The Noises in Your Head

The link isn't working :(

Worked for me, but someone may have fixed it in the meantime.

Very interesting article, thanks!

So what I'm wondering now (as are other communication engineering types, I'm sure) is what the Bit Error Rate (BER) is for your average neuron/axon? Do the false positives happen with the same frequency as the false negatives, or is there a big difference in the two rates? Does this kill the hypothesis of determinism within the brain? So many questions from my noisy neurons...

-mark.

So, can neuronal noise cause hallucinations? Is it the source of the tunes that get stuck in your head for days at a time? (You know the one's I mean, the catchy melodies that you hear once on the radio then have to hear over and over again on the inner radio until they wear out or something. For my part, I've had TMBG's _Birdhouse in Your Soul_ on auto-repeat in my head for nearly a week now.)

Random comments:

1. bit-error rate. I'd guess that it can't be determined now, and varies greatly from neuron to neuron. As Carl summarized, it is affected by the size of the neuron. Also, a lot of the error talked about in the article was action potential timing, not the either-or of spikes. As the article authors noted, its hard to determine how much noise there is. Some of the statistical variation may be true noise, and some may be significant sources of information that haven't been decoded yet.

2. Yes, hallucinations can be due to noise. Have you ever driven late at night and suddenly thought you saw a bear or something like that, later to discover that it is a road sign? Its happened to me a bunch of times. The "bear" is a real perception, but akin to a hallucination. Our brain, supplied with a very weak light signals tries to make best guesses about the world. These are the Baysian processes described in the article. The guesses are derived from sense data, expectancy, and survival importance. The guesses can be right or wrong. When they are obviously wrong we laugh. But all of perception is better described as inference and guesses than passive responses. Another example. In sleep we dream and the dreams seem like hallucinations. In fact, the content of dreams can be affected by sense data. For example, if a sleep scientist sprays water on the hand of a dreamer, when the dreamer wakes up, it is likely he or she will report that in the dream it started to rain or that, during the dream, the dreamer started to exercise and sweat profusely. Apparently, the "water on the skin" stimulus is getting to the brain, and the dreaming brain is making a best guess about what it is. In this case, the guess tries to incorporate the stimulus into a coherent dream narrative.

3. Carl didn't mention an important distinction between narrow and wide-diameter axons. Axons with wide diameter, particularly myelinated axons, are much faster. In some cases speed is very important. Large axons are most common in long-distance connections in the nervous system. Brain size and neuron size are highly correlated. building big brains is a big engineering problem.

Excellent article, Carl. I thoroughly enjoyed that one - keep them coming!

Excellent article, Carl.

LOVED that article Carl. Very interesting stuff. If anyone knows of websites with more info like this, I'd love a link or two.

Thanks for the extra info, John.