Brain

tim pattison timpatt at augean.eleceng.adelaide.edu.AU
Wed Mar 11 23:54:36 EST 1992


In article <4379 at cluster.cs.su.oz.au> tom at minnie.cs.su.OZ.AU (Thomas James Jones) writes:

>       4) What evidence of RECURRENCY (feedback) is there in the
> activity of the brain with respect to a particular task. Personal view:
> tasks of perception are 'feed-forward' , do not require feedback
> tasks of reasoning do require feedback.

On the contrary, there is good evidence for various types of recurrency
in early visual processing.

Lateral (recurrent) interactions between the simple cells in primary
visual cortex --- mediated probably by interneurons --- have been
demonstrated using various techniques, and are believed to mediate the
oriented receptive field profiles (RFPs) observed for these cells.

There is also a sizeable corticofugal output from layer VI in primary
visual cortex to the lateral geniculate nucleus (LGN). The LGN may be
(crudely) viewed as a relay station between the retina and primary visual
cortex. This sort of feedback may be important in adaptively tuning the
response properties of LGN neurons according to the input stimuli and our
recent perceptions thereof.

Finally I would not be surprised if primary visual cortex itself receives
feedback from higher visual centres, although I don't know if this is or
isn't the case -- my studies haven't extended that far! Such feedback
could be advantageous for expectation: it is easier to interpret a visual
scene if you know what to expect. Consider the classic black and white
blob picture of a Dalmation (oops, what a give-away!).

Feel free to contact me via email if you require references for the above
assertions.

Cheers,
Tim Pattison.
-- 

Tim Pattison
Dept of Electrical and Electronic Eng.
University of Adelaide



More information about the Neur-sci mailing list