EP1.40 Defining the Spectrum of Acetabular Dysplasia: A 3D Clockface Classification System in a Large Periacetabular Osteotomy Cohort

Abstract

Abstract Introduction Traditional classification of acetabular dysplasia in mature hips relies on plain radiographs and the lateral center edge angle (LCEA), which fail to capture the full complexity of the femoral head-acetabulum relationship. While CT-based measurements and early 3D analyses have improved morphological understanding, existing classification systems remain limited by small patient cohorts and incomplete characterization of acetabular deficiency patterns. This study aims to define distinct 3D morphological subtypes of acetabular dysplasia in a large cohort of 245 patients undergoing periacetabular osteotomy (PAO), to improve diagnostic accuracy and guide surgical planning. Methods This retrospective study reviewed 245 symptomatic female patients with acetabular dysplasia who underwent PAO, compared to a normative cohort of 961 healthy females aged 18–25 years identified from a hospital radiology database using a validated natural language processing pipeline. Preoperative CT scans were processed using a hierarchical UNet-based convolutional neural network to segment the pelvis and femur, generate three-dimensional surface mesh models, and align them to a standardized pelvic coordinate system. An automated clockface method measured center-edge angles at seven positions (9 to 3 o'clock) around the acetabulum. The classification was based on the specific regions: anterior, from 1, 2, or 3 o’clock; posterior, from 9, 10, or 11 o’clock; and superior, at the 12 o’clock position. Results Among 245 female patients with acetabular dysplasia undergoing PAO, eight distinct morphological subtypes were identified using the 3D clockface classification system. Global deficiency was by far the most prevalent subtype, accounting for more than half of all cases (N=138, 56%), followed by posterior deficiency (N=32, 13%) and posterior superior deficiency (N=26, 11%). Anterior superior deficiency represented 8% of cases (N=19), while anterior-only and superior-only deficiencies were less common (4% and 2%, respectively). Notably, 6% of patients (N=15) demonstrated no measurable 3D deficiency despite clinical recommendation of PAO for hip instability. Conclusion This large-cohort 3D analysis reveals that acetabular dysplasia encompasses six distinct morphological subtypes, with global deficiency being the most prevalent. The 3D clockface classification system provides a more comprehensive characterization of acetabular deficiency than traditional radiograph-based methods, with important implications for surgical planning and individualized PAO correction.