Quantitative video analysis of head acceleration events: a review

Abstract

The biomechanics of head acceleration events (HAEs) in sport have received increasing attention due to growing concern over concussion and long-term neurodegenerative disease risk. While wearable sensors, such as instrumented mouthguards (iMGs), are now commonly used to measure HAEs, these devices face well-documented challenges, including poor skull coupling, limited compliance, and high false-positive rates. Video footage is routinely collected in sports for performance analysis, and is a perhaps underutilised source for both retrospective and in situ measurement surrounding HAEs. Traditionally used to confirm HAE exposure in wearable sensor studies, video has more recently been explored as a quantitative tool in its own right. This review synthesises the current state of the art in video-based measurement of HAEs, with a particular focus on videogrammetric methods, including manual point tracking and model-based image matching. Recent advances in computer vision and deep learning that offer the potential to automate and extend these approaches are also examined. Key limitations of current video-based methods are discussed, alongside opportunities to improve their scalability, accuracy, and biomechanical insight. By consolidating evidence across traditional and emerging approaches, this review highlights the potential of video as a practical and valuable measurement source for quantitative measurement and modelling of HAEs in sport.