Investigation of the Sources of Error and Variability of On-line Motion Analysis

Motion Anaylsis;Variablity;Movement;Biomechanics;On-line Motion Analysis

Evans, Justin Noel ; Bartlett, Roger

University of Otago

关键词: Motion Anaylsis;    Variablity;    Movement;    Biomechanics;    On-line Motion Analysis;   

Others  :  https://ourarchive.otago.ac.nz/bitstream/10523/4264/1/EvansJustin2013MPhEd.pdf

美国|英语

来源: Otago University Research Archive

【 摘 要 】

Movement variability is of increasing interest to biomechanists as it has been hypothesised to play a functional role in movement such as facilitating coordination changes, allowing adaptation to environmental changes or reducing the risk of injury. To investigate this variability, online motion analysis provides an essential tool, however data collection can introduce errors which may be mistaken for movement variability. The aim of the study was to investigate whether movement variability could be assessed accurately by discerning it from other sources of variability introduced while using online motion analysis. The project was divided into two parts with the first part partitioning the total variability into; movement variability, inter-session variability and inter- operator variability, while the second study investigated the amount of inter-session variability introduced due to incorrect marker placement.The first study sought to quantify the proportion of all variability that was due to movement variability to determine if it could be accurately discerned from the other sources of variability. A single participant walked on a treadmill at a set speed with kinematics determined using the Plug-in- Gait biomechanical model. Markers were placed by three operators for three sessions each in three different conditions; markers placed by the operator before each session, marker position marked on the participant’s skin after first placement with the marker replaced on the index mark for each subsequent session, and markers left in position for all sessions. The total variability was then partitioned using a repeated measures ANOVA. The second study investigated the amount of variability introduced due to changes in the kinematics of the knee because of incorrect placement of the thigh marker. The kinematics of the knee were calculated for a number of marker placements following a simple movement.When partitioning the variability for the operators separately it was determined the largest contributor to the total variability was due to inter-session variability for all conditions but this decreased for condition 2 and then again for condition 3. When looking at the operators as a group, the largest proportion of the total variability was a result of inter-operator variability for all conditions, followed by inter-session variability. This is most likely due to marker placement differences and that conclusion was supported by the results of the second study where the kinematics were found to be considerably different due to differing marker placements. These results indicate that in studies where movement variability is the focus, it cannot be accurately investigated over a number of sessions when markers are replaced between sessions, either by the same or different operators as the variability introduced from inter-session and inter-operator variability overwhelms that due to movement variability.