Inter-Session Reliability of Markerless Motion Capture for Return-to-Sport Screening Tasks

November 26, 2025
Theia
BIO
Recent Posts
How Northwestern University Cuts Data-Collection Time by One-Third Using Markerless Motion Capture
Athletic Movement Performance (AMP) Lab, directed by Dr. Yuki A. Sugimoto in the Department of Physical Therapy and Human Movement Sciences (PTHMS) at Northwestern University’s Feinberg School of Medicine, is advancing ACL-injury and chronic ankle instability (CAI) research by integrating Theia3D markerless motion capture to analyze high-risk movements – including jumping, landing, sprinting, and cutting – in adolescent and collegiate athletes. With Theia3D, the AMP Lab has streamlined key components of real-world biomechanics data collection, completing sessions approximately one-third faster than before. This increased efficiency allows the team to evaluate more athletes, run more trials, and accelerate investigations into lower-limb injury mechanisms that directly inform sport-performance and sports-medicine decision-making.
A Decade of Sports Validation: Which Motion Capture Tech Performs Best?
A 2025 Sensors review from Auburn University compares how optical, IMU, and markerless motion capture systems perform across real-world sports environments. The findings reveal that while optical systems remain industry standard, markerless motion capture is quickly closing the gap, offering unmatched scalability, flexibility, and ecological validity for multi-sport organizations.
Concurrent Assessment of Sprint Running Kinematics Using Marker-Based and Markerless Motion Capture
A July 2025 study published in the Journal of Sports Sciences evaluated the reliability of Theia3D’s markerless motion capture system for assessing lower-limb and trunk kinematics during sprint running. Researchers compared Theia3D against a marker-based optical reference system (Motion Analysis Corp. Raptor) across multiple sprint phases, including early acceleration, mid-acceleration, and top speed. Results showed good to excellent agreement between systems for most joints and movement phases, with some variability at top speed, particularly in pelvic and foot angles. The findings highlight both the promise and the challenges of measuring sprint mechanics with markerless motion capture in high-speed, real-world conditions.

Summary

A 2025 study from KU Leuven, Liverpool John Moores University, and collaborators evaluated whether Theia3D can reliably measure lower-extremity joint kinematics and kinetics across two separate testing sessions during common return-to-sport (RTS) tasks.

Why This Matters


Return-to-sport decisions increasingly rely on movement-quality analysis, not just performance metrics (e.g. hop distance or T-test time). Traditional marker-based motion capture offers high accuracy but requires considerable setup time and specialized resources, which can make frequent RTS assessments challenging in many clinical or field environments.

Markerless systems promise faster, more scalable biomechanical assessment, but only if they demonstrate strong test-retest reliability, meaning clinicians can trust that measured changes between sessions reflect meaningful changes in the movement. 

Interested in how reliability fits within a broader evaluation of sports technology? Explore our full breakdown of the STRN Sports Technology Framework and how it applies to markerless motion capture, including reliability, validity, usability, and data aesthetics.

This study directly addresses the key question clinicians, sports scientists, and performance staff have been asking: 

‍Can markerless motion capture reliably capture high-impact, multiplanar RTS tasks – across different days – well enough to track rehab progress?

‍

Study Overview

‍
Participants

  • 18 healthy adults (12 male, 6 female)
  • Two test sessions ~1 week apart

Tasks Assessed 

A comprehensive battery representing real RTS challenges:

  • Countermovement jump (CMJ)
  • Crossover hop (COH)
  • Drop vertical jump (DVJ)
  • Single-leg drop vertical jump (SLDVJ)
  • Single-leg hop
  • Sidestep cutting (SSC)
    ‍
    ‍

Measurement Setup
‍

  • Recorded with 8 Qualisys Miqus video cameras + dual force plates
  • Processed through Theia3D → Visual3D for joint angles & external moments
  • Evaluated sagittal & frontal planes (transverse excluded due to known unreliability)

‍

Analysis Metrics
‍

  • Integrated ICCs for full time-series reliability
  • Discrete ICCs for key events (initial contact, peak values)
  • SEM and MDC to determine whether real clinical change can be detected

‍

Key Findings
‍

1. Strong Reliability for Most Kinematic Measures
‍

Across all tasks and joints, joint angles showed moderate-to-excellent test-retest reliability:

  • Sagittal plane: ICC = 0.61–0.86
  • Frontal plane: ICC = 0.62–0.90

This indicates clinicians can confidently use markerless systems to evaluate joint flexion/extension and ab/adduction between sessions.
‍

2. Kinetics: Sagittal Plane Strong, Frontal Plan Mixed
‍

Peak joint moments showed:

  • Good-excellent reliability in sagittal-plane knee and ankle moments
  • Moderate-excellent reliability in sagittal-plane hip moments
  • Notably poorer reliability for frontal-plane hip and ankle moments, especially in tasks revolving rapid direction change (DVJ, SSC, SLH)

This is consistent with previous literature on both marker-based and markerless systems, reflecting the known difficulty of detecting subtle frontal-plane kinetic changes for these challenging joints.
‍

3. Minimal Session-to-Session Variation in Kinematics
‍

SEM values were small:

  • Hip flexion: < 9°
  • Knee flexion: ≤ 4.3°
  • Ankle flexion: < 3.5°

These values are comparable to (or better than) marker-based systems, and the majority across all movements and joints fell within the widely considered ‘reasonable’ range of 2-5 degrees.
‍

4. MDC Values Suggest good Sensitivity for Kinematics
‍

MDC ranges for joint angles aligned with clinically observed improvements after ACL rehab (e.g., 5–10° changes).

However, kinetic MDCs – especially hip/ankle – were sometimes larger than typical rehab-driven changes, meaning kinetics should be interpreted cautiously if looking for subtle week-to-week improvements. 
‍

5. Cutting, Single-Leg Tasks = Most Challenging
‍

More complex tasks (SLH, SSC) showed:

  • Slightly lower reliability
  • Higher sensitivity to small variations in movement execution 

This reflects real-world variability rather than system limitations alone.
‍
‍

What This Means for Clinicians, Researchers, and Performance Staff 
‍

For Sports Scientists / Performance Coaches
‍

  • Multi-session testing (e.g. weekly monitoring) is feasible without concern for major measurement drift.
  • Cutting and single-leg tasks remain highly reliable in kinematics, even if kinetics show variability.
  • Avoid overinterpreting small-force asymmetries in frontal plane mechanics.

For Researchers
‍

  • Provides foundational reference values for designing longitudinal or intervention studies in dynamic tasks.
  • Highlights the importance of plane-specific reliability when selecting outcome variables.
  • Supports growing evidence that markerless systems can replace marker-based setups for field-relevant RTS tasks.

Read the full article here.
‍
‍

Interested in learning more?
‍

Contact us today to learn how Theia3D can support your lab, clinic, or program.

‍

Recent Posts
How Northwestern University Cuts Data-Collection Time by One-Third Using Markerless Motion Capture
Athletic Movement Performance (AMP) Lab, directed by Dr. Yuki A. Sugimoto in the Department of Physical Therapy and Human Movement Sciences (PTHMS) at Northwestern University’s Feinberg School of Medicine, is advancing ACL-injury and chronic ankle instability (CAI) research by integrating Theia3D markerless motion capture to analyze high-risk movements – including jumping, landing, sprinting, and cutting – in adolescent and collegiate athletes. With Theia3D, the AMP Lab has streamlined key components of real-world biomechanics data collection, completing sessions approximately one-third faster than before. This increased efficiency allows the team to evaluate more athletes, run more trials, and accelerate investigations into lower-limb injury mechanisms that directly inform sport-performance and sports-medicine decision-making.
A Decade of Sports Validation: Which Motion Capture Tech Performs Best?
A 2025 Sensors review from Auburn University compares how optical, IMU, and markerless motion capture systems perform across real-world sports environments. The findings reveal that while optical systems remain industry standard, markerless motion capture is quickly closing the gap, offering unmatched scalability, flexibility, and ecological validity for multi-sport organizations.
Concurrent Assessment of Sprint Running Kinematics Using Marker-Based and Markerless Motion Capture
A July 2025 study published in the Journal of Sports Sciences evaluated the reliability of Theia3D’s markerless motion capture system for assessing lower-limb and trunk kinematics during sprint running. Researchers compared Theia3D against a marker-based optical reference system (Motion Analysis Corp. Raptor) across multiple sprint phases, including early acceleration, mid-acceleration, and top speed. Results showed good to excellent agreement between systems for most joints and movement phases, with some variability at top speed, particularly in pelvic and foot angles. The findings highlight both the promise and the challenges of measuring sprint mechanics with markerless motion capture in high-speed, real-world conditions.
In this blog
Summary