Concurrent assessment of sprint running lower-limb and trunk kinematics using marker-based and markerless motion capture.

Reliable and accurate sprint kinematics assessment is of utmost interest for both performance and prevention purposes. To guide field practice, new technologies must be tested against optical motion capture references. We compared markerless motion capture (ML)(Theia Markerless) to the reference method, marker-based motion capture (MB)(CORTEX). Three-dimension kinematic data from MB and ML systems and ground reaction force were collected simultaneously for 14 participants, at three stages of the sprint acceleration spectrum (early, middle acceleration and top speed). Sagittal plane angles for the foot, ankle, shank, knee, thigh, hip, pelvis and trunk were computed through biomechanical modelling. Analysis on the whole curves showed reasonable reliability for all sprint running phases for all angles and joints except hip, pelvis and foot at top speed. Discrete points analysis identified greater discrepancies at top speed touchdown. These contrasting results do not allow for clear conclusions about the ability of ML technology to reliably quantify sprint kinematics in the sagittal plane, especially regarding the pelvis. However, due to some limitations of both methodologies (soft tissue vibration (MB) and joint center position accuracy (ML)), it is not clear whether these inter-system differences are due to inaccuracies of one system or the other, or a combination of both.