Background/Purpose Protrusio acetabuli and acetabular dysplasia are biomechanical antipodes with distinct risk profiles for chondrolabral damage and OA progression. We asked whether volumetric, AI-assisted dGEMRIC can detect topographic differences in acetabular cartilage composition that are clinically relevant for PAO planning, rim work, and counseling versus early THA. Methods Retrospective cohort (2020–2024) of 55 hips (11 protrusio; 44 dysplasia) with standardized 3T direct MR arthrography. A validated nnU-Net pipeline generated patient-specific 3D acetabular cartilage models and sector-wise dGEMRIC (T1, ms) using a clock-face (1–12 o'clock; incisura-adjacent sectors excluded) and central/peripheralzoning (2:1). Groups were defined radiographically (protrusio sign; LCE < 23°). Statistics compared global and regional composition with multiplicity control. Results Global dGEMRIC did not differ (protrusio 595 ± 129 ms vs dysplasia 576 ± 89 ms; p = 0.70). However, dysplasia showed focal biochemical vulnerability with significantly lower dGEMRIC at the anterior and posterior horns (central 4–5 and 7–8 o'clock; peripheral 4–5 and 7–9 o'clock; Δ≍ 80–200 ms; p≤0.04). Both morphologies peaked superiorly (11–1 o'clock). Protrusio displayed biochemically more uniform acetabular cartilage without significant intra-acetabular dispersion. Discussion Although protrusio and dysplasia represent opposite morphologies, overall cartilage biochemistry was largely similar; only the anterior and posterior acetabular horns showed lower dGEMRIC values in dysplastic hips. The topographic pattern was identical in both groups (highest superiorly, lowest at the horns), indicating that cartilage quality is mainly intra-individual load-dependent rather than morphology-dependent. These results offer a solid baseline for future longitudinal studies and for using 3D dGEMRIC as a potential prognostic marker after reorientation surgery or rim trimming.
Domenech et al. (Thu,) studied this question.