Smallest axon diameter measurable by diffusion MRI in humans
In plain words
Most human axons are less than a thousandth of a millimetre across, near or below what the scanner gradient strength can resolve. Whether reliable axon-size maps are possible is unsettled.
Precise statement
Determine the resolution limit on axon radius $r$ from the diffusion signal perpendicular to axons as a function of maximal gradient amplitude $G$ (30 to 50 G/cm in dedicated human scanners), diffusion time and signal-to-noise ratio, given that the signal attenuation scales as $r^4 G^2$ in the wide-pulse limit, and whether the estimated effective radius, weighted toward the largest axons, can be related to histological radius distributions. An answer is the limit in micron and a demonstrated agreement or disagreement with histology.
What would settle it
Ex vivo and in vivo diffusion measurements on the same tissue followed by electron-microscopy histology of axon radii.
Status in the literature
Resolution-limit analyses place the threshold near the upper end of the human axon radius distribution, so only large-axon tails are accessible with current gradients.