Placecell research. That’s what it is called.
In our brains, there are cells that become active when you are somewhere familiar. This is how we can drive home and not even remember taking the exit ramp off the highway.
Anyway, I was recruited to help a memory and learning researcher correlate the physical location of a live mouse to the particular part of the brain that is active.
I decided to use a video processing approach to doing this, but the software to identify a nuique, living thing was not very evolved and living things change shape as well.
SO I created a simpler solution. I used an Infrared LED and visible light filters to turn a 30 inch arena into a black screen with a white dot. That made the software much easier.
The algorithm merely looked for a concentration of “white” and then found the center of the dot. The, Based on projecting 30 inches onto a 200×200 pixel image, I needed to determine which 2 inch square the dot fell into.
That was not too difficult either.
Processing this was pretty fast as I was using C and the image capture driver could produce something like 15 images per second.
The tricky part was taking the location, and integrating that information into a brain activity measauring machine. It did have an output stream, but there were something like 100 sensors.
In the end, I ended up with a 6 dimension array to track the data I needed.
I needed to count and average the number of brain Spikes or firings per brain cell. Then, I needed the ability to show a particular brain cell’s activity correlated with the “where is the mouse” data.
I ended up with a color scale related to brain activity and the UI had a drop down where one could look at a particular cells’ activity drawn out with “where” the mouse was when the cell fired.