Merging electronics and the brain - towards a fully biocompatible neural interface

Sammanfattning: Techniques for interfacing the nervous system using chronically implanted electrodes are turning into invaluable tools for neurophysiological research and treatment of neurological disorders. These methods have been developed and refined over the last decades. However, they still suffer from deteriorating electrical capabilities and the fact that they inevitably damage the tissue into which they are implanted, thus interrupting the very circuits aimed to be studied or stimulated. This problem could be alleviated by constructing truly biocompatible interfaces, which causes minimal damage during implantation, and minimal chronic tissue reactions. This thesis is aimed at identifying which factors that gives rise to these reactions, and finding ways to construct implants that as far as possible integrate these factors. As it was hypothesized that electrodes should be ultraflexible, the problem of how to implant such structures with positional control into soft tissue, such as the brain, had to be solved. To this end, we developed a novel gelatine based vehicle, into which the electrodes were embedded. This vehicle permits any structure, even nanoscaled, to be implanted. Furthermore, the gelatine coating was shown to have highly biocompatible. We also show that the number of electrodes placed in a rat brain does not greatly affect the resulting tissue reactions to each implant. This means that a large number of brain areas can be interfaced simultaneously without risking a snow-balling tissue reaction. We have explored the effect of the density discrepancies between electrodes and the neural tissue on the ensuing tissue reactions. We show that when an electrode with a density close to that of the tissue is constructed the chronic glial scar is almost completely eliminated. This surprisingly strong effect shows that even the minute inertial forces arising from movements of the animal are enough to stimulate a substantial glial scarring. The fourth paper is a qualitative assessment of the cells immediately interfacing the implants. We find a substantial cell population between the astrocytic scar and the implants, which for the most part is constituted by stromal cells. These cells may provide the missing link between the degree of tissue reactions and the deterioration of electrode recording capabilities. In total, this thesis has provided fundamental knowledge about how to design a fully biocompatible neural implant. The discovery that inertial forces are a major factor in the glial scar formation suggests that it is possible to develop neural implants that cause minute tissue reactions, which can enable valid long-term neurophysiological recordings.