Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!sun-barr!rutgers!mit-eddie!media-lab.media.mit.edu!minsky
From: minsky@media-lab.media.mit.edu.MEDIA.MIT.EDU (Marvin Minsky)
Newsgroups: sci.bio
Subject: Re: Simulated Vision Article Search
Message-ID: <5677@media-lab.media.mit.edu.MEDIA.MIT.EDU>
Date: 19 Apr 91 14:25:46 GMT
References: <72559@brunix.UUCP> <41524@netnews.upenn.edu>
Reply-To: minsky@media-lab.media.mit.edu (Marvin Minsky)
Organization: MIT Media Lab, Cambridge MA
Lines: 48

In article <41524@netnews.upenn.edu> rowe@pender.ee.upenn.edu (Mickey Rowe) writes:
>In article <72559@brunix.UUCP> cs012136@cs.brown.edu (Garth
>Shaneyfelt) writes: 
>>A little while ago someone posted a reference to a series of
>>experiments involving placing a wire mesh on the occipital lobe of
>>blind persons which allowed them to "see" a 3X3 grid.  
>> 				Thanks in advance 
>>						-Arioch
>
>Are you sure that you're remembering this correctly?  It sounds a bit
>like bad science fiction to me.  I've spoken to a neurologist who
>performs cortical stimulation on epileptic patients prior to surgery,
>and he says that direct stimulations lead (sighted patients) to see
>irregular blobs of light that are essentially uncontrollable (that is
>the experimenter has little control over the size, shape, and location
>of the percept). 
>
It was good science.  Look at the book, @i[Visual Prosthesis,] Academic 
Press, 1971, pp. 315-319.   A neurosurgeon named Brindley implanted an
8X8 array of inductively coupled electrodes on the occipital cortex of
a woman who had become blind because of bilateral retinal detachments
(a thoroughly peripheral injury that almost surely left her cortex in
normal condition).  The pad was under the skull and, I think, under
the dura mater, but I'm not sure of the latter.
  Of the 64 electrodes, as I recall (I can't find the book here) about
8 failed and about 8 hurt -- presumably because of stimulating nerves
in the pia mater -- but the remaining 48 electrodes produced images which she
described as being like half a matchstick at arm's length.  And they
were indeed arranged topographically, so that Brindly was able to
produce various recognizable patterns.

   One deficiency was that parts of the fovea were missed because of
lying in a sulcus -- space between convolutions.  Brindley did not
attempt to push his silicone pad into that fissure.  In fact he did
not want to risk any serious brain injury, so the device was removed
after a few days, as I recall, because of fear that the implant might
cause irritation and scarring.  

  Because of the lawyers and ethicists, no such experiment could be
done now, or (?) ever again.  The lady actually volunteered for the
experiment because of her long interest in neuroscience!  But no
hospital would now cooperate, I think.

  Last word.  When Brindley did the experiment in the late 60s Warren
McCulloch told me that if I wanted to do something like it, it must be
done quickly because in later years it would become ethically and
politically impossible.  I thought he was being depressed or
something, but evidently he was just correctly predicting that trend.