7.27.2007

Brain Scan

I participated in another fMRI experiment the other day. This one attempted to study latent memory.

The experiment comprised three parts. The first was a guided learning session, I believe, lasting 24 minutes. I held a two-button controller and a series of instructions was flashed on the screen of the goggles I wore into the MRI machine, taking me through a decision tree with images of fractals indicating the "state," and the dynamics being controlled by the left and right buttons of the controller.

In the second session, also 24 minutes long, I had to try to reach end states for a given reward ($0.25 for one, $0.10 for another, and $0.00 for another).

Lastly, some anatomical scans were taken to ensure, I guess, that I had a normal brain. This process took about 10 minutes, though it seemed to take too long.

The first stage had a single possible value, and you could make two decisions- left and right. With a certain probability, depending on which button you pushed, one of four images would come up. Then, there was a second step where the process repeated, only the tree collapsed down to three different states there, again with differing probabilities.

The second 24 minute session was the moneymaking phase. The prompt would come up and I was supposed to navigate to the final state, attempting to maximize my earnings. It was fairly simple to do so, as I had figured this was the goal coming in and paid attention mostly to the consequences of the first choice in the learning session. I noticed that one choice led overwhelmingly to outcomes one and three ($0.25 and $0.00 payouts, respectively), and the other overwhelmingly to two and three ($0.10 and $0.00, respectively). I took the first few trials on the payout round to further prune the decision tree.

In the end, I won more money than the experimenter had ever seen. Plenty of money for lunch this week!

After the experiment, I asked if I could get a picture of my brain. Well, he wouldn't give me one, but he showed me some of the anatomical scans with the 3D rendering. The cerebellum looks delightfully leafy. The major blood vessels are huge--and one of my eyes is noticeably lower than the other. My corpus callosum seemed huge, and the lobes were generally quite fluffy in appearance.

No comments: