Brad Theilman









Watching a brain in action

30 Jan 2019

As I’m writing this, I’m in hour 35 of the experiment I started yesterday. It is going very well so far - excellent signals, no catastrophes. Some of the auditory responses are so strong, they remind me why I’m even here in the first place.

Once you get the bird situated in the apparatus and the probe installed, these acute experiments settle into a rhythm. You move the probe to a new position in the brain, play some test stimuli to look for auditory responses, and then run a block.

The block is made up of repetitions of a fixed library of stimuli that you use to test whatever hypothesis the experiment is designed for. As I mentioned yesterday, this experiment is to get data on longer, minute long songs. We have 12 of these songs repeated 20 times each per block. So, along with the inter-trial intervals of about 5 seconds, this ends up making each block last about 4 hours long.

While a block is running, you only need to pop your head in every few minutes to make sure the bird is ok, and that the recording system is functioning properly. Otherwise, you want to reduce any sources of interference - noise, electrical,etc. So we turn off all the lights and shut all the doors.

I’m not sure that all the precautions we take make a huge difference, but despite what you may think, there’s a lot of superstition that goes along with electrophysiology.

We’ve been running blocks constantly since about 10:30 a.m. yesterday.
But before I started this latest block, I had the chance to have a little fun with the experiment and demonstrate why neuroscience is so cool.

I had driven the electrode to a depth of 2500 micrometers below the surface of the brain. This should have placed our electrode squarely in one of the secondary auditory regions of the bird. It was clear right away that the spikes in this location were ideal. They were large and stable - you could endlessly watch them fire in mesmerizing patterns we don’t yet understand. Since we have a rotation student in the lab this quarter, I wanted to demonstrate what an auditory response looks like. Usually we just do a few claps or taps on the box to make sure the neurons respond. Even more appropriately, we play test songs recorded from real birds.

However, it’s always more impactful to be the cause of some effect yourself. So I opened the box and began speaking to the bird. Now, this was not entirely unexpected, but immediately you saw a complicated pattern of neural activity explode on the computer monitor!

What made it so striking was how loud the response was. I’m not talking loudness of the sound, but the magnitude of the change in the neural activity in response to my voice. Many large spikes appeared. And these spikes were not just a random burst of activity - you could observe structure in how the spikes danced around the screen due to the modulations in my voice. It was so dynamic you couldn’t take your eyes away - much like watching a campfire.

So the rotation student and I both sat there for a few minutes making all kinds of strange noises, experimenting with different ways to drive a large response. It was very similar to how the early experiments were done on the nervous system, trying any sort of stimulus you can to get an effect.

This isn’t my first recording, and I’ve seen these kinds of effects before, but today it really hit me for whatever reason. We were watching in real time a bird’s brain process our own voices. Our readout of this processing was a beautiful multicolored pattern of activity on the screen. And this mystery remains: how does what we see happening in the bird’s brain reflect what is happening in the real world!? Is this even an appropriate question?

The practical details of what made it so striking are that we are using a new recording system that has much lower noise than our previous one, we are using a denser, more sensitive electrode, and I’ve gotten better at these recordings through graduate school. But all of that was just technicality - seeing the response of a live brain to my own voice in real time reminded me why I went to graduate school in neuroscience in the first place, and has reopened all the naive questions about the brain we still don’t know the full answers to.

I took some videos of the responses that you can see here:

Responding to my voice

Responding to whistles