When it was time for a project to graduate on (yes, I'm talking about my university days here, this is around 1996), I couldn't find what I was initially looking for but ended up with a reasonably interesting project in Budapest where I developed a, well, groundbreaking statistical method for approximating multidimensional chance distributions.
But enough about that.
What I initially was looking for was to do research on how to operate a computer through brain waves. At the time, not much was there, except for the US military working on how to have pilots fly planes with their brains (and I specifically didn't want to work for the military, let alone the US military) and a vague project, somewhere (I never found out where), where they were trying to have the user of a wheelchair steer the wheelchair with his thoughts. At that time, that project wasn't very successful.
Indeed, after some searches and basically finding nothing, I went for the project in Budapest. Keep in mind, by the way, that the internet was rather small at the time and looking for projects like this was indeed like looking for a needle in an offline haystack.
Anyway, things have slightly changed since and on this year's CEBIT, a 'mental typewriter' was demonstrated. It's finally here folks. Maybe I'm in the wrong line of business after all.