Bergen Braun, Andrew Anderson, PhD, Penny Atkins, PhD Hip disease affects many individuals and is detrimental to quality of life. It is generally understood that sex-based differences in hip shape may predispose individuals to certain hip diseases. Specifically, females are more likely to develop acetabular dysplasia, while males are more likely to develop acetabular retroversion and FAI. Further, there is a high prevalence of hip disease related morphology in male and female athletes, especially, acetabular dysplasia and FAI. However, many of these athletes do not have any pain or motion-related symptoms, which indicates that pathological morphology is not well defined. Previously, hip shape and morphology has been evaluated using 2D radiographic measurements, however, the pelvis is a complex shape and cannot be fully represented in 2D. We believe that 3D analysis of the pelvis shape will help us to gain a better insight into shape variation of the pelvis between males and females. Herein, we used 3D statistical shape modeling to determine the variation in male and female pelvis shape for a cohort including patients with hip disease, athletes, and control subjects. The mean male and female pelvic shapes will be used to evaluate the morphological variation of the pelvis and to identify the less obvious anatomical differences between male and female pelvis. Imaging data from fifty-five computed topography (CT) scans and 10 double echo steady state (DESS) magnetic resonance (MR) scans were acquired after obtaining informed consent. The subjects were separated into five cohorts, based on diagnosis or level of activity. These groups included three patient groups based on diagnosis of dysplasia, retroversion or cam FAI, a control group, and a group of collegiate athletes. Each of these groups contained a mix of male and female participants (38 males, 25 females overall). The images were segmented to isolate the pelvic bone based on voxel intensities in the images, using Corview and Amira. Here, all three planes of the images were segmented, and a 3D reconstruction of each pelvis was generated, smoothed, and decimated. The 3D reconstructions of the pelvii were then used to create a statistical shape model using ShapeWorks. ShapeWorks provides methods to automatically place and optimize correspondence particle locations over subject-specific surfaces which provides correspondence across a population of shapes. From the particle locations, the mean shapes were generated for the entire cohort, as well as for all males and females and then for each group. The statistical shape model of the pelvis is still being finalized, but preliminary comparisons between males and females show large shape differences in superolateral iliac crest and posterolateral ischium (Figure 1), which aligns with previously described sex-based differences in pelvic morphology. Results will be finalized within the next few weeks and the analysis expanded to evaluate the five groups.