Inter-system differences in modelling approaches for shoulder joint kinematics between markerless and marker-based...

Area of Science:

• Biomechanics
• Sports Science
• Motion Capture Technology

Background:

• Markerless motion capture systems offer potential advantages for analyzing high-speed movements.
• Limited comparative data exists for upper-extremity kinematics using markerless systems versus traditional marker-based methods.

Purpose of the Study:

• To compare shoulder joint kinematics between a markerless system (Theia3D) and a marker-based system during baseball pitching.
• To identify inter-system modeling differences in shoulder angles and joint center positions.

Main Methods:

• Simultaneous recording of six collegiate pitchers during 117 baseball pitching trials using both markerless and marker-based systems.
• Analysis using nested Bland-Altman analysis, intraclass correlation coefficients (ICC), and statistical parametric mapping (SPM).

Main Results:

• Small differences at maximum external rotation (MER) for external rotation (ICC=0.77), but significant bias in horizontal angle (-19.1°).
• Increased inter-system differences at ball release (BR), particularly for external rotation (ICC=0.33), with proportional bias observed.
• Statistical parametric mapping revealed significant differences during the acceleration phase.

Conclusions:

• Markerless and marker-based systems exhibit notable discrepancies in shoulder kinematics during baseball pitching, especially at ball release.
• Participant-specific offsets contribute to agreement limits, highlighting the need for careful interpretation of markerless data.
• Understanding system-specific characteristics is vital for applying markerless motion capture in overhead athletes.