Study at a glance
Summary
Disclaimer: The following summary details independent academic research. Theia3D is intended for research, sports performance and educational use only. It is not cleared as a medical device and is not intended to be used for the diagnosis, treatment, mitigation or prevention of any disease, injury or medical condition.
Editor's note: Theia3D version 2022.1.0.2309, as stated in the paper.
Study Overview
Design
A single-group, repeated-measures reliability study assessing the inter-session reliability of lower-extremity joint kinematics and kinetics obtained using Theia3D markerless motion capture across a battery of six dynamic return-to-sport screening tasks.
Participants
Eighteen healthy participants (12 men, 6 women; age 24.8 years; height 174.1 cm; mass 70.8 kg), free of neuromuscular or musculoskeletal impairments and free of lower-limb trauma in the six months prior to testing.
Data Collection
Participants completed two testing sessions, six plus or minus two days apart, each involving a countermovement jump, a crossover hop, a bilateral drop vertical jump, a single-leg drop vertical jump, a single-leg hop, and a sidestep cutting maneuver, performed in random order with at least five successful trials of each task recorded per session. Only the non-dominant limb was assessed for single-leg tasks. Markerless kinematic data were recorded using eight Qualisys Miqus video cameras (180 Hz) processed through Theia3D, synchronized with two force platforms (2700 Hz), with all kinematic and kinetic modeling performed in Visual3D.
Analysis
Hip, knee, and ankle joint angles and moments in the sagittal and frontal planes were calculated for both discrete events (initial contact and peak values) and full movement-curve ("integrated") profiles. Inter-session reliability was assessed using intraclass correlation coefficients (ICC), standard error of measurement (SEM and SEM%), and minimal detectable change (MDC), which together indicate how consistent the measurements were between the two test sessions and how much change would be needed to be confident a true difference, rather than measurement noise, had occurred.
Key Findings
- The majority of hip, knee, and ankle joint angles and moments showed moderate-to-good inter-session reliability overall, with combined integrated reliability across kinematics and kinetics ranging from moderate to good depending on the specific task (ranging roughly from 0.64 to 0.79 across the six tasks).
- Joint angle reliability was slightly better in the frontal plane (average integrated reliability of 0.80) than in the sagittal plane (average of 0.73), a pattern the authors attribute to the frontal plane's more limited range of motion during these tasks rather than to any inherent measurement advantage.
- Sagittal-plane knee and ankle joint moments generally showed good-to-excellent reliability across most tasks, while frontal-plane hip and especially ankle joint moments showed considerably weaker and more variable reliability, with several task-and-joint combinations falling into the poor reliability category.
- Measurement error (SEM) for joint angles was generally low, under roughly 9 degrees for peak hip flexion and under about 4 degrees for peak knee and ankle flexion, broadly comparable to reliability levels previously reported for less demanding tasks like walking, though measurement error was notably higher for frontal-plane and kinetic parameters.
- The minimal detectable change values, particularly for joint moments, were often large relative to the magnitude of change typically expected following a training or recovery-focused intervention, leading the authors to recommend caution when using this system to track subtle, meaningful change in kinetic measures over time, and to identify frontal-plane hip and ankle joint moments specifically as the measures currently requiring the most caution.
What This Means for Movement Science Researchers and Return-to-Sport Screening and Athlete Monitoring Teams
For researchers and performance staff involved in return-to-sport screening, this study offers a specific, practically useful roadmap: markerless motion capture's session-to-session reliability was generally solid for joint angles and sagittal-plane joint moments across a demanding battery of jumping, hopping, and cutting tasks, supporting its use for the kind of repeated, longitudinal movement-quality screening that marker-based systems' setup demands often make impractical. The frontal-plane kinetic findings offer an equally concrete, actionable caution: teams relying specifically on frontal-plane hip or ankle moment measurements to track an athlete's progress over time, measures often of particular interest given their relevance to ACL-related loading mechanisms, should interpret session-to-session changes in these specific values with added caution given the relatively high measurement variability and minimal detectable change identified here. This kind of task-and-plane-specific reliability data gives practitioners a concrete basis for deciding which measures from a return-to-sport screening battery can be trusted to reflect real change, and which should be interpreted more conservatively.
The Role of Theia3D in Movement Analysis
Theia3D served as the markerless motion-capture tool directly assessed for its own session-to-session measurement reliability in this study, specifically across a battery of demanding jumping, hopping, and cutting tasks that had not previously been comprehensively evaluated for inter-session reliability using this system. The authors describe this as, to their knowledge, the first study to comprehensively assess markerless test-retest reliability across this specific set of return-to-sport screening tasks, extending prior reliability research that had focused mainly on walking and running.
Study Limitations and Future Directions
The authors note that their convenience sample was imbalanced by sex (12 men, 6 women) rather than deliberately sex-balanced, which may be relevant given that prior research has reported sex differences in lower-extremity kinematics and kinetics during dynamic tasks. Their sample consisted entirely of healthy participants with no recent trauma history, so these findings provide baseline reference data under standardized conditions but do not yet reflect how reliability might differ in individuals recovering from a specific lower-limb trauma, whose movement patterns and compensatory strategies may differ from those of healthy participants. The authors also note that markerless pose-estimation algorithms like Theia3D are still relatively young and under active development, so future software versions may further change test-retest reliability, particularly for the frontal-plane hip and ankle kinetic measures where marker-based motion capture currently still appears to outperform markerless data in terms of session-to-session consistency. The authors identify extending this kind of reliability assessment to populations recovering from a specific lower-limb trauma, and continuing to track how software updates affect reliability for the specific measures identified as weaker here, as important next steps.
Citation: Xu, J., Wen, Z., Augustine, S., Sharir, R., De Bleecker, C., Robinson, M. A., Jamkrajang, P., Smeets, A., & Vanrenterghem, J. (2025). Inter-session reliability of markerless lower extremity joint kinematics and kinetics for return-to-sport screening tasks. Journal of Sports Sciences. Advance online publication. https://doi.org/10.1080/02640414.2025.2541490
Link: Read the full study
How Reliable Is Markerless Motion Capture for Return-to-Sport Screening Tasks?
This study of 18 healthy participants tested across two sessions finds that Theia3D markerless motion capture shows moderate-to-good inter-session reliability for most lower-limb joint angles and sagittal-plane joint moments during jumping, hopping, and cutting tasks, with weaker reliability specifically for frontal-plane hip and ankle moments. If your work involves return-to-sport screening or athlete monitoring research, talk to our team about applying Theia3D to your own screening protocols.