This thesis contains four studies, each of which involves measurement of water-NMR relaxation in peripheral nerve in some manner. The central basis of these studies is that peripheral nerve water exists in three broadly unique environments—myelinic, intra-axonal, and extra-axonal.

In the first study, multi-echo imaging identified three transverse relaxation (T₂) components in the frog sciatic nerve, including a long-lived component (T₂ > 200 ms) which previously had only been identified in vitro. The existence of a long-lived T₂ component indicated echo times of 200-300 ms may provide maximal contrast-to-noise (CNR) (nerve to muscle) in T₂-weighted images, Averaging selected images from the multi-echo image set, the CNR was increased by a factor of nearly three.

In the second study, multi-echo imaging and in-vitro measurements showed progressive changes in the T₂-spectra of frog sciatic nerve undergoing Wallerian degeneration. The two most apparent changes as degeneration progressed were a reduction from three well-resolved T₂ components to one and a decline in the fraction of the spectra associated with short-lived T₂. The former change appears to reflect a collapse of mjelinated fibres, while the latter a combination of interstitial oedema and myelin loss.

The third study found that each of the three T₂ components of peripheral nerve water exhibited unique longitudinal relaxation (T1) and magnetisation transfer characteristics. Simulations demonstrated that mobile water exchange between axonal and myelinic components was not necessary to explain their similar steady-state magnetisation transfer contrast (MTC)s, and reasoning dictated that water exchange cannot be the primary mechanism for this similarity. Rather, the similar MTC of the two shorter-lived T₂ components results from differing intrinsic T_1s. Therefore, interpreting MTC change to solely reflect a change in degree of myelination could lead to erroneous conclusions.

Finally, the fourth study used computer simulations and experimental data to demonstrate that when using sub-optimal spoiler gradients in a multi-echo imaging sequence, increasing the first spoiler gradient slightly reduces the fraction of unwanted signal by several times, resulting in T₂ measurements within 1% of those obtained using optimal spoiler gradients. Use of this spoiler adjustment reduces the peak spoiler gradient requirement by a factor of 2-4.