Inter-system reliability of markerless motion capture in older hip and knee patients during walking

This study evaluated the inter-system reliability of multi-camera markerless motion capture (MMC) during walking in older adults with hip or knee pathology and compared inter-system disagreement with a healthy control group. Forty-one participants were included: 15 hip patients, 13 knee patients, and 13 healthy controls. All participants performed walking trials recorded simultaneously by two temporally aligned MMC systems. Inter-system consistency was assessed using waveform-based comparison of joint angle trajectories across the gait cycle and discrete range of motion parameters. Waveform disagreement was quantified using one-dimensional statistical parametric mapping and root-mean-square error, while discrete outcomes were evaluated using intraclass correlation coefficient (ICC), standard error of measurement (SEM), minimal detectable change at the 95% confidence level, and Bland-Altman metrics. Group differences in subject-level inter-system disagreement metrics were tested using linear models. ICC-based inter-system reliability was predominantly high, with 18 of 36 parameters classified as excellent, 10 good, 6 moderate, and 2 poor. SEM values were below $2^{\circ }$ for most parameters. The least robust results were consistently observed for trunk flexion and selected transverse-plane variables. Exploratory group comparisons did not reveal a consistent directional pattern of increased disagreement in the pathological groups, and clinically plausible group differences were similarly reflected by both systems. However, the modest group sizes limit the ability to exclude smaller pathology-related effects on measurement consistency. These findings indicate small within-platform inter-system disagreement in most lower-limb outcomes for gait analysis in older adults with hip or knee pathology. Trunk and some transverse-plane outputs should be interpreted with greater caution.