Agreement Between Marker-Based and Markerless Kinematic Methods: Clinical Application in the Assessment of Trunk-Lower Limb Coordination in Runners With Chronic Ankle Instability

• Grant number: 24/10736-9
• Support Opportunities: Scholarships in Brazil - Master
• Start date: August 01, 2025
• End date: February 28, 2026
• Field of knowledge: Health Sciences - Physical Education
• Principal Investigator: Bruno Luiz de Souza Bedo
• Grantee: Dayanne Rodrigues Pereira
• Host Institution: Escola de Educação Física e Esporte (EEFE). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

The use of marker-based human movement analysis systems has been increasingly questioned in terms of their cost and benefits. Their high cost in software, the need for skilled labor, and the restricted use in controlled environments are some of the limitations associated with these systems. Therefore, machine learning-based systems (markerless) may represent a more affordable and lower-operational-cost alternative for human movement analysis. However, most of these tools have been validated in healthy populations, making it difficult to generalize their results and reliability to populations with motor deficits. Thus, the aim of this study is to compare marker-based and markerless kinematic analysis systems in assessing trunk and lower limb coordination in runners with chronic ankle instability. Two simultaneous kinematic analyses will be conducted using eight OptiTrack infrared video cameras for the marker-based system and a markerless system with the OpenCap tool during the Single-Leg Step Down task, involving 42 runners divided into a control group and a chronic ankle instability group. Coupling angles will be calculated and time-normalized. To quantify joint coordination patterns, relative motion diagrams will be constructed, and the classification of joint movement patterns will be performed using the Vector Coding technique. Accuracy between systems will be assessed using the Bland-Altman test, and group data will be compared through unpaired analysis using Statistical non-Parametric Mapping. (AU)