seminars & bulletins

2009-2010
PBIO SEMINAR SERIES

THURSDAYS in
G-328
10:30am (unless otherwise noted)
Refreshments follow the seminar

pbio news

pbio special events

general notices

health sciences bulletin

seminar series archives

2008-2009
2007-2008
2006-2007
2005-2006
2004-2005
2003-2004

pbio seminars rss feed
hs bulletin rss feed

 

 


PBIO SEMINAR SERIES

Lettvin's bug detector: multiple systems in the visual system

October 29, 2009

Harvey Karten

Professor of Neurosciences UC San Diego
Speaker's website

Host: Adrienne Fairhall and Ed Rubel
Co-hosting Department: Bloedell Center

The vertebrate visual "system" consists of multiple parallel central channels generated by 18-20 different types of retinal ganglion cells. The "information content" of each of these ganglion cells remains obscure. The ganglion cells form the optic nerve and terminate in varying degrees of complexity upon 15-20 different cell types organized into 6 major clusters - the dorsal nucleus of the lateral geniculate, the ventral nucleus of the lateral geniculate, the optic tectum, pretectum, accessory optic system and suprachiasmatic nucelus. The vast majority of research on the visual "system" has centered on studies of the geniculostriate "system". There is no organizing principle that provides a model as to how these multiple "systems" are integrated into a functional interdependent "multi-system". Attempts to account for the operations of the visual system that center almost exclusively on the highly retinotopic geniculostriate system have ignored major contributions of dedicated pathways concerned with extracting unique features, such as Lettvin's Bug Detector of the optic tectum, first described in his landmark paper "What the Frog's Eye Tells the Frog's Brain" (Proc. IEEE 1959). Analysis of the anatomy and physiology of Lettvin's Bug Detector provided a lovely example of Neuromorphic Engineering, and revealed a surprising source of motion signals to the cortex. This "subsystem" relies on a markedly different information processing paradigm than that employed in the geniculostriate system. I hope that the audience will help me understand the marked dichotomy of the vertebrate visual system in coding of motion vs. retinotopy.