AUTOMATED VIRTUAL REDUCTION OF DISPLACED DISTAL RADIUS FRACTURES

A distal radius fracture is one of the most common fractures treated in emergency departments. Pre-operative 3D virtual reduction of distal radius fractures is key as it improves the recovery process and reduces complications. Unfortunately, most state-of-the-art virtual reduction methods are either interactive or semi-automatic, which are labour intensive and suffer from high inter-observer variability. In this work, we propose a fully automated algorithm for the reduction of displaced fractures that is robust to both fracture segmentation errors and large fragment displacements. A strategy is proposed to adaptively select landmarks for fragment replacement and iteratively moving the fragments towards a template extracted from the contralateral radius. Real datasets of distal radius fractures with fragment displacements were used to validate our method. Results show that, in terms of distance to the mirrored contralateral radius, the automated reductions closely resemble the manual reductions by the surgeon.