Magnetic resonance neurography can be performed on 1.5 and 3.0 Tesla clinical scanners. When available, 3.0 Tesla is preferred due to the increased signal-to-noise ratio and superior contrast-to-noise ratio at higher spatial resolution. Especially 3-dimensional (3D) isotropic imaging benefits from higher field strength and shorter acquisition times. With high-end field strength, some drawbacks may arise such as increased B0 inhomogeneities that might cause poor fat suppression when spectral saturation pulses are used. Therefore, the use of STIR (short tau inversion recovery) or Dixon type fat suppression may be beneficial at 3.0 Tesla [17]. Another concern at 3.0 Tesla is the increased specific absorption rate, which might cause longer examination times. Of course, technical tricks exist to address these latter limitations such as reducing the refocusing flip angle to shortened acquisition time (e.g. 130° instead of 150°). However, despite some drawbacks, most experts favor the use of 3.0 Tesla scanners.

Whenever possible, the use dedicated multichannel surface coils is advised. These should be—as always in magnetic resonance imaging—be placed as close as possible to the anatomy of interest. While seemingly simple, this can be quite challenging when imaging several anatomic regions needing different coils or if a larger field-of-view needs to be covered. Some scanners allow combined use of multiple coils, whereas with other scanners, coils need to be changed during the study.. Dedicated joint coils can also be used in combination with flex or surface coils in some scanners which, for example, allows the use of high-resolution wrist coils to examine the median nerve in the wrist in combination with a flex coil that covers the forearm. This combination allows seamless examination of the median nerve in the forearm, wrist and hand at the same time. If a specific field-of-view is still too large to be covered even with the use of coil combination, then a stepwise-approach, first using nerve ultrasound to narrow down the possible site of injury and with subsequent dedicated magnetic resonance imaging of that remaining nerve section may be helpful. Along these lines, it is important that referring physicians understand the tradeoffs between spatial resolution and anatomic coverage in magnetic resonance neurography. It is almost always preferable to scan a small region in high detail than to try to survey a large amount of anatomy for possible nerve pathology.

Training of the technician/technologist is an often forgotten but nevertheless extremely important issue. Technicians should also be taught to identify target anatomic structures (they should be able to identify the nerve-of-interest); to correctly position patients (e.g. the region-of-interest should be placed in the scanners isocenter whenever possible); and to correctly plan pulse sequences (e.g. to avoid wraparound artifacts).

Much has been written about pulse sequence developments for magnetic resonance neurography and several new techniques have been published including variations of 3D FSE (fast spin echo) acquisitions. The latter are based on technical advancements and are available as product sequences (e.g. Siemens; SPACE; Philips: VISTA, Views; and GE: Cube) [4]. The basic principle of the sequences when used and optimized for magnetic resonance neurography is to reduce the refocusing flip angles to decrease the sensitivity to flow. When assessing images, it is obvious that these sequences adequately suppress the arterial signal, however, venous signal contamination remains a frequent problem due to slow flow. Hence, these 3D FSE sequences are often combined with either a weak gradient pulse to reduce venous signal or with motion sensitizing driven equilibrium (MSDE) for further reduction of the venous flow signal. For example, SHINKEI (nerve-SHeath signal increased with INKed rest-tissue RARE Imaging) is a 3D FSE sequence variant with SPAIR for fat suppression and MSDE for blood suppression [18]. These 3D sequences are especially helpful for brachial and/or lumbosacral plexus imaging [19].