In a previous article, cs60a-ae at po.EECS.Berkeley.EDU (Class Account) says:
> My name is Seth Piezas and I have been active in lucid dreaming
> I think I have got some sort of solution, however. A little
>while ago I read about fast MRI and the ability to localize responses
>to sayign certain words and thinking certain thoughts.
What you are referring to is the PET (Positron Emission Tomography)
scan. By superimposing PET scan with MRI, one can better see which
region of the brain is responsible for a certain activity. I believe
MRI cannot detect any activities since it simply detects the concentration
of hydrogen ions.
As for PET scan, a radioactive substance (usually glucose) is injected
into blood (it's non-toxic, obviously). Since there is a correlation between
the amount of the blood being used in a certain region of the brain
while performing certain tasks, this technique has been used to localize
the specific functions of brain areas. For instance, it was used to
demonstrate the different functions of generating speech and comprehending
speech were localized separtely in Broca's Area and Wernikle's Area.
Kendell's _Principle_of_Neuroscience_ Chap. 1 or 2 has extensively
referrences. Send me an email if you are interested.
However, I doubt the PET scan locate the difference of two words.
The resolution required would not only be high, but whether the
brain actually works that way is unclear. Your proposal implicitely
implies that memory is localized in such a way that when you think
of one word a certain spot will be active while when you think of
another word, a different spot will be active. I don't know if that's
true. (It surely will be nice...) However, the current thinking is
that the memory is distributed and redudant. Therefore, the two
words may not be easily detected.
> Seth Piezas
Case Western Reserve BioMed Engineering