Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Medical Can Brain Scans See Depression? [NY Times]

Tags:
  1. Oct 19, 2005 #1

    hypnagogue

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    http://www.nytimes.com/2005/10/18/health/psychology/18imag.html?incamp=article_popular
     
  2. jcsd
  3. Oct 19, 2005 #2

    somasimple

    User Avatar
    Gold Member

    Hi,

    It is well known that depressed states are corelated with decreased brain activity (see A Damasio, O Sacks). But it is ure that defining a "normal" activity will be a huge problem.
     
  4. Oct 19, 2005 #3

    DocToxyn

    User Avatar
    Science Advisor

    The application of functional brain scans like fMRI, PET, SPECT and others should allow the activity in specific regions of the brain of a depressed person to be compared to someone without clinically diagnosed depression. Keep in mind that it may not simply be an overall decrease in brain activity, perhaps some areas are more active in the depressed brain, it depends on the underlying pathology/condition.
    Testing the hypothesis that locus coeruleus hyperactivity produces depression-related changes via galanin.
    The use of single photon emission computed tomography in depressive disorders.
    Neuroplasticity: from MRI to depressive symptoms.
    Is major depression a neurologic disorder with psychiatric symptoms?
     
  5. Oct 19, 2005 #4

    somasimple

    User Avatar
    Gold Member

    Yes, that's true and many scientists say that brain functioning becomes draft with depression. The slow functioning is under the serotonin production.
     
  6. Oct 19, 2005 #5

    hypnagogue

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    According to the article:
    So a big problem seems to be how to decide whether observed variation is due to some (say) depressive disorder, or whether it falls under the expected range of variation in 'normal' subjects. Of course we can always observe the differences between populations once we've already sorted out who is clinically depressed and who isn't, but using imaging techniques as a diagnostic tool to decide whether a given subject should or should not be categorized as depressed seems to be another issue altogether. Also, statistically speaking, there might be some difference on average between the depressed and normal brain types that is relatively clear, but such a difference would be much more difficult to discern on a case-by-case basis.

    I wonder how much of these difficulties can be chalked up to idiosyncratic/developmental differences in individuals in terms of how various things in the brain are encoded? This might be a nut complex enough that getting imaging tools with better spatial/temporal resolutions won't be enough to crack it. Then again, it might not.
     
    Last edited: Oct 19, 2005
  7. Oct 19, 2005 #6

    somasimple

    User Avatar
    Gold Member

    That is what I have said.
     
  8. Oct 20, 2005 #7
    Seems to me scanning the brain to find out what is going on psycologically is like xraying a computer to find out what progam is running. Bah, humbug.
     
  9. Oct 20, 2005 #8

    somasimple

    User Avatar
    Gold Member

    But mind is a brain affair! :rolleyes:
     
  10. Oct 20, 2005 #9

    hypnagogue

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    kublai wasn't debating that, as should have been clear from the hardware/software analogy.

    I think your concern is valid, though it's not as bad as you might think. X-raying a computer isn't likely to tell you anything about what the computer is doing; a better analogy would be trying to figure out what software a computer is running by analyzing the electrical signals it send back and forth across its processing/memory components.

    Also, keep in mind that imaging studies of the brain do not stop at merely imaging the brain. Rather, the brain is imaged while subjects perform some kind of cognitive tasks, so we do have a pretty good high-level idea of what kinds of input, processing of input, and output are involved. For instance, if we see that a particular set of brain regions is consistently more active than baseline while subjects perform a visual memory task, we have good reason to believe that these regions are involved with visual memory. Of course it's all much more complex than that, and there's lots of further refinements and dissociations and so on that need to be done, but the point is that we're not completely in the dark about what the brain is doing at the outset.

    So to further refine your analogy, we should say that brain imaging studies are like observing the pattern of electrical activity in a computer while we feed certain inputs into it, ask the computer to perform certain computations on the input, and observe the output. In this way we can slowly begin to refine our knowledge of what exactly the computer is doing during this input/computation/output process, and how it's doing it (i.e., we slowly figure out the nature of the software that the computer is running by observing how its internal workings are related to its external 'behavior').
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Can Brain Scans See Depression? [NY Times]
Loading...