Researchers have long attempted to unravel the cryptic code used by the neurons of the brain to represent our visual world. By studying the way the brain rapidly and precisely encodes natural visual events that occur on a slower timescale, a team of Harvard bioengineers and brain scientists from the State University of New York have moved one step closer towards solving this riddle. The findings were reported in a September 6th Nature article.
"Visual perception is limited by the relatively slow way in which the neurons in our eyes integrate light. This is why, for example, a Hollywood movie consisting of a series of flickering images appears to us as seamless motion," explains Garrett Stanley, Associate Professor of Biomedical Engineering at the Harvard School of Engineering and Applied Sciences. "However, when the brain responds to some kind of visual event, such as a ball bouncing, the activity of the neurons responsible for sending information can be precise down to the millisecond, despite the fact that the motion of the ball is much slower."
To determine why the brain might encode visual information with such precision, the researchers relied on data obtained by directly recording neuronal activity in animals while they viewed natural scene movies. Doing so enabled Garrett and his colleagues to pinpoint the pattern of neuronal firings in cells that respond to form and motion.
Their analysis of the data suggests that the brain's timescale depends on the nature of the visual stimulus. In other words, the precise timing of the neurons (i.e. their internal clock) changes relative to the timescale of the visual scene. For example, a faster bouncing ball results in more precise brain activity than a slower one. In each case, however, the precision of the neurons' activity was several times that of the speed of the bouncing ball.
It turns out that the extreme precision of the brain's neural response to visual stimuli is, paradoxically, necessary to accurately represent the more slowly changing visual world. The neuron's response must be more precise to recover the important aspects of the visual environment.
"We believe that this type of relative precision may be a general feature of sensory neuron communication," says Stanley. "You can think of it like digital sampling used for audio recordings. The brain 'digitizes' the visual stimulus. As with digital audio recordings, for clear and representational 'playback', the encoding frequencies must be at least double that of the signal information."
In future research, the researchers plan to further clarify why and how the brain encodes visual information across larger networks of cells and across functional units of the brain. They also will investigate how the visual pathway of the brain adapts to changes in the visual scene. They believe cracking the neural code will help other scientists and engineers better "communicate" with the brain. Understanding the speed at which the brain encodes information is critical for designing interfaces such as neural prosthetics, that seek to augment or replace brain function lost to trauma or disease.
Stanley's coauthors were Daniel A. Butts, Weill Medical College of Cornell University; Chong Weng, SUNY College of Optometry and University of Connecticut, Storrs; Nicholas A. Lesica, Harvard School of Engineering and Applied Sciences, and Jose-Manuel Alonso, SUNY College of Optometry. The research was funded by a Charles King Trust Postdoctoral Fellowship, the NGIA, NIH, and SUNY Research Foundation.