ChildBone: A novel digital twin to assess bone strength in children

When a child attends hospital with unexplained fractures, doctors must rule out all diseases that weaken the bones, making them more likely to fracture, before considering child abuse. Incorrect diagnoses risk the child being taken away from loving carers, while missed diagnoses risk returning the child to an abusive environment, where abuse may continue and perhaps lead to death. However, conditions such as isolated low vitamin D (VD) in children might lead to weaker bones without visible pathology on X-rays and nearly 40% of children below 2 years of age have low VD. The debate among doctors (and sometimes in courts) is whether isolated low VD (i.e., Low VD with normal X-rays) is a cause of fracture. This is difficult to prove or disprove in live children. There is no external gold standard to reliably quantify bone strength in children that could accurately predict the risk of fracture. However, sub-optimal VD level has been used in court cases as a cause of fracture in children with otherwise unexplained injury and researchers have been challenged to replicate and validate the traumatic nature of contentious fracture injuries. We believe that by using engineering technology and Digital Twins (DTs, computer models that replicate an object such as the femur that contain personalised shape and material properties), the effects of isolated low VD on bone strength (i.e., failure load of the bone) can be demonstrated. DTs have been successfully developed and validated for adults, but there is no such validated model for children due to the scarcity of child-specific medical imaging and experimental data to inform these models. This project aims to address these challenges by developing child-specific DTs of the femora and ribs across a wide age range (0-16 years) of children and young people with both sufficient and sub-optimal VD. The current scarcity of data will be addressed by prospective recruitment and collection of paired Computed Tomography and Magnetic Resonance Imaging scans of children to capture the shapes of both the mineralised bone and the non-mineralised cartilaginous region (including the growth plate). Mechanical testing (nanoindentation and in situ mechanical testing/digital volume correlation) of rib samples will also be carried out to evaluate personal-specific material properties and validate the DTs. The personalised DTs will be used to simulate different loading conditions to obtain bone strength to evaluate the effect of sub-optimal VD levels. Correlations will be explored between in vivo (i.e., anatomical features, blood serum VD level) and ex vivo (i.e., nanoindentation modulus, histology score) biomarkers and the predicted bone strength to identify measurements that can be used to speed up diagnosis between fragility and injury fractures. We will create an open-source database containing 3D geometries, material properties, finite element models and predicted bone strength derived from this study to share with the research community and the public. This database can be used in combination with low-radiation EOS® and machine learning methods in the future for faster diagnosis in a larger clinical study. The project aims to lower the uncertainty associated with the diagnosis of unexplained fracture in children, particularly differentiating child abuse from metabolic bone disease. This work will shape our understanding of juvenile bone properties, developmental biology, multiscale modelling in children and will have impact on medico-socio-legal fields for more robust safeguarding of our children.