Sparing nerves of the autonomic nervous system during pelvic surgery is highly desirable to avoid postoperative complications like incontinence or impotence. When combined with augmented reality, individualized preoperative mapping of those nerves could provide a valuable support to guide surgeons during interventions. However, while nerve roots and large bundles of nerves in the pelvis are clearly visible with MR neurography^1,2, it is much more difficult to distinguish small nerves surrounding the rectum, the bladder or the prostate, and to fully delineate them. A previous study carried out on fixed cadaveric pelvises showed that autonomic innervation could be reconstructed thanks to an imaging protocol combining several contrasts³. In this preliminary study, we investigate the value of a 3D high-resolution sequence with a particular contrast to visualize and delineate pelvic autonomic nerves in a healthy volunteer. A young female volunteer (26y) underwent a scan of the pelvis at iCube laboratory (IMIS platform, university of Strasbourg) without any injection. The subject was examined with a 3T MR system (Verio, Siemens Healthcare, Erlangen, Germany) using a 6-channel phased-array flexible coil and a spine coil for signal reception. A dielectric pad was used to prevent standing wave artifact and limit breathing motion in the upper part of the pelvis. MR neurography was performed with a 3D PD SPACE (Sampling Perfection with Application optimized Contrasts using different flip angle Evolution) sequence with the following parameters: FOV=240*160mm², TE/TR=23/519 ms, GRAPPA factor 2, echo train length 43, 2 excitations, in plane resolution 0.77*0.77 mm², 208 1-mm thick axial slices, TA=12min. Nerves were manually segmented in consecutive slices using ITK-SNAP software (Penn Image Computing and Science Laboratory, Department of Radiology, University of Pennsylvania, USA). Particular attention was paid to the smaller pelvic nerves like the obturator and pudendal nerves and the inferior hypogastric plexus. Errors were minimized by navigating through various orientations in the image volume to check continuity of structures and anatomical accuracy. The image quality was scored on a 5-point Likert scale (0=not visible, 1=only the root is visible, 2=root and small parts of the nerve are visible, 3=almost fully visible with uncertainties, 4=visible from the root to the ending) for the nerves of interest, ie the pudendal nerves and the hypogastric plexus that are likely to be damaged during surgery.Figure 1 shows delineation of nerves in the three planes. The annotated final 3D reconstruction is depicted in Figure 2. The sacral plexus and the obturator nerves were easily segmented on consecutive slices. The major branches of the pudendal nerve innervating the inferior part of the rectum, the perineum and the genital tract were segmented, but the position of the proximal part after splitting of the sacral nerves to the end of Alcock’s channel was difficult to locate. This part of the nerve was not represented and it received a quality score of 3. Owing to the complex structure of the hypogastric plexus and the presence of small blood vessels having a similar contrast, segmentation of the smallest branches of the hypogastric plexus was also hard. To avoid any wrong reconstruction, these branches were not represented. As a consequence, the images received a quality score of 3 for the hypogastric plexus. Contrary to most of the studies stating that T2-based or diffusion-based sequences are best suited for MR neurography^1,2,3, we chose here a PD-weighted SPACE sequence with the shortest TE and TR so as to acquire T1-weighted-like images and get the best contrast between pelvic fat and nerves. This feature, as well as high-resolution three-dimensional imaging, was found to be essential in tracing the course of the nerves. The pudendal nerve could not be fully delineated but the branches most likely to be impacted during surgery were successfully reconstructed. We plan to acquire angiographic images to avoid confusion between the hypogastric plexus and small blood vessels.Our method proved to be efficient to delineate nerves of the sacral plexus and the obturator nerves, and offers interesting results for the pudendal nerves and the hypogastric plexus. Even if these preliminary results have to be corroborated by other acquisitions to check applicability to other morphologies, they are promising for the future of nerve-sparing surgery by opening avenues for individualized preoperative maps of autonomic innervation.