The sounds and noises that a typical MRI scanner produces are problematic for several reasons:
they give - annoying as they are - a strong feel of uncomfortableness to any subject or patient
they are quite loud, dominant, and distracting and disturb any kind of processing a task, not only auditory related tasks (but obviously these)
they especially strongly disturb the measurement of any subtle emotional response.
I have heard of several solutions to this problem (partly successful):
physical (hardware-based) attenuation and prevention
software-based attenuation or "softening"
making the sounds and noises part of the experimental design, e.g. in this arcticle: The MR tomograph as a sound generator
Some overviews can be found here:
My question is:
Are there other - maybe completely different - approaches to solve the problem of sounds and noises of MRI scanners?
(I assume that not everyone would have immediately come up with the third approach!)