fMRI and PIN numbers?

[penguin42]

Hi,
I was curious, is fMRI upto trying to figure out peoples PIN numbers?

The experiment I suggest is that you stuff someone in an fMRI with a numeric keypad and get them to tap out on the keypad a series of random numbers and see if you can discern a pattern for particular numbers.

Then after a while you suddenly show them a picture of an ATM and a card and get them to think of the PIN - would that work ?

Dave

[Paralith]

Hopefully neuro will be able to shed more light, but my gut reaction is no. An fMRI detects increases in blood flow to certain areas of the brain, not actual neuronal activity itself. It shows you which parts of the brain are awake and being used, but there are a lot of subtleties about the neuron activity itself that are simply not captured by the information you can get with an fMRI.

[penguin42]

Yes I'm aware it tracks oxygen usage rather than individual nerve spikes; but I was thinking about the experiments done where they were tracking when the motor cortex was active versus when you believed you had made a decision. I realise that was controversial; and I found a reference saying the decision was whether to use the left or right hand which I guess would be much different areas of the brain than individual finger movements.

Dave

[CanadysPeak]

Yes I'm aware it tracks oxygen usage rather than individual nerve spikes; but I was thinking about the experiments done where they were tracking when the motor cortex was active versus when you believed you had made a decision. I realise that was controversial; and I found a reference saying the decision was whether to use the left or right hand which I guess would be much different areas of the brain than individual finger movements.

Dave

At present, they are identifying concrete nouns. Thus, you would have to do some association thing where "one" is always associated with Lee Marvin, for example and "two" is always associated with kangaroos. Even then, I believe the hit rate is considerably less than what you'd need for a PIN. Perhaps you could figure out how to do a checksum?

[BioWizard]

Neuro will probably have a definitive answer, but I don't think we know enough about how a human brain stores numbers to be able to extrapolate fMRI readouts in the manner your describe.

[penguin42]

I'd wondered if there was enough from the motor movements that were recorded during the 1st phase to allow it to be inferred rather than actually having to know about how the brain stored numbers. So you wouldn't actually be looking for the thought '1' you would be looking for the motor neurons about to hit the 1 button.

Dave