Brain scans admissible in Indian Court

[Stark]

I also have questions about whether it can tell the difference between a memory of an act, or thinking about an act that you did not do.

Or even 'thinking about a similar act you may have witnessed in order to understand the question'. I mean, if I heard the question 'did you blow up the government building, I might hit my memory to pull up some recollections of exploding buildings in movies. It's not like association is voluntary, to my understanding.

[Winston Blake]

One problem I can see with this is remembering dreams. What if someone sees a news report about a local knife crime, and has a nightmare about it? Can fMRI distinguish remembering real events from remembering vivid dream experiences?

[Fire Fly]

There's a lot of evidence from fMRI studies show that this does work but the big problems I see are that you need good baseline images to compare with, the margin of error is quite significant (at least I would consider it to be if you're going to convict someone of a crime), and its possible to train your brain to suppress activity.

But, supposedly if you lie, your brain is aware that you've told a lie; you have to fabricate the lie, you need to know the consequences of the lie, you need to know work out all of the intricacies of lying where as a person telling the truth does not. So a person who lies will have their brain light up like a Christmas tree whereas the person telling the truth, their brain will be relatively quiescent.

If they used converging lines of evidence combined with multiple brain scans (fMRI, EEG, etc), I would be less cautious of a guilty conviction.

[Fire Fly]

One problem I can see with this is remembering dreams. What if someone sees a news report about a local knife crime, and has a nightmare about it? Can fMRI distinguish remembering real events from remembering vivid dream experiences?

Real events and dreams are nothing but memories, so fMRI will see them as the same.