Gait analysis has demonstrated to efficaciously support clinical investigations. The patterns of the outcome variables (joint angles, moments and powers) are characterized by an intrinsic and extrinsic variability. Particularly, extrinsic variability is induced by operator-dependent differences in markers' placement, with errors propagating non-linearly to alter outcome variable patterns. The aims of this study are (1) to consider a specific gait analysis protocol named LAMB and provide a description of its procedures, (2) to experimentally assess the between-operator and within-operator variability induced by operator-dependent marking of required anatomical landmarks and (3) to evidence how such inaccuracies propagates to the gait analysis kinematic and kinetic outcome variables. Six expert gait analysis operators performed LAMB anatomical landmarks marking on three healthy adult participants; moreover, one operator repeated three times the marking on one participant. The participants then performed a set of locomotor tasks including stair negotiation and heel- and toe-walking. An anatomical calibration approach let to register each marking and to compute, starting from one single raw data set, a set of outcome variables for each marking/operator. The between-operator variability of gait analysis outcome was assessed in terms of mean absolute variability to quantify offsets and minimal correlation coefficient to quantify patterns' similarity. The results evidence average minimal correlation coefficient ranging from 0.857 for moments to 0.907 for angles and average mean absolute variability accounted for few degrees in angular variables (worst between-operator mean absolute variability is 7.3°), while dynamic variables mean absolute variability, relative to the variable range, was below 5% for moment and below 10% for powers. The variability indexes are comparable to those related to previously published protocols and are independent from the considered task, thus suggesting that the LAMB is a reliable protocol suitable for the analysis of different locomotor tasks.
Keywords: Gait analysis; biomechanical testing/analysis; motion analysis systems; photogrammetry/stereo-photogrammetry; variability.