If clinical and/or EDX studies cannot precisely localize a lesion, it is recommended to initially focus imaging at the most symptomatic site or the area of previous injury or surgery. Skin markers should be placed over palpable abnormalities and/or surgical scars, when present, to assist in anatomic localization. When feasible, real-time supervision by an experienced MRN radiologist facilitates protocol adjustments and coil repositioning to maximize the diagnostic yield.

MRN is ideally performed at 3.0 Tesla (T) due to its superior gradient performance and nearly double signal-to-noise ratio (SNR) compared to 1.5 T. To optimize SNR and spatial resolution, use of a flexible phased-array surface coil that closely conforms to anatomic contours is preferred, as this reduces the distance between the coil elements and the region of interest. MRN relies on heavily T2-weighted (T2w), fat-suppressed sequences—both 2-dimensional (2D) and 3-dimensional (3D)—which typically provide the necessary contrast to delineate the nerve from background tissues and sufficient fluid-sensitivity to maximize sensitivity to nerve pathology. Such 2D sequences provide high in-plane resolution (typically 0.3–0.5 mm), and when acquired orthogonal to the longitudinal course of the nerve, facilitate detailed assessment of the nerve’s fascicular architecture. 3D MRN commonly includes either fat-suppressed fast spin echo (FSE) (including motion-sensitive driven-equilibrium based techniques such as 3D CRANI ), or steady-state free-precession based sequences, such as time-reversed fast imaging with steady-state precession (PSIF) and double-echo steady-state (DESS) sequences. When acquired isotropically, 3D MRN is particularly effective for long-axis visualization of peripheral nerves to detect subtle morphologic changes and deviations in their course. An MRN examination also typically includes either a 2D proton-density weighted (PDw) FSE sequence for anatomic delineation, acquired axially and sometimes with additional planes to provide higher in-plane resolution imaging than T2w sequences for evaluating the surrounding anatomy and fascicular architecture.

Recommended MRN protocols for the lumbosacral plexus and lower extremity are summarized in Tables 1 and 2 , respectively. ^, Optionally, intravenous gadolinium administration can facilitate vascular signal suppression, improving the visualization of small nerve branches that may be obscured by adjacent isointense vessels. Susceptibility artifacts from metallic hardware can be mitigated using multispectral metal artifact reduction techniques, such as multiple-acquisition variable-resonance image combination (MAVRIC, GE HealthCare, Waukesha, WI) and slice encoding for metal artifact correction (SEMAC, Siemens Healthineers, Erlangen, Germany). However, these are often associated with trade-offs in reduced in-plane spatial resolution or increased scan time. In some cases, severe susceptibility artifact may necessitate imaging at 1.5 T rather than 3T, when the nerves of interest are near the hardware.

Fig. 1 provides an illustrative overview of the key lower extremity nerves, their normal MRN appearances, and anatomic courses.

Overview of key lower extremity nerves: Normal MRN appearance and anatomic course.

The lumbosacral plexus (LSP) comprises the lumbar (T12–L4) and sacral (L4–S4) plexi and its anatomy has been comprehensively reviewed in a recent publication. The optimal coil configuration typically involves combining a posterior spine array coil with a flexible anterior surface array coil wrapped around the lumbar and pelvic regions. Coronal imaging usually uses a wide field-of-view (FOV) of 29 to 32 cm, spanning from the L3 superior endplate to the lesser trochanters. The small-caliber nerves arising from the L1 and L2 roots (iliohypogastric, ilioinguinal, and particularly genitofemoral nerves) may be difficult to reliably visualize on both steady-state free-precession sequences (due to motion) and short tau inversion recovery (STIR) FSE sequences (due to T2 blurring). For protocol efficiency, the anteroposterior scanning field can be tailored to the specific nerves of interest. For sciatic nerve imaging, posterior coverage should include the greater sciatic foramen to visualize the nerve as it exits the pelvis. In contrast, assessing the femoral and lateral femoral cutaneous nerves requires inclusion of the subcutaneous fat anterior to the anterosuperior iliac spine to fully capture their course along the anterior pelvis and thigh. For a targeted assessment, an oblique-sagittal T2w Dixon scan with a smaller FOV (∼20 cm)—plotted orthogonally to the nerve—enhances visualization of fascicular architecture.

A high-density flexible surface coil, typically 16 to 23 channels, is essential for optimizing SNR and maintaining high spatial resolution across multiple stations in the thigh and lower leg. This is preferred over a combination of several flexible coils that typically results in poorer conformity to the skin and reduced density of elements covering a particular region, thereby resulting in lower SNR. Key nerves relevant to MRN of the thigh and lower leg include the sciatic nerve and its branches—the tibial and common peroneal nerves—as well as the sural and saphenous nerves.

To image the ankle and foot nerves, a dedicated ankle coil or multipurpose high-density flexible array can be used; the tradeoff of the latter option is that the ankle cannot be placed in full dorsiflexion to obtain standard anatomic imaging planes. For small-caliber lower extremity nerves, such as the saphenous and deep peroneal—where interpretation may be hindered by adjacent vessels of similar size and signal intensity—the inclusion of fat-suppressed 3D gradient-echo techniques (eg, PSIF and DESS) are especially helpful as they provide simultaneous vascular suppression, high through-plane spatial resolution, and when acquired isotropically, or nearly isotropically, afford the ability to create multiplanar reconstructions to evaluate the nerve’s course in multiple arbitrary planes.

Deep learning-based reconstruction methods have been shown to enhance nerve conspicuity, and have made traditionally time-consuming 3D high-resolution imaging more clinically feasible. Bone-sensitive MR imaging sequences, such as gradient-echo or zero echo time imaging, are valuable for assessing adjacent osseous structures—particularly when CT has not been performed—by aiding in fracture assessment or preexisting osseous anomalies, such as osteochondromas, in relation to regional nerves ^,

While not routinely used in practice, quantitative T2 mapping shows promise in assessing active muscle denervation, correlating with motor unit recruitment loss observed on EMG. Recent sequence innovations have enabled simultaneous qualitative assessment and quantitative T2 mapping, providing image quality comparable to conventional T2w imaging while facilitating the detection of abnormal T2 values in nerves and muscles, as demonstrated in patients with foot drop. Additionally, T1w Dixon-based fat quantification can be used to assess fat infiltration in chronically denervated muscles.