Word comes that David Hubel passed away last night. A nobel laureate who studied the visual system, he was a legend in many ways.
First and foremost are his investigations into the basic representations of the world in the visual cortex. It was known prior to his (and, importantly, Torsten Wiesel) experiments that neurons in the retina respond to ON/OFF changes in light intensity at a specific part of the visual field – an area referred to as that neuron’s receptive field. In the illustration above, an ON-center cell responds best to a bright spot surrounded by a dark spot while an OFF-center cell responds best to the inverse light pattern. These cells are excellent at finding the edges of the visual world.
Hubel expected the visual cortex would contain neurons that responded in the same way. As happens so often in science, the visual cortex was uncooperative with his pet theory and did not respond to light and dark patches. Famously, it was only an accident that he discovered that visual cortical neurons respond to moving patches of light and dark! In particular, the cortical neurons will respond to precisely-oriented edges of light; a bar place in an identical location but rotated ninety degrees will elicit no response at all from the cell! Hubel and Wiesel found two distinct types of cells (although there is a contemporary debate as to whether these categories actually are distinct). They were classified into “simple” and “complex” cells. Whereas simple cells will only respond to a bar at a precise angle at a precise location, complex cells will respond to bars of a specific orientation located anywhere within the receptive field of the cell.
Hubel and Wiesel proposed a simple model of how these cells build up their behavior from the retinal cells that they (indirectly) receive input from. Simple cells gather input from a line of ON/OFF retinal cells while complex cells gather input from a collection of simple cells. Here is their original drawing:
On top you can see the proposed pooling of the ON/OFF receptive fields to form a line, and on bottom you can see the pooling of simple cells to enable an invariant response to lines regardless of location. Beautiful and simple! It is my understanding that much of this has been shown to be pretty much on the mark (though I have not paid attention to this subfield in years).
Hubel’s research extended so much further in its characterization of the visual system, but I’m not going to go into that. Rather, it is worth reading the speech he gave upon receiving his Nobel Prize. It is enjoyable and surprisingly gripping! As someone who knows the research inside and out, I still found a lot to learn from it. There are some great historical tidbits including this:
Many of the ideas about cortical function then in circulation seem in retro- spect almost outrageous. One has only to remember the talk of “suppressor strips”, reverberating circuits. or electrical field effects. This last notion was taken so seriously that no less a figure than our laureate-colleague Roger Sperry had had to put it to rest, in 1955, by dicing up the cortex with mica plates to insulate the subdivisions, and by skewering it with tantalum wire to short out the fields, neither of which procedures seriously impaired cortical function (7, 8). Nevertheless the idea of ephaptic interactions was slow to die out. There were even doubts as to the existence of topographic representation, which was viewed by some as a kind of artifact.
Hubel’s work was absolutely fundamental to our current understanding of cortical function. Something like 10% of all searches for ‘neuroscience’ on pubmed are related to vision, which is a shockingly high number for a field studying the whole nervous system. It goes without saying that Hubel’s work is the progenitor of much of this work and what enabled the visual system to be the model system of cortical function that it is today.
If you want to get the chills and see real science done, watch this movie of Hubel and Wiesel mapping out receptive fields.