Repeatability of an optimised lower body model

The optimisation technique, optimised lower-limb gait analysis (OLGA), is described together with a preliminary study of repeatability compared to an implementation of the Newington-Helen Hayes gait model. The study of repeatability used a single healthy subject, three physiotherapists as observers and provided approximately 100 gait cycles. Improvement in intra- and inter-observer repeatability of the lower limb model was found for OLGA, indicated by significantly lower standard deviations (S.D.s) in local marker co-ordinate (a measure of rigidity of the marker attachment), together with reduced S.D. in the estimated length of the bone segments. The S.D. in the inter-hip distance measured by OLGA (N = 25) was found to be only 2.4 mm. The repeatability of clinically significant output variables (joint angles, forces and moments) was also improved, with the inter-observer variations for joint angles and forces being significantly lower for OLGA. Euler angle component cross-talk effects frequently reported at the hip, knee and ankle were also successfully reduced by OLGA, this being the chief cause of the improvement in inter-observer repeatability. (C) 2003 Published by Elsevier B.V.