Thursday 5 July 2012

Erik Cook: Are Neural Fluctuations in Cortex Causally Linked to Visual Perception?


      Abstract: Our perception fluctuates when we view images at the threshold of our visual capabilities.  It has been widely shown that fluctuations in visual cortical activity are correlated with fluctuations in perception.  The source of these neural fluctuations, however, is not clear.  Are they causal, such as bottom-up sensory noise, that directly influences perception? Or are these fluctuations non-causal, such as top-down attentional modulation, that produce correlations between sensory neural activity and perception when no functional link actually exists between the two?  In this presentation, I will present accumulating evidence that both causal and non-causal processes are responsible for this functional link and that careful electrophysiological observations can distinguish between these two sources of neural fluctuations.

      Smith, J.E., Zhan, C.A. and Cook, E.P.  The functional link between area MT neural fluctuations and detection of a brief motion stimulus.  Journal of Neuroscience 31: 13458 - 68, 2011. http://www.jneurosci.org/content/31/38/13458.long

Comments invited

16 comments:

  1. I was just wondering to what extent we can really extend the findings obtained in monkeys and rats to human beings. We talked earlier that debate if it's about consciousness being there or not, or if the interest is rather to understand consciousness as a continuum. A. Cook's presentation is interesting, but I'm just thinking about how would the findings presented be applicable to human beings as well. When we see a bunch of dots moving to our right, how many things else are getting activated at the same time in our head and that influence the "consciousness", the "perception", of it all? I'm just having a moment of worry about the ecological validity of all these studies on consciousness. Of course I understand that we have to isolate variables, this is science after all.

    ReplyDelete
    Replies
    1. Shady - your misgivings are well founded and justified. All this would be better studied in humans if we had brain imaging methods with much better spatial and temporal resolution. Then we could go right to the source. I expect these brain imaging technologies will come along eventually, but it may be a while. In the meantime, animal-based studies provide us a glimpse of how consciousness in humans may be approached and understood.

      Delete
  2. I agree, and I think that was the concern of many people during the question session. We saw so many animal studies during this summer school, but how many can really be interpreted according to the human consciousness ? Perhaps we should first look at meta analysis and litterature review. Obviously there are some advantages to animal study, as the single neurons recording, but where that lead us ? I believe that consciousness is still such an abstract subject, we feel the need to study several research paradigms in different fields of study (philosophy, anthropology, neuropsychology, neurology, etc.) and animal experiments is one of it.
    Claudia Polevoy - UQAM

    ReplyDelete
    Replies
    1. Claudia - So you and others are asking “why study the neural mechanisms of consciousness in animals?” We all agree it would be better to study this in humans if we could only record human brain activity with sufficient resolution. For example, the temporal response of fMRI BOLD signals is 10X too slow compared to what we think are the time-scale of neural signals associated with consciousness. There is still a lot of work to do while we are waiting for advanced human brain imaging technologies to come along - which includes defining the variables that we believe are associated with consciousness.

      Delete
  3. Claudia and Shady, you are making an assumption that human consciousness is dramatically different from consciousness in other animals. It might be, but the differences could also be relatively small (as a metaphor, there are differences in the visual sensory systems of mammals but there are still large similarities). We can hash out the details of the differences once we have generated some data-based theories.

    ReplyDelete
    Replies
    1. I agree. I guess for now it's just finding what is consciousness : how, what and why. I guess the answer to these questions should be the same for all organisms. Yes of course, our consciousness might be different due to our language for example, but Mr. Harnad keep telling us that it's not a matter of degree: you are conscious or you're not. Then I guess for now animals can give us a lot of informations about these questions. When we will know more it should be interesting to see the differences between animals' and humans' consciousness.

      Delete
    2. Martha & Marjorie - I'm with you in that consciousness is probably a matter of degrees until we can prove otherwise.

      Delete
  4. I think Dr. Cook's approach of looking at both single cell activity as well as population activity (EEG or LFP or at least multiunit) is important, not just because we get more info that way but also the population activity has a bearing on the single cell activity. In the case of oscillatory activity measurable using LFP, it can modulate the cell's firing timing as we saw with phase locking which can have its own meaning, as seen in Dr. Singer's talk.

    ReplyDelete
  5. Carey - I agree that the more independent measurements the better. However, the key word is independent. For example, if gamma band LFPs are correlated with neural spiking, then it does not add much to our understanding.

    ReplyDelete
    Replies
    1. Dr. Cook, I completely agree. I have always been skeptical of gamma band LFPs because they could potentially just be poorly isolated spikes! And indeed they appear to co-occur very often.

      And what I still do not get is, how does gamma oscillation synchronize across brain regions, even distant ones? If you see synchrony pop up in EEG/MEG, how synchronized is it really in terms of exact timings? I consider gamma rhythms as something that arises from more local networks, because higher-frequency oscillations/signals do not propagate over large distances well. I wonder if gamma synchrony is just a by-product of the faster-oscillations riding on slower-oscillations that are truly synchronized across distant regions (due to common pacemaker)? Any help to clarify this matter will be greatly appreciated!

      Delete
    2. The how and why of synchrony in the gamma frequency band is an ongoing point of study in systems neuroscience. At this point, I don't think my opinion would contribute to much on this subject. What we need are good experiments that measure the behavioral effects of modulating gamma synchrony with no modulation to other measurements of neural activity!

      Delete
  6. Thanks to Dr Cook knowledge of statistics for reminding us that if we want to study the neural correlates of consciousness, we need variability. What is it then that you vary in order to correlate neural consciousness with another variable?
    That’s an important question.
    First, we would need to agree on if consciousness can be mapped on a continuum. Although it seems to be the case for neurological scales of consciousness (awareness … confusion … coma…), it doubt there is such a consensus on the case of the phenomenological consciousness (feeling) we are referring to in this conference.

    ReplyDelete
    Replies
    1. I am not sure you need to say that consciousness is a matter of degree. What you can say, rather, is that it is a matter of degree whether or not a certain stimulus is perceived (or maybe ‘felt’) as moving in a certain direction. Take, for example, the experiment with the two random-dot patches designed by Dr. Cook. It might be that the monkeys’ feeling that there is motion in one direction or another is a matter of degree. The perceptual states themselves are fully conscious, but it is the content of these states that can vary on a continuum.

      Delete
    2. I agree with Alexandre's comment. I recall that after the talk, I thought for a second that the Gamma-Beta temporal interactions described by Dr. Cook might be compatible with the all-or-none consciousness approach (where one is either conscious or not, and, as Alexandre said, it is only the content of consciousness that varies on a continuum).

      Delete
  7. Xavier Dery ‏@XavierDery

    Cook admits he has bouts of "Robot Envy" and it must take some courage. Robots ARE cool but please neuroscientists, don't give up! #TuringC

    9:57 AM - 5 Jul 12 via Twitter for Android

    ReplyDelete
  8. Xavier Dery ‏@XavierDery

    Cook: if only everyone trying to find the link between neronal activity and behavior would be as prudent as you about causality! #TuringC

    10:37 AM - 5 Jul 12 via Twicca Twitter app

    ReplyDelete