Slashdot Mirror
Scientists Use fMRI To (Sort of) Read Minds
NigelTheFrog writes "Researchers in England have used fMRI to map the activity in volunteers' hippocampuses. From these scans, they could pinpoint exactly where they were in a virtual reality landscape. 'Specific parts of each participant's hippocampus were active after that person had navigated to particular places in the room. A few practice rounds provided fodder for creating algorithms for each participant that correlated different brain activity patterns with different virtual locations. The algorithms, the team found, could in turn "predict" new virtual locations, not those used during practice rounds, based on each person's pattern of brain activity.'"
Shouldn't that be hippocampi?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Well it would be simillar to that. aluminum is not magnetic so you wouldn't notice anything when getting into the machine, but as soon as the scan started, the ultrafast sweeping of the gradient magnet's fields that's needed to perform echo sequences with the time resolution relevant to fMRI would create HUGE ohmic heating in the conductive metal and severely burn you, if not light your hair on fire.
- "Hear that?! The percolations are imminent! Cease your ingress!"
These studies are actually about as unrelated as you can get. Both are trying to decode a parameter from fMRI BOLD response (as do hundreds of other studies), but they are looking at very different brain areas and different tasks. There are hundreds of rodent electrophysiology studies showing that specific hippocampal cells respond when the rodent is in a specific position in its cage. There is even a study showing spatially selective cells in the human hippocampus (Ekstrom et al., 2003).
The really remarkable thing about this study is that they have managed to find spatially selective voxels in an fMRI. Electrophysiology has provided similar results, but records from a region hundreds of thousands of times smaller and requires invasive surgery. Results here are patient-specific; there is no spatial map in the hippocampus, and place cells (or "place voxels," in this case) are located in different places for each subject.
Scientists know enough about the brain today that we are can design procedures that target-train a cognitive processes and see results in the neural tissues responsible for those cognitve processes.
An example with working memory. If you commit to doing the n-back test 20 minutes daily, and gradually increase difficulty, you will not only see improvements in the post-test working memory performance, but you will see structural changes in the neural tissues in the cortex responsible for working memory. Moreover, to the extent that working memory improves recall and recognition memory, you will see improvements there too.
A Good Troll is better than a Bad Human.
CBS 60 Minutes did a piece on FMRI at CMU in January. Watch it -- http://www.cbsnews.com/video/watch/?id=4697682n.
Here in the US we call it the "freshman fifteen". Or how we referred to it when I was in college: "dorm butt" (and "dorm gut")