Is Markerless Motion Analysis a Valid Tool to Study Knee Joint Kinematics?

Description

This presentation from the ISTA 2024 conference held in Nashville, discusses advancements in motion capture technology for clinical gait analysis, focusing on comparing markerless systems that utilize deep learning with traditional marker-based systems. The speaker emphasizes the cumbersome nature of the conventional method, which relies on reflective markers and expensive equipment, and introduces markerless technology as a viable alternative. Key findings indicate significant discrepancies in kinematic measurements between the two systems, particularly in knee joint angles and center positions. These differences are attributed to misalignments in joint axes, prompting a research question on how closely the two systems' outputs align when accounting for these misalignments. A new method called 'Reframe' is introduced to optimize and realign the reference frames of the markerless system, leading to improved agreement between measurements post-analysis. While the results demonstrate that the underlying motion is consistent across systems, the study concludes with a note on the importance of assessing the accuracy of markerless technology and suggests future research directions for expanding its application in analyzing other joints.

Specialties

Conferences