Corrupted file for brain DICOM data

[mmuckley]

As documented in Discussion #287, the tar.gz file for the brain DICOM data seems to be corrupted. I have downloaded a couple of times, observing an SHA256 hash of 68e19447b428ab723d62bcd09a2b859a06eac52f4d5751d5b66570de0f7064db vs. the expected 04da798eda6f49986f79c8b8758d80d12d943d29776070befdf54cb5682359de. I'm currently trying to reach out to NYU to see if we can recompress the data and reupload an uncorrupted file.