New Research on Theia3D: Comparing Markerless and Marker-Based Shoulder Kinematics Across Full Arm Range of Motion

Summary

A 2025 study published in the Journal of Biomechanics compared shoulder kinematics measured with Theia3D markerless motion capture against gold-standard marker-based motion capture across full arm ranges of motion. The results show strong agreement for many movements, particularly in the coronal plane, while also highlighting where caution is still required, especially at the extremes of ranges of motion and in transverse-plane rotations.

Why This Matters

Until recently, most validation work for Theia3D focused on gait and lower-limb biomechanics. This study addresses a major gap by rigorously evaluating upper-limb and shoulder kinematics across full, isolated ranges of motion, including overhead tasks that are especially challenging to capture reliably.
‍

Study Overview

Participants
‍

20 healthy adults (18–30 years)
10 male, 10 female
40 shoulders analyzed‍

Experimental Setup
‍

Participants performed three repetitions of each bilateral movement while data were collected simultaneously using:
‍

Markerless: Qualisys Miqus cameras (100 Hz) processed with Theia3D Apollo v2024.1.24‍
Marker-based: Qualisys Arqus cameras (200 Hz), using ISB-recommended marker placement

*Fig. 2. Shoulder tasks. Scaption (A), abduction (B), flexion (C), internal rotation (D), external rotation (E), and internal and external rotation at 90° abduction (F).*

Tasks Analyzed
‍

Full arm scaption
Abduction
Flexion
Internal rotation
External rotation
Internal and external rotation at 90° abduction

Analysis Methods
‍

Coefficients of Multiple Correlation (CMC) for waveform similarity
Root Mean Square Differences (RMSD)
Bland–Altman limits of agreement
Correlations of ROM measures
Minimal Clinically Important Differences (MCID) to assess clinical relevance
‍

Key Findings

Strong Agreement in the Coronal Plane
‍

Abduction angles showed the smallest RMSDs (<6°, within the MCID) and excellent CMCs across most tasks
Mean ROM differences in the coronal plane were generally within MCID thresholds
This indicates strong suitability for evaluating shoulder abduction movements
‍

Moderate Agreement in the Sagittal Plane
‍

Flexion waveforms were broadly similar between systems
ROM differences were statistically significant for several tasks, but within clinically meaningful limits
Correlations between systems were strong to very strong
‍

Larger Discrepancies in the Transverse Plane
‍

External and internal rotation angles showed the largest RMSDs (≈15–22°)
Markerless ROM values tended to be overestimated compared to marker-based measurements
Differences were most pronounced at the extremes of motion
Despite this, waveform similarity (CMC) was often still good to excellent
‍

Differences Increase at Extreme Ranges of Motion
‍

Across planes, the largest disagreements occurred near maximal joint excursions
This aligns with known challenges in both marker-based and markerless shoulder modeling
‍

What This Means for Research Teams

Where Theia3D Performs Well
‍

Evaluating shoulder abduction and coronal-plane kinematics
Monitoring movement patterns and trends over time
Applications where repeatability and workflow efficiency matter more than precise transverse-plane rotation values
‍

Where Caution Is Still Needed
‍

Transverse-plane rotations (internal/external rotation), especially at end-range
Clinical decisions that rely on small absolute differences in rotational ROM
Interpretation of extreme joint positions without complementary assessment methods

Overall, the authors conclude that Theia3D provides promising shoulder kinematic measurements, particularly for coronal-plane analysis, while emphasizing that further improvements are needed before full clinical equivalence with marker-based systems can be assumed across all planes.

Read the full study here.
‍

Interested in Markerless Shoulder Analysis?

If you’re exploring markerless motion capture for shoulder biomechanics, movement analysis, or research applications, contact us today to see how Theia3D fits into your workflow.

Recent Posts

Baseball Tracking System: A Practical Guide to Evaluating Your Options

Baseball tracking systems can measure everything from ball flight and bat path to workload and full-body biomechanics, but each category answers a different question. This guide compares leading options and explains how to choose the right system based on what your program needs to measure, where it will operate, and whether it can connect athlete mechanics to baseball outcomes.

How to Choose Motion Analysis Software: Features, Accuracy, and Workflow Fit

Choosing motion analysis software depends on more than whether it can track movement. This guide compares leading tools by output type, setup complexity, validation, data ownership, and workflow fit so research and performance teams can choose the right system for their needs.

The Gold Standard for Motion Capture Only Works Inside a Lab. Here's What Comes Next.

Traditional marker-based motion capture has long been considered the gold standard for biomechanics, but its reliance on lab conditions, markers, and lengthy setup limits where and how movement can be studied. This article explores how markerless motion capture systems like Theia3D make it possible to collect accurate, repeatable movement data in real-world research and performance environments.