3D magnetic resonance neurography (MRN) enables multiplanar coverage and characterization of the peripheral nerves. 3D Short tau inversion recovery (STIR) or spectral adiabatic inversion recovery (SPAIR) are typically used to suppress fat and improve the conspicuity of the nerves. However, considerable B1 and B0 inhomogeneities in difficult regions such as brachial plexus limit the performance of these techniques, resulting in inhomogeneous fat suppression and reduced SNR, particularly on 3T scanners. Recently introduced Dixon method with 3D TSE achieves homogenous fat suppression with high SNR, often at the expense of scan time^[1,2]. However, the scan time can be made equivalent to 3D STIR or SPAIR by acquiring both echoes in the same repetition using dual-echo Dixon TSE^[3]. The purpose of this work was to compare the performance of 3D dual-echo Dixon TSE against current standard of care 3D STIR and SPAIR in brachial plexus of volunteers and patients at 3T using similar acquisition times.All experiments were performed on a 3T Ingenia scanner (Philips Healthcare, The Netherlands). Dual-echo Dixon TSE was designed to acquire both in-phase and out-of-phase images in the same repetition. 3D STIR and 3D dual-echo Dixon TSE were performed on the brachial plexus of 3 healthy volunteers and 4 patients with IRB approval and written informed consent. 3D SPAIR TSE was performed on 4 normal volunteers, but was not performed subsequently on the patients due to incomplete fat suppression observed in the preliminary studies. The dual-echo Dixon images (Δt=1.1 ms) were acquired using a modulated refocusing flip angle 3D TSE with the following parameters: α_min=20^o, α_center=100^o and α_max=120^o; TR = 2000ms; TE = 107ms; FOV = 320×321×123mm; resolution = 1.4×1.4×1.4mm (reconstructed to isotropic 0.7 mm); SENSE = 2.4 (phase encoding direction) and total scan time = 6 minutes (coronal orientation). 3D STIR and 3D SPAIR were acquired using the same parameters except for a TR of 3000ms to increase the SNR; TE of 68ms; and TI of 240ms. All images were qualitatively evaluated by a musculoskeletal radiologist (AC) with 19 years of experience. The images were scored on a three-point grading scale. Visualization of the nerves, fat suppression: 0 – poor and non-diagnostic; 1 – moderate and non-uniform; and 2 – excellent and uniform. Blood suppression: 0 – unsuppressed and non-diagnostic; 1 – unsuppressed but diagnostic; and 2 – suppressed and diagnostic. These scores were analyzed with a Wilcoxon signed-rank test. Additionally, quantitative signal intensities were measured on the dorsal nerve root ganglion and C6 nerve. The SNR was calculated as the ratio of the mean to the standard deviation and the measurements were compared using a paired Student’s t-test.Figure 1 shows targeted maximum intensity projections (MIP) of a representative volunteer comparing 3D STIR, 3D SPAIR and water-only images from 3D dual-echo Dixon TSE. The conspicuity of the nerves is relatively poor on 3D STIR (fig.1a) due to reduced SNR combined with non-uniform fat suppression due to its sensitivity to both B1 and B0 inhomogeneities. The conspicuity of the nerves is improved in 3D SPAIR (fig.1b), but still suffers from incomplete fat suppression in the areas of increased B0 inhomogeneity (dashed arrow in fig.1b). The 3D dual-echo Dixon TSE (fig.1c) achieves uniform fat suppression increasing the conspicuity of the nerves throughout the volume. The brachial plexus images of a patient with stretch injury of the left upper trunk, axillary and suprascapular nerve are shown in figure 2, where the B1 inhomogeneity caused significant shading artifacts on the right side in 3D STIR (fig.2a). The uniform fat suppression achieved with 3D dual-echo Dixon TSE (fig.2b) provided enhanced visualization of the entire brachial plexus as well as the lesion. The fat suppression, arterial and venous suppression and nerve visualization scores across all subjects (3 normal volunteers and 4 patients) are summarized in figure 3a. The fat suppression and nerve visualization with 3D dual-echo Dixon TSE were significantly higher than 3D STIR (P<0.05), while the arterial and venous suppression were better, but not statistically significant. Figure 3b shows statistically significant SNR increase with 3D dual-echo Dixon TSE compared to 3D STIR (P<0.001).Compared to 3D STIR and SPAIR, 3D dual-echo Dixon TSE sequence provided robust fat suppression with superior SNR and better nerve visualization in similar acquisition times. The modulated refocusing flip angle 3D TSE sequence with low refocusing flip angles has the intrinsic ability to suppress vessels increasing the conspicuity of the nerve visualization.3D dual-echo Dixon TSE sequence can be used for optimal brachial plexus MR Neurography for selective nerve visualization at 3T.