A Deep Dive into EEG with Firefly Neuroscience's Dave DeCaprio

I started out as an audio guy.

I'm sitting, literally my computer is sitting on top of a Rupert Neve Designs console.

I have a bunch of A to D and D to A converters.

So when we're looking at EEG data, are we measuring it at high fidelity? Maybe you could just tell us something about what we get out of the brain.

For those who've never heard of EEG before, what are we measuring with EEG and how detailed is that information?

What you're ultimately measuring is your neurons firing.

When your neurons fire, there's an electrical impulse that happens.

billions of neurons in your brain.

With EEG, we're putting sensors on the outside of your scalp.

We don't really have the ability to measure one neuron at a time, but what actually happens in your brain is there are these sort of networks of how neurons fire, and they don't fire individually, they fire in groups.

With EEG, what we start to measure is these groups of neurons from these brain networks, when they all start to fire together, All of those little signals, the little tiny bit of electrical pulses from one neuron add up across billions of neurons, and then you can start to actually get a signal that you can read on the top of your scalp.

And so we're measuring changes in voltage? Is that the right way to look at it? Yeah.

Just electrical potential.

A little bit different from sound, but basically if your brain was like a whole bunch of people clapping and some louder and some softer, we're hearing that kind of far away on the surface of the scalp.

And if people are yelling loud enough, we can catch it.

And these are very mercurial crowds.

So sometimes they, it kind of goes up and down in, in something like real time.

Yep, yep.

How quickly are those changes happening that you're measuring, at least? Do we know the theoretical limit of how fast they change versus how fast they measure?

Generally, e.g., 250 hertz is the 250 samples a second.

There's sort of math on the signal processing side that at 250 samples a second, you can roughly measure things that are half that.

So you can measure cycles that are 125 times a second, which is about 8 milliseconds.

And I think...

It's an interesting question.

I'm sure if you look at the biology of how fast neurons fire, that's something I should know more

I started out as an audio guy.

I'm sitting, literally my computer is sitting on top of a Rupert Neve Designs console.

I have a bunch of A to D and D to A converters.

So when we're looking at EEG data, are we measuring it at high fidelity? Maybe you could just tell us something about what we get out of the brain.

For those who've never heard of EEG before, what are we measuring with EEG and how detailed is that information?

What you're ultimately measuring is your neurons firing.

When your neurons fire, there's an electrical impulse that happens.

billions of neurons in your brain.

With EEG, we're putting sensors on the outside of your scalp.

We don't really have the ability to measure one neuron at a time, but what actually happens in your brain is there are these sort of networks of how neurons fire, and they don't fire individually, they fire in groups.

With EEG, what we start to measure is these groups of neurons from these brain networks, when they all start to fire together, All of those little signals, the little tiny bit of electrical pulses from one neuron add up across billions of neurons, and then you can start to actually get a signal that you can read on the top of your scalp.

And so we're measuring changes in voltage? Is that the right way to look at it? Yeah.

Just electrical potential.

A little bit different from sound, but basically if your brain was like a whole bunch of people clapping and some louder and some softer, we're hearing that kind of far away on the surface of the scalp.

And if people are yelling loud enough, we can catch it.

And these are very mercurial crowds.

So sometimes they, it kind of goes up and down in, in something like real time.

Yep, yep.

How quickly are those changes happening that you're measuring, at least? Do we know the theoretical limit of how fast they change versus how fast they measure?

Generally, e.g., 250 hertz is the 250 samples a second.

There's sort of math on the signal processing side that at 250 samples a second, you can roughly measure things that are half that.

So you can measure cycles that are 125 times a second, which is about 8 milliseconds.

And I think...

It's an interesting question.

I'm sure if you look at the biology of how fast neurons fire, that's something I should know more