the tissue. Kozai argues that the long-term function of a neural interface also depends on preserving blood vessels, controlling the cellular response, and selecting an appropriate stimulation protocol.
His group tested in animals whether the tissue response could be altered after implantation. Following placement of an intracortical electrode, low-intensity pulsed ultrasound reduced microglial coverage of the contact by 50%, reduced astrocytic scarring by 36%, and improved chronic recording, the authors reported. The experiment was conducted in mice and rats. It shows that the tissue response around an implanted electrode can be modified after implantation.
Stimulation presents a separate problem. A 2026 review of intracortical microstimulation notes that the relationships among pulse parameters, neuronal activation, and the resulting sensation remain only partly understood. There is also evidence from a human participant: a BrainGate participant used an intracortical interface at home for 22.6 months. Kozai's goal spans decades. The electrode must record from and stimulate the brain for years while maintaining contact with living tissue.