Monday, March 24, 2008

Oscillations Part VIII: Deep Brain Function or "Noise"

(This is a continuation of notes taken from Ginger Campbell's podcast #31 with György Buzsáki, and a continuation from Part VII):

Independence of "deep" brain function

* the brain maintains its own excitability, activity, in the absence of external information, or external perturbations
* what is it good for?
* physicists and computational neuroscientists like noise, because many of the systems they study wouldn't be able to operate without some energy which is usually supplied with the mechanism of noise
* this noise is generated by the internal activity of the brain at a time when the brain cannot rely on outside sources
* this is the time when brain activity moves forward independently
* sometimes it coincides with sleep - it's very difficult to understand what kind of processes go on in sleep but one of them seems to be consolidation: e.g., remember a conversation with someone - you will:
"recall that information with just one single question and you go through a whole long process of thinking, and that implies that there are cell assemblies that produce self-organized and perpetuating sequences without any further input. And you can think and talk about these events for a long time. And this is all maintained by internally generated activity."

Why is brain "noise" necessary?
"(Deep brain activity) is the internally-generated activity that is independent from the outside world and that is the one that interferes with the incoming inputs. This is the source of noise that is annoying when you are trying to understand the impact of a stimulus on the brain. So this noise, which is considered noise for many, is my most important signal. That's in fact the only source of cognitions - the self-generated or inside-generated activity or internally generated activity of the brain is an absolute requirement for cognition. And there has been very little thought about this, and they are very difficult experiments."

* self-generated, inside-generated activity, or internally generated activity of the brain is an absolute requirement for cognition
* it is very difficult to approach this experimentally, especially in animals
* but if you are seriously interested about how cognitive information is transferred around from one place to the other in the brain, then we have to consider seriously how self-generated activity is serving important functions

* various things must be energy efficient, volume-efficient, wiring-efficient etc., but certain problems must be solved - a balance becomes struck between needs and goals and energy efficiency
* it seems a waste of energy to maintain spontaneous activity - why do we have a lot of neurons firing when we are drowsy?
* this activity/noise makes:
"perception less effective, the inputs are not the same so in fact when you are driving on the highway and getting tired, then your brain takes control and your input is becoming less efficient. So from this point of view it's a disaster. But we have to understand and give some good rationale why is it good in the long run for the brain that it has its own fluctuations."

Some benefits of sleep:

* activity during sleep is not a random useless activity
* in many structures where this has been looked at, it turns out that the activity of sequences and patterns in the neurons that are being active in our sleep processes are pretty much the same as the ones that have been used in our waking experience
* a recent book about creativity discusses what could be the mechanism of creativity in the brain; many creative people reported to the author that they try to create in an environment which is peaceful, calm, and they try to fall asleep, but not quite sleep, and whenever the associations occur in an interesting and random manner they can bring it back to wakefulness.
"I have fantastic associations every single night in my dreams, but I don't have the ability to bring them back to the waking states. These are the states when the activity of the previous day dominate, but at the same time, other things that happened in our previous life will also emerge, and mix with the recently acquired information. Sleep is the best mixing phase because it's noisier so to speak, but perhaps this noisiness is creating, is using something or serving something very important."

Does sleep make you 'smarter'?

* sleep seems to be very important for consolidating memories
* Jan Born, in Lübeck, Germany has done work on this:
* students are given a puzzle to solve
* the puzzles are of sufficient difficulty that it takes more than two or three times to solve them
* they go and sleep on it:
"...when they wake up, a fairly high percentage of students can solve the task right away, compared with those students who have been engaged doing something else, implying that during sleep, those representations about the unsolved problem went through various synapses and various parts of the brain, and in the morning, they were, so to speak, consolidated, and the answer was available."

* many examples show that sleep provides an advantage - there may be a savings of 10, maybe 15% (more information retained)
* interestingly this can be enhanced beautifully by oscillations
* the same lab followed up the idea that there is a slow oscillation in the neo-cortex capable of entraining these fast ripples that seems critical for storing and consolidating episodic memories
* the rationale was, if this is the case, then we should be able to enhance the magnitude of these slow oscillations during sleep
* EEG electrodes were placed on the scalp, electrical currents were applied while the students were in stage 4 sleep
" a result of the electrical sinusoid stimulation, the power of the slow oscillation increased, and perhaps this enhanced activity increased the probability and timing of hippocampal ripples (although this hasn't been shown...). But this entrainment of oscillations enhanced the ability of sleep for an additional 10% or so (more information retained).

* students whose brains were stimulated during stage 4 sleep remembered the memorized items even better than those who had just had a regular sleep

* a counter-argument might be: oscillations not only produce oscillations but they made neurons fire more synchronously, contact with the balancing partners were discharged more effectively
* if perturbations did not occur in an oscillatory manner but in a random manner, perhaps the outcome would be the same

* it's a very difficult issue to answer -
"... it's a bi-directional causation, and the system that oscillates uses every ingredient, and every ingredient when it works perfectly together produces an oscillation. So whether it's an epiphenomenon or not, it doesn't have a trivial solution. However, it does not matter. If you are looking at the practical aspects of this, because they are there, they can be used, they can be useful for a researcher who is interested in how internal activity of the brain or internal generated activity of the brain is generated, and whether oscillations are epiphenomena or not doesn't really matter as long as he understands the processes that underlie behavior and commission."


So ends the interview. I will bring a list of references and further suggested reading here in a separate blogpost, then I will hook all these posts together and post a single link in the menu.

This has been a great few weeks of soaking in this information. I want to thank Ginger Campbell and György Buzsáki for having such a good informative conversation about oscillations.

No comments: