Reliability of Markerless 2D Video Analysis for Calculating Gait Joint Angles in Uncontrolled Environments

Background: Traditional 3D gait analysis constrains natural movement variability due to artificial laboratory settings; thus, this study evaluated the intra- and inter-rater reliability of a low-cost, 2D markerless smartphone video protocol for assessing lower limb sagittal kinematics in an unconstrained environment. Methods: Forty-five healthy young adults walked at a self-selected speed in a non-laboratory corridor, where sagittal hip, knee, and ankle angles were recorded at 60 fps and analyzed using Kinovea^® software by two independent raters, with one rater repeating measurements seven days later. Results: Inter-rater reliability ranged from moderate to good across most joint variables (ICC = 0.628-0.789), except for the ankle during loading response, which showed poor consistency (ICC = 0.381). Between-session intra-rater reliability was generally lower, ranging from poor to moderate (ICC = 0.278-0.580), with the lowest values observed at the hip (ICC = 0.278) and ankle (ICC = 0.391) during loading response. Minimal Detectable Change (MDC95) values ranged from 7.07° to 8.95° for inter-rater assessments and from 7.40° to 16.55° for intra-rater assessments, with Bland-Altman analysis displaying wide limits of agreement longitudinally. Conclusions: In conclusion, while prior observer training standardizes criteria for immediate cross-sectional evaluations, the protocol's high sensitivity to measurement variability and unconstrained motor execution limits its clinical application for longitudinal tracking across different sessions. Between-session intra-rater variability reflects evaluator subjectivity in manual frame selection and landmark digitization over time when no physical markers are present, rather than biological gait fluctuations.