It is well known that spatially restricted incoherent (diffusive) motion of molecules inside macroscopically ordered samples can give resolution in the sub-micrometer regime, which is much smaller than the spatial resolution of any magnetic resonance imaging (MRI) experiment. However these two sources of spatial information can be combined in experiments, which are denoted as diffusion-weighted MRI.
However, the situation is somewhat different if
the scope of interest is shifted from confined liquids to gases. Due to their
lower density it is almost indispensable to hyperpolarize them for MRI
experiments, a process by which they can exceed the signal from e.g. protons in
water. The spatial resolution limit in MRI is due to low sensitivity, hence it
is often speculated that by increasing the polarization by 4-5 orders of
magnitude (loosing 2-3 in density) by using hyperpolarized gases, extremely
high resolved images (e.g. of alveolae in lungs) can be obtained. On the other
hand the self-diffusion coefficient increases also by 4-5 orders of magnitude
(when comparing 3He to water).
Such extreme diffusive motion must be taken into account when designing MRI sequences for high resolution, because there is a trade-off between signal loss due to incoherent motion and the coherent ?imaging? of space. These limits are discussed in a simple theory of general spatial resolution and demonstrated in simulations and images of phantoms and lungs of small animals.