The human eye can handle 10 million bits per second

Researchers at the University of Pennsylvania School of Medicine estimate that the human retina can transmit visual signals equivalent to an Ethernet connection. Ethernet is the most popular local area network connection standard, allowing data transfer between computers at speeds of 10 to 100 million bits per second, with actual throughput of 2 to 3 million bits per second. The study aims to compare the information processing system of the nervous system and the artificial visual system to provide the basis for designing artificial visual systems.

Many basic studies aim to understand the type of information that the brain receives; This study is directed to the question of how. Using intact retina from guinea-pigs, scientists recorded branches of electrical impulses emitted from ganglion cells with a miniature network of multi-electrode antennas. Surveyers calculate that the human eye can transmit approximately 10 million bits of data every second .

The retina is actually a part of the brain that grows in the eye, taking on the processing of nerve signals when light is detected. Ganglion cells will relay data from the retina to the brain center. Other neurons in the retina perform visual analysis. Nerve fibers connected to ganglion cells on the retina along with other supporting cells make up the nerve nerve that will send signals to the brain.

The team said there are about 10 to 15 ganglion cell types concentrated in the retina that are responsible for classifying motion & working together to fully visualize the brain. The study was conducted on assessing the amount of information transmitted to the brain of these seven ganglion cell groups.

The hamster retina is placed on a plate and is viewed with 4 types of biological motion. For example, a salamander swims in a water tank, representing an object - motion stimulation. The electrical impulse branches will be captured & recorded by a multi-electrode antenna network. Researchers will classify cells into two categories: " agile " or "sluggish" based on their speed.

The magnified image of a ' fast ' and ' sluggish ' cell type in the hamster retina. (Photo: Uphs.upenn.edu)

The researchers found that there are different branching pathways between cell types. In general, in every second, agile cells stimulate many branches of electrical impulses and get a lot of feedback. In contrast, slow cells stimulate only a few branches of electrical impulses every second and get very few feedback signals.

But what is in each relationship between the electrical impulse branches and the information sent?"It is a combination of different types of information transmission lines. These lines have many different meanings," said Vijay Balasubramanian, a professor and doctor of physics at the University of Pennsylvania. "We determine the number of lines & calculate how much information is transmitted by each line, ie the number of bits transmitted per second".

Calculating corresponding to each type of retinal cell, the team estimates that about 100,000 guinea-ganglion cells take on 875000 bits of data per second. Because the number of " slow " cells is so large, they take on the role of information transmission. With 1000000 ganglion cells, the human eye can transmit data approximately the connection speed of an Ethernet network, equivalent to 10 million bits per second.

"The more electrical impulses that work, the higher the metabolism," said graduate student Kristin Koch, who works at Sterling's lab. We found that the cells were slower to "metabolize" cheaper, because they sent more information to each branch. If the information is sent at a high rate, the brain will use fast channels. But if an information is sent more slowly, the brain uses slow channels because the brain will only have lower metabolism. "

Peter Sterling, a professor of neuroscience in Pennsylvania, said: "Until now, the slow cells have not been fully studied. What is surprising is that the activities they decide on and control are important. slow cells take on most of the crucial role in all the transmitted information ".

Co-authors of this work include: Judith McLean and Michael A. Freed of the University of Pennsylvania, Ronen Segev and Michael J. Berry III, of Princeton University.