Happy New Year!
Last week, John’s commentary with Michael Platt on recent results by Read Montague’s group at Virginia Tech went live. While the commentary is behind a paywall, the original article is open access, and we encourage you to take a look.
The study, spearheaded by Ken Kishida, looked at dopamine levels in humans undergoing surgery for deep brain stimulation to treat their Parkinson’s Disease. In P[λ]ab, we perform studies under similar conditions in collaboration with surgeons at Duke, and while it’s tough to pull off a well-controlled study in the operating room, the chance to record data from individual nerves inside a human brain is worth the trouble and then some.
What makes the Kishida study particularly interesting, though, is that it takes advantage of a special probe developed by Ken and Stefan Sandberg that allows for the direct measurement of dopamine levels in the human brain. This probe is an adaptation of the same measurement technology (fast scan cyclic voltammetry) used in other studies to read out dopamine levels in rats with sub-second accuracy. In this case, the investigators examined dopamine responses to gains and losses in a fictitious stock market, a task Read’s group has used before.
In this case, though, the results were pretty surprising. What turned up in the dopamine measurements were not the typical reward prediction errors (the difference between expected and actual outcomes), but something more complex. We cover some of the implications of this in the commentary, but the details, as usual, are in the paper, and well worth checking out.