Artificial bone models (ABMs) are commonly used in traumatology and orthopedics for training, education, research and development purposes. The aim of this study was to develop the first evidence-based generic Asian pelvic bone model and compare it to an existing pelvic model. A hundred clinical CT scans of intact adult pelvises (54.8±16.4 years, 161.3±8.3 cm) were acquired. They represented evenly distributed female and male patients of Malay (n=33), Chinese (n=34) and Indian (n=33) descent. The CTs were segmented and defined landmarks were placed. By this means, 100 individual three-dimensional models were calculated using thin plate spline transformation. Following, three statistical mean pelvic models (male, female, unisex) were generated. Anatomical variations were analyzed using principal component analysis (PCA). To quantify length variations, the distances between the anterior superior iliac spines (ASIS), the anterior inferior iliac spines (AIIS), the promontory and symphysis (conjugate vera) as well as the ischial spines (diameter transversa) were measured for the three mean models and the existing ABM. PCA demonstrated large variability regarding pelvic surface and size. Principal component one (PC 1) contributed to 24% of the total anatomical variation and predominantly displayed a size variation pattern. PC 2 (17.7% of variation) mainly exhibited anatomical variations originating from differences in shape. Female and male models were similar in ASIS (225±20 mm; 227±13 mm) and AIIS (185±11 mm; 187±10 mm), whereas differed in conjugate vera (116±10 mm; 105±10 mm) and diameter transversa (105±7 mm; 88±8 mm). Comparing the Asian unisex model to the existing ABM, the external pelvic measurements ASIS (22.6 cm; 27.5 cm) and AIIS (186 mm; 209 mm) differed notably. Conjugate vera (111 mm; 105 mm) and diameter transversa (97 mm; 95 mm) were similar in both models. Low variability of mean distances (3.78±1.7 mm) was found beyond a sample number of 30 CTs. Our analysis revealed notable anatomical variations regarding size dominating over shape and gender-specific variability. Dimensions of the generated mean models were comparatively smaller compared to the existing ABM. This highlights the necessity for generation of Asian ABMs by evidence-based modeling techniques.
