Estimation of hip joint center from the external body shape: a preliminary study

International audienceHuman external shape model is becoming easily available using scanning technology, stereophotogram- metry, or Microsoft Kinect systems. However locating the internal skeleton from the external shape remains a challenging issue. Using low dose biplanar radiographs of the whole body, subject-specific 3D reconstructions of the bones and the surface envelope have been developed, providing data of both external body shape and internal skeleton (Dubousset et al. 2010). The basic idea of this research project was to explore the relationships between characteristics of internal skeleton and those of external body shape. Hip joint center (HJC) prediction was considered in this preliminary study. Subject-specific HJC prediction is cur- rently performed either using functional methods based on the relative motion of the femur and pelvis, or predictive methods relying on empirical regression equations using palpable femoral and pelvic landmarks as predictors. Yet functional methods may be ineffective when hip motion is limited. Different regressions between HJC and predictive landmarks were established based on direct measurements on pelvic and femoral bone specimen surface, using cadav- ers, medical imaging or 3D CT-scans (Peng et al. 2015). However, in most applications when medical imaging is not available, manual palpation over the skin of the bony predictors might increase the prediction error on HJC compared to cases where bone information is available directly (Sholukha et al. 2011, Sangeux et al. 2014). In this paper, new predictors from external body shape were explored for HJC prediction