As you picked up your morning cup of coffee, your brain was actively examining the shape of the mug. This might seem like a simple feat, or at least something that our brain should be able to do with ease.
Yet, considering that our eyes only give two-dimensional information to our brains, there must be a lot of computation necessary in order to perceive three-dimensional forms, both for the sense of vision and the sense of touch.
In the past, many investigators had focused on how the brain represents 2-D figures, which is slightly less complicated. Researchers in the Zanvyl Krieger Mind/Brain Institute, led by Charles Connor, wanted to explore exactly how the brain is able to perceive 3-D objects.
In a recently published Nature Neuroscience article, they found a population of neurons that responded specifically to certain 3-D shapes. For instance, some cells might respond to the point of a pyramid, while others respond to curves taken from the surface of a sphere.
Does this mean that there is a neuron that responds to books, one that responds to coffee mugs and one that responds to pencils? It would probably be impossible to have neurons devoted specifically to one object; we would most likely have much larger brains if this were the case.
Connor and his team focused on the question of a neural code, or a mechanism that translates images from the eye to perceptions in cortex. This code would represent a systematic approach to understanding everyday objects as well as new ones.
They trained rhesus monkeys to look at various shapes that were computer generated. At the same time, they inserted electrodes into certain neurons of the monkey and were able to record these neurons' activities.
Amazingly, they discovered a region in inferotemporal (IT) cortex, a region involved in complex visual processing, that contains a population of neurons tuned to 3-D shape. Each of these neurons responds to certain salient features of shape.
For example, one neuron might respond to a curve in one direction, while another neuron responds in the other direction. Specific neurons will respond only when a certain collection of 2-D features (curves, lines, edges) are combined.
Furthermore, these neurons do not care about the size of the object. Also, the neurons cannot be tricked into responding. The researchers presented visual stimuli of 2-D objects that had been shaded to give a 3-D appearance, like something you would do in art class.
Even though there is the illusion of depth perception, the neurons in the IT cortex will not respond to these cues.
This evidence strongly suggests that the IT neurons are responding to 3-D shape rather than other visual features. Thus, our 3-D perception may be based on neurons that code for certain parts and features of an object as well as neurons that code for more than one part.
It seems that our visual system builds shapes piece by piece rather than having one neuron for every known object. While this research shows strong evidence that the IT region is involved, other cortical areas might also contribute to 3-D shape.
Additionally, it remains to be seen how responses to specific 3-D features can be coded into a response that can differentiate between two similar shapes.
