A Comparison of Marker-Based and AI-Driven Markerless Motion Capture Technologies for Orthopedic Gait Analysis: Accuracy, Limitations, and Clinical Applications

Accurate assessment of human movement is crucial for identifying abnormal gait patterns and evaluating the effectiveness of orthopedic treatments. Meanwhile, advances in artificial intelligence and computer vision have promoted the development of markerless motion capture as an alternative to traditional marker-based systems. This paper primarily compares marker-based and AI-based markerless motion capture in orthopedic gait analysis, including their principles, errors, accuracy, and clinical applications. Marker-based systems have relatively high accuracy and are already well developed, but they can still be affected by soft tissue artifacts. Moreover, this system requires the pre-installation of markers before testing, and the equipment cost is also relatively high. While the markerless system does not need to attach markers to the body, it is therefore more convenient to use. Yet, limitations remain in complex 3D movements, pathological gait, and some dynamic measurements. Based on relevant research, this paper compares the features and application scenarios of the two technologies, and discusses their future development directions and their different roles in high-precision assessment and daily gait screening.