fMRI is better suited for long-ish tasks. EEG-based techniques are better for responses to sudden events. It would be interesting to analyse the brain activit

fMRI is better suited for long-ish tasks. EEG-based techniques are better for responses to sudden events. It would be interesting to analyse the brain activity while devising algorithms to solve problems, for example. The algorithm design doesn't have to be explicit. Just ask people to solve a problem with different solutions, some of which are algorithmically more elegant than others. You could also ask programmers to write programs in the MRI, and check what brain regions are activated. I wouldn't be surprised if functional and imperative programming activated different brain areas.