A new study in mice has revealed never-before-seen details about how the complex visual network forms in them. This research could inform future research on the treatment of congenital blindness. But given the parallels between biological neural tissue and digital artificial intelligence, this research could also help software engineers develop better and more versatile artificial intelligences.
If you could see the web-like nature of the neurons and constructs that make up the brain and sensory systems of animals, you might think it’s just a random complicated mess. But researchers such as neuroscientists are able to examine this chaos and deduce not only discrete buildings, but also determine their functions. Recently, Professor Kenichi Ohki and Assistant Professor Tomonari Murakami from the Department of Physiology at the University of Tokyo and their team studied a particular formation to understand how it forms: the eyesight system.
“The eyes, certain parts of the brain and the neural network that connects them form the vision system. A rough analogy might be a camera connected by a wire to a screen that your conscious self can watch. But an accurate biological description of this system is extremely complicated,” Murakami said. “There are a large number of visual cortical areas involved and these are arranged in layers that form a kind of hierarchical structure. This idea is not new, but it was unclear how the connections between the early stages of this network, or primary areas, and areas involved in processing visual signals, or higher visual cortical areas, form over the course of of development. We set out to find out how this happens.
The team studied the developing vision systems of mice. In particular, they looked at areas called cortical and thalamic regions. By seeing how the neural networks of these regions grow in newborn mice, and when those networks become active, the team was able to describe more generally the mechanisms governing the growth of the visual system.
“As we recorded the increasingly dense web of connections over time, something popped up that surprised us,” Murakami said. “We expected the visual network to form a lot of connections between the cortical area first, reflecting the hierarchical structure of the whole system. But in fact, the parallel neural pathways from the retinas in the eyes leading to the cortical areas form earlier than those among the cortical areas. This new fact adjusts what we know about this area of cortical development.
This study was carried out not only to satisfy curiosity, but also because basic research of this type may form the foundation of future medical research that could improve people’s lives: in this case, the team’s hypothesis that their mouse research can likely explain visual development in primates, including humans. And that in turn could help researchers aiming to treat congenital blindness.
“There is another area of research that can also learn from what we have done here”, Ohki said. “Artificial intelligence is often based on digital artificial neural networks. These are usually structured in several layers, which can give them complex functionality. But now that we’ve shown that at least some biological neural systems develop parallel buildings before layers, software engineers could take inspiration from this to experiment with new design methodologies. Conceivably, this could help them in their goal of creating ever more versatile intelligences capable of solving a wide variety of problems.