Brain cells in a “Brain Dish” learn to play pong within 5 minutes

This study, published in the journal Neuron, shares something unprecedented.

Researchers led by Bret J. Kagan placed human and mice brain cells in a special kind of petri dish.

This dish, called the brain dish, provides and studies electrical impulses among and within the cells.

These electrical impulses are the “language” of cells.

Without the ability to transmit these impulses to each other, cells cannot communicate with each other.

If one has to give any feedback to the cells, that also is possible only through electrical impulses.

So, in brief, this study:

A. Created an environment where brain cells derived from stem cells in humans, and foetal cells from mice, could communicate within and among themselves. Their electrical impulse activity was monitored.

B. Feedback was given to the unit of cells on how they were performing at a task (in this case, playing a simple computer game).

C. Researchers checked whether there was activity between them, and more importantly, whether that activity led to improvement in their performance on the game.

The entire experiment has been visualised by the researchers here.

The Result

The biological cells started communicating with each other almost immediately and within 5 minutes, they were winning the game.

Why is this important?

A neural network is a group of nerve cells that work with each other. Nerve cells are basically communication cells.

The two types of neural networks – Biological neuronal network (BNN) and artificial neural network (ANN) Image Credit: clevertap.com

There are two types of neural networks.

A. Biological Neural Networks (BNN)

Image Credit: Commons.Wikimedia.com

These are the neural networks in the bodies of living beings. To understand how our neural networks work, scientists have been trying to study the electric impulse activity in the brain for many decades. We now know, thanks to this work, that the human brain has dedicated areas for different types of activity.

B. Artificial Neural Networks

Based on whatever we know about how the brain operates, we have created computer networks that rely on each other’s resources to generate great computing power and run complex algorithms.

These computer networks are called ANN – Artificial Neural Network.

While we know that neurons talk to each other using electrical impulses and that there are specialised regions in the brain, we still do not know HOW these neurons draw upon each other and learn as a cohesive unit.

The second challenge is – we have been able to conduct learning studies on brain cells in vivo – in other living organisms, but never in vitro – in a glass dish.

This is the first time that brain cells have demonstrated in a lab dish how they learn from and with each other and are able to learn very fast.

This means, in short, that we have taken stem cells and created working brain cells in a dish.

This also means that we can now learn from BNNs to create more efficient and powerful ANNs.