Discussion

The Neurotrophic Electrode is based on regeneration. It is composed of biocompatible glass, gold wires, Teflon and methyl-methacrylate glue. It is basically a hollow glass cone with wires inside, strain relief, and growth factors. Of course it damages the cortex, otherwise regeneration would not occur. Regeneration is encouraged by the growth factors (rat NGF in rats and monkeys). So, the neuropil grows inside the electrode and reconstitutes the brain inside the electrode without neurons but with copious amounts of myelinated axons. Survival? 16 months in rats when they died, 14 months in monkeys until they damaged the implant, four years in patient JR until death, four years in patient TT until death, 13 years in ER until he died, and 7 months in me until the electronics had to be removed due to separation of the incision edges (the electrodes glass tips are still in my speech motor cortex, and hey, I can speak).

The 13 years in ER is obviously a record with recordings up to a decade. The references are in my attached CV. ER’s histological analysis demonstrated no scarring, no gliosis, and abundant myelinated axons (never any neurons), just like all the other histological analyses in rats and monkeys.

Furthermore, recordings persisted in all these participants without changing the parameters of single units (unless we had to change out the single channel amplifiers). This is true of ER at years 9 and 10 when we were able to condition the single units, which is more evidence these signals were not just noise. Recording after year 10 was not possible because his brainstem stroke extended and he almost passed out when we lifted his head (the stroke damage involved the brainstem control of is blood pressure). So, we stopped recording at year 10.

I often wonder why more people do not use this electrode. I believe it is mainly because the funding is following the fashion of trying to improve the otherwise excellent Utah array and other electrodes that hold the signals for weeks, months and years, but not long term. They have contributed immensely to understanding brain function. The Neurotrophic electrode looks horrible and everybody seems to believe that we must not damage the brain. Funny what happens when you take the opposite approach: The Neurotrophic Electrode damages the brain and reconstitutes the neuropil inside the electrode tip. And it works! For evidence, please check out the references in my CV.

I want to discuss the data from each electrode and how come I get so many (15 to 20). I need to show you how I do it. I takes several days per electrode to sort out the single units and show that they are verifiably single. Briefly, I use the convex hull technique that is unique to the Neuralynx Cheetah system. First, the continuous stream of data is separated using two thresholds, one above and one below the data stream. Then the system lets me sort by amplitude (peak to trough), amplitude above the baseline, the width, area under curve and other parameters. The convex hull technique allows me to delete units so I don’t have to worry about including other units, that is, any outliers are deleted. I need to show you. Then each unit is auto-correlated to detect just one peak. If two or more peaks are found I start over. Most people are afraid to tackle the small units because they dismiss them as noise. We know exactly where the system noise floor is (11 uV) so anything above that is ‘ripe for the picking’! Much of the data above that is physiological noise, I agree, which is why we reject 80 to 90% of these waveshapes.