Video analysis of ACL injuries in sports

Abstract

A complete understanding of the causes for injury requires a description of the mechanism of injury. Video analysis is the only method where we can objectively describe the motion patterns that are involved in a real injury situation. A new model-based image-matching (MBIM) method for 3D motion reconstruction from uncalibrated camera sequences was developed and validated against a marker-based 3D motion analysis system for running and side-step cutting. Flexion/extension joint angles agreed well in both hip and knee, but some underestimation from the video analysis was seen in knee flexion and hip abduction. The internal/external rotation angles varied the most. Next, the model was successfully applied on real injury situations outside the lab. A four-camera basketball video, a three-camera handball video and a one-camera video of an alpine skier was analysed using the MBIM method. Although it was demonstrated that important information for understanding the mechanism of injury could be extracted, some limitations exist. Poor picture quality, occluded body parts (due to clothing or other players), lack of visible landmarks or having only one camera angle may give sub-optimal matchings. It is therefore crucial to obtain high-resolution video footage, as this will open for improved understanding of injury mechanisms.