Katana_Mind

What I was trying to say with my first post was that unless your non invasive headset also has a way of actively passing information to the user, a lot of the proposed uses for the technology are no longer possible. An example would be a user of a prosthetic limb - ideally you want your implanted electrodes to not just detect the impulse to move the limb, but also to return information to the brain about touch, orientation etc (in some fashion that can be interpreted by the brain, but that is a whole other problem). I agree that for the enthusiast user the non invasive approach is always going to be more attractive, but for therapeutic uses I think that invasive is most likely the more appropriate.

Motivated

1) Couldn't the neurons be stimulated by rTMS or something similar?

Or couldn't the replies just be sent through displays in your contact lenses?

2) & 3) Yes, there are problems with both methods. The least of which would be that MRI machines are giant hulking monsters.

Basically I'm positing enormous advances in remote brain imaging, data processing speeds, miniaturization and remote neural stimulation by magnetic fields.

The ideal would be a hat you just put on and it reads your brain and sends complex messages to it. I don't know if it's merely an engineering problem or just scientifically impossible.

But I'm just saying that for every attempt to hack the meat directly there is non-invasive fyborg solution lagging not far behind. Why take steroids to boost my strength when there's a perfectly good powered exoskeleton or forklift nearby? Why jam a soon to be obsolete PDA socket in my skull when my PDA can be made noninvasive and wearable?

corpore-metal

The second and third problems are latency and resolution respectively. EEG based systems have very low latency, but the resolution of infomation they provide is very low. fMRI is the opposite - very high resolution, but the post proccessing of the data means that it can take up to 10 seconds to get infomation on the neural activity of the user.

Of course there are just as many problems with the direct implant approach as well, not the least of which is that long term implantation of electrodes leads to the death of nearby neurons...

Motivated

I think a better way to do this brain-computer interface is wear some kind of helmet that's a portable fMRI/EEG/PET thing. It just scans your brain in realtime at molecular resolution. Basically it reads your mind.

corpore-metal

Katana_Mind

Jeff-Minor

Jeff-Minor