The Utility of Gait in Forensic Human Identification: An Empirical Investigation Using Biomechanical and Anthropological Principals
conference contributionposted on 27.11.2020, 12:06 by Ioana Macoveciuc
Forensic gait analysis is generally defined as the analysis of walking features of individuals from video footage, to assist in criminal investigations. Although an attractive approach to detecting suspects since data can be collected from a distance without their knowledge (e.g. from public CCTV cameras), this field of study currently lacks validation and quality standards not only due to insufficient research, but also because certain scientific foundations, such as the assumption of gait uniqueness, have not been adequately investigated. To test the scientific basis of this premise, a suitable dataset replicating an ideal forensic gait analysis scenario was compiled from the Karlsruhe Institute of Technology (Germany) database. Biomechanical analyses (i.e. joint angles) of human walking motion (i.e. gait cycles) in the left and right shoulder, elbow, hip, knee, and ankle joints of twenty participants was conducted using the R program. The purpose of these analyses was to investigate the degree to which differences between walking activities of the same participant (i.e. intraindividual variation) impact differences amongst participants (i.e. interindividual variation), according to the following aims: (1) to better understand the relationship between form (anatomy) and function (physiology) of human gait, (2) to investigate the basis of gait uniqueness by examining similarities and differences in joint angles, and (3) to build upon current theoretical foundations of gait-based human identification. The findings indicate different degrees of gait asymmetry given anatomical body region and gait sub-phase (i.e. component of a given cycle), thereby challenging previous methods employing interchangeable use of gait data irrespective of body side, and the use of ‘average’ gait cycles to represent the gait of an individual irrespective of body side. Furthermore, interindividual variability in all five joints is influenced by body side to different extents depending on gait sub-phase and body region, thereby challenging the claim of holistic uniqueness of gait features across all body regions and gait sub-phases. The results therefore underline that previously held assumptions may not hold true, and that their continued use irrespective of innumerable recommendations previously made and in light of novel research, may be detrimental to judicial situations where guilt or innocence is established based on such evidence. Given the findings of this research and paucity regarding empirical basis to support expertise, exerting caution when evaluating gait-based evidence admissibility in the courtroom is highly recommended, since the utility of gait in identification is currently limited.