>You mean someone will combine this with the netflix database and will proceed to de-anonymize this based on what they were thinking during the scan?
There is a saying on the web site Less Wrong that "nobody knows what science does not know." In other words, people tend to be too quick to leap to the conclusion that a particular feat involving science is impossible.
I also note that you have no personal stake in the question. If your scientific career or your ability to continue to win research grants were at risk, you might not be so confident in your belief that it would be impossible for anyone to misuse the data in a way that harms your reputation or the reputations of people you depend on or care about.
(There is a small chance I could be wrong in inferring that that is indeed your belief, in which case, ignore the above.)
Note that I am not asserting that the researchers should limit distribution of the data or that doing so would have net positive effect on society. I am just saying that is a rational decision for researchers who care about their careers to make.
I think that being careful with medical data is an excellent idea. At the same time I can't see any harm coming from releasing this particular data as long as it has absolutely no meta data associated with it. Just the raw scans should be enough to satisfy anybody's curiosity. If they need more for whatever reason then they should have their credentials and needs checked on a case-by-case basis.
Mis-use of such data would require at least a hypothetical scenario, and stating that would bolster the reason why the data was not released without restriction.
The AOL data release was a total fiasco in that it wasn't anonymized properly and because they were not aware of the state of the art.
And it is not like they are not releasing it, they are releasing it! Just not on an instantaneous download.
As for AOL, I'm quite grateful they did what they did, I got quite a bit of mileage out of their data.