Early quantification of hip-related abnormalities in children using statistical shape models

Osteoarthritis (OA) is one of the most common musculoskeletal disorder in adults, but several hip-related pathologies that increase the risk of OA appear already during growth (from early childhood to young adulthood).

However, there are no well-defined thresholds to diagnose hip-related deformities during growth.This proposal develops a novel approach to detect hip abnormalities during growth.

The hypothesis is that a combination of 1) accurate anatomical definition of the hip in 3D; 2) subject-specific predictions of the stress state of the hip cartilage, and 3) of the overall mechanical competence of the bones, can lead to a much-improved identification of hip developmental diseases and of the long-term risk for OA.

This assessment is clinically feasible thanks to a statistical shape and appearance model of the hip.

The model works as an atlas of normal hip anatomy in 3D as a function of age and sex and allows to obtain 3D subject-specific models of pelvis, femur, and the articular cartilage from one 2D x-ray image.

An extensive clinical assessment on different cohorts of children is performed to verify the diagnostic ability of the approach.

The long-term risk of OA is investigated longitudinally using data from registers to identify new anatomical and functional cues linked to OA.A successful project may improve diagnosis of hip-related deformities during growth and open new ways for long-term OA prevention, resulting in huge socioeconomic benefits.