How does our brain process and store movements? Scientists find the answer, with implications for multiple diseases as well as machine learning

From our birth, and even before, we interact with the world through movement. We move our lips to smile or to speak. We reach out to touch. We move our eyes to see. We squirm, we walk, we gesture, we dance. How does our brain remember this wide range of motion? How does he learn new ones? How does it do the calculations necessary for us to grab a glass of water, without dropping it, smashing it, or missing it?

Ruth and Bruce Rappaport Medical School Technion Professor Jackie Schiller and his team examined the brain at the level of a single neuron to shed light on this mystery. They discovered that computation occurs not only in the interaction between neurons (nerve cells), but within each individual neuron. It turns out that each of these cells is not a simple switch, but a complicated calculating machine. This discovery, recently published in the Science magazine, promises changes not only in our understanding of how the brain works, but also in a better understanding of conditions ranging from Parkinson’s disease to autism. And as if that weren’t enough, these same discoveries should advance machine learning, inspiring new architectures.

Movement is controlled by the primary motor cortex of the brain. In this field, researchers are able to identify exactly which neuron(s) are firing at any given time to produce the movement we see. Professor Schiller’s team was the first to get even closer, examining the activity not of the entire neuron as a single unit, but of its parts.

Each neuron has branching extensions called dendrites. These dendrites are in close contact with the terminals (called axons) of other nerve cells, allowing communication between them. A signal travels from the dendrites to the cell body and then is transferred through the axon. The number and structure of dendrites vary greatly between nerve cells, as the crown of one tree differs from the crown of another.

The particular neurons Professor Schiller’s team focused on were the largest pyramidal neurons in the cortex. These cells, known to be strongly involved in movement, have a large dendritic tree, with many branches, sub-branches and sub-sub-branches. What the team discovered is that these branches don’t just pass on information. Each sub-sub-branch performs a calculation on the information it receives and passes the result to the larger sub-branch. The sub-branch then performs a calculation on the information received from all its subsidiaries and transmits it. Additionally, multiple dendritic branches can interact with each other to amplify their combined calculus product. The result is a complex calculation performed within each individual neuron. For the first time, Professor Schiller’s team has shown that the neuron is compartmentalized, and that its branches carry out calculations independently.

“We used to think of each neuron as a kind of whistle, whistling or not,” says Professor Schiller. “Instead, we’re looking at a piano. Its keys can be struck simultaneously or in sequence, producing endless different melodies.” This intricate symphony playing in our brain is what allows us to learn and perform endless different, complex and precise movements.

Multiple neurodegenerative and neurodevelopmental disorders are likely to be linked to alterations in the neuron’s ability to process data. In Parkinson’s disease, it has been observed that the dendritic tree undergoes anatomical and physiological changes. In light of the Technion team’s new findings, we understand that as a result of these changes, the neuron’s ability to perform parallel computations is reduced. In autism, it seems possible that the excitability of dendritic branches is impaired, leading to the many effects associated with the condition. The new understanding of how neurons work opens up new avenues of research into these and other disorders, in hopes of alleviating them.

These same results can also serve as inspiration for the machine learning community. Deep neural networks, as their name suggests, attempt to create software that learns and functions somewhat similarly to a human brain. Although their advances are constantly in the news, these networks are primitive compared to a living brain. A better understanding of how our brains actually work can help design more complex neural networks, allowing them to perform more complex tasks.

This study was led by two of the MD-Ph.D. of Prof. Schiller. candidate students Yara Otor and Shay Achvat, who contributed equally to the research. The team also included postdoctoral fellow Nate Cermak (now a neuroengineer) and Ph.D. Hadas Benisty, along with three collaborators: Professors Omri Barak, Yitzhak Schiller, and Alon Poleg-Polsky.

The study was partially funded by the Israel Science Foundation, Prince Funds, Rappaport Foundation and the Zuckerman Postdoctoral Fellowship.

Comments are closed.