An interesting blog article named "Hallucinatory neurophysics
" can be found here: http://preposterousuniverse.blogspot.com/2005/02/hallucinatory-neurophysics.html. Unfortuntally I wasn't able to find the paper by now. (There are, by the way, some nice videos about hallucinatory drugs on youtube, and some good articles in Spektrum). And Jack Cowans appointment to math and neurology also sounds pretty interesting. Computational neuroscience seems to be a field of its own.
I came there via a blog about "imagining more than 3 dimensions" at http://blogs.discovermagazine.com/cosmicvariance/2009/03/30/why-cant-we-visualize-more-than-three-dimensions/.
That previous blog post reminded me of the "Imagining the tenth dimenson" video ( http://www.youtube.com/watch?v=XjsgoXvnStY ) which I first watched years ago. This version contains some annotations taken from the book (for which the video is supposed to be pomotional material I guess). I found them not too easy to follow and I doubt the correctness of some claims. But to discuss this I'd need the book, but one at the first glace strange thing is his integration of Everett's many-world interpretation of quantum physics.
edit: I found a paper which seems to have a similar focus: "Geometric Visual Hallucinations, Euclidean Symmetry, and the Functional Architecture of Striate Cortex" at http://www.math.utah.edu/~bresslof/publications/01-1.pdf