Limits of neuroscience are tested by a massive new data collection
A video is required viewing in nearly all first-year neuroscience classes. The visuals aren’t impressive; all you see is a bar of light moving and rotating over a dark screen while distant explosions can be heard in the background. The content seems dry at first, but then you realise that each pop represents the activity of a single neuron in a cat’s brain as it watches the bar on the screen move. A final climactic burst of frenetic activity occurs when the bar reaches a certain location and lies at a certain angle. This neuron is sending a very strong signal that it is really interested in that bar.
This experiment, conducted by David Hubel and Torsten Wiesel in the 1960s, provided crucial insight into the functioning of the visual system. Since then, neuroscientists have implanted thin metal electrodes into the brains of mice, finches, and monkeys to see and study individual neurons and learn their triggers. There are neurons that only react to certain colours or forms, some that only react to certain points in space or the direction in which one’s head is pointed, and still others that only react to certain facial characteristics or the entire face.
Anne Churchland, a professor of neuroscience at the University of California, Los Angeles, notes that “everyone always wants more neurons,” despite the fact that single-cell study has shown the brain to be a powerful engine. As a matter of basic statistics, having more data from an experiment is always preferable. However, when studying neurons in isolation, researchers hit analytical roadblocks. Neurons in the prefrontal cortex, the area of the brain responsible for planning, decision-making, and social conduct, respond to such a wide variety of stimuli (visual characteristics, tasks, decisions) that scientists have been unable to categorise their function, at least on an individual level. When an animal views an orientation cue, just a small percentage of its neurons fire, and this is true even in the primary visual cortex, the portion of the brain recorded by Hubel and Wiesel.
Using the methods developed by Hubel and Wiesel, studying more than a few neurons simultaneously was not practical. Yet engineers have continued to push that limit, ultimately leading to 2017’s introduction of Neuropixels probes. Small enough to insert several into an animal’s brain, a single probe measuring only one centimetre in length and built of silicon can listen to hundreds of neurons simultaneously. The Allen Institute, founded by Microsoft co-founder Paul Allen, is a non-profit research organisation that used six Neuropixels probes to record simultaneously from eight regions of the mouse visual system. The institute unveiled its findings in August, which included information from 81 mice and the actions of over 300,000. All researchers are welcome to utilise the data at no cost.
The publication of this data collection, three times larger than the previous record holder, enables scientists to examine the coordinated behaviour of massive networks of neurons for the first time. Those hitherto inaccessible facets of cognition may finally be within scientists’ grasp because to the scale at which the data is being collected. Shawn Olsen, a key investigator on the project from the Allen Institute, has said, “We want to understand how we think and see and make decisions.” Furthermore, “it simply does not occur at the level of single neurons.”
Exactly how to process all that information is the current obstacle. Huge data sets can be challenging to work with, especially when it comes to sharing and downloading. For many researchers, the challenges of analysing such large data sets are well worth it in order to study the brain in its natural environment.
The brain, as seen by Hubel and Wiesel, is like an assembly line, with different sets of neurons performing different tasks. If you show someone a red balloon, their neurons that respond to red and their neurons that respond to circles will do so separately. Unfortunately, the brain is so intricately wired that no single neuron ever acts in a vacuum, therefore this method has never been a good fit for how the brain truly operates. Professor of neuroscience at Columbia University Stefano Fusi states, “The brain is not looking at one cell at a time.” They are observing thousands upon thousands of other neurons. Consequently, we need to share the same point of view.