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 v2023.1.0.3161, as stated in the paper.
Study Overview
Design
A within-subject, cross-sectional study comparing lower-limb kinematics, participant and caregiver perceptions, and assessment duration between two clothing conditions during a markerless gait assessment in typically developing children: a "Conventional" condition designed to mimic the clothing requirements of a traditional marker-based assessment, and a "Street" condition using participants' own, unconstrained clothing.
Participants
Thirty typically developing children (15 boys, 15 girls; mean age 8.6 years; mean height 136 cm; mean mass 34.3 kg), screened to exclude any self-reported lower-body musculoskeletal or neurological condition that could affect typical gait.
Data Collection
Each child completed eight passes of a 4.2 × 10 m raised walkway at a self-selected pace (four in each direction) under each of the two clothing conditions, in a block-randomized order, during a single visit. The Conventional condition used fitted shorts with the waistband rolled down and a t-shirt taped to expose landmarks on the pelvis and spine, with reflective markers placed but not used in the kinematic calculations; the Street condition used whatever clothing the child arrived in, with no restrictions. Eight synchronized Sony DSC-RX0M2 cameras recorded each trial at 119.8 Hz and were processed through Theia3D to generate hip, knee, and ankle joint angles in all three planes of motion. Participants and caregivers each completed a survey after the assessments to report their comfort and preference between conditions, and the duration of each condition was recorded.
Analysis
Subject-specific root-mean-square deviations (RMSD) between the Conventional and Street clothing conditions were calculated for each joint and plane, with an interquartile-range-based outlier analysis used to flag participants whose clothing may have meaningfully affected their kinematic data. Survey responses were tallied descriptively, and a Wilcoxon signed-rank test compared assessment duration between the two conditions.
Key Findings
- Kinematic differences between the Conventional and Street clothing conditions were small across the board, with median RMSDs below 4 degrees at every joint and plane of motion tested, and no consistent pattern indicating that one particular plane or joint was systematically more sensitive to clothing condition than another.
- Eighteen of thirty participants (60%) showed no statistical outlier in any joint or plane between conditions, while twelve (40%) had an outlier in at least one joint, most commonly associated with specific clothing characteristics: shirts extending below the hips, and baggy or loose pants or shirts that extended past or obscured the ankle joint centre.
- Clothing characteristics associated with outliers did not reliably predict them: baggy clothing resulted in a kinematic outlier only 31% of the time, and loose clothing only 27% of the time, suggesting clothing fit alone is not a fully deterministic driver of kinematic quality.
- Participants reported feeling notably happier during the Street condition than the Conventional condition. No participants reported feeling unhappy during the Street condition, compared with one participant (4%) during the Conventional condition, and 95% reported feeling happy or happiest during the Street condition compared with 85% during the Conventional condition.
- The Street condition was both strongly preferred and substantially faster to complete. Over 80% of participants were willing to repeat the Street condition compared with fewer than one-third for the Conventional condition, and the Street condition took a median of 3 minutes to complete compared with 14 minutes for the Conventional condition, a significant median difference of 11 minutes (p < 0.001).
What This Means for Movement Science Researchers and Care Teams
For teams running pediatric gait assessment services, this study offers concrete, evidence-based guidance rather than a general endorsement of "any clothing works": specific characteristics, loose or baggy shirts and pants, shirts extending below the hips, and pants obscuring the ankle, were the clothing features most associated with kinematic outliers, and the study's accompanying visual guide gives practical recommendations for what to wear and what to avoid. Beyond the kinematics, the clear preference for, and speed of, the unconstrained Street condition is a meaningful finding on its own: for a population where the burden of a traditional assessment can itself be a barrier to access, a markerless approach that doesn't require specific clothing or marker placement may make gait assessment more tolerable and more broadly accessible, without a meaningful tradeoff in lower-limb kinematic quality.
The Role of Theia3D in Movement Analysis
Theia3D was the markerless motion capture system used throughout this study, generating hip, knee, and ankle kinematics from eight synchronized cameras under two different clothing conditions in the same group of children. Because Theia3D requires no markers or specific clothing to operate, this study was able to directly isolate the effect of clothing choice itself on kinematic quality, a comparison that would not have been possible using a traditional marker-based system, while also demonstrating that removing clothing and marker constraints meaningfully improved the children's and caregivers' experience of the assessment.
Study Limitations and Future Directions
The authors note that fourteen of the thirty participants wore the same shirt in both clothing conditions, which may have reduced the overall diversity of clothing styles tested. Kinematic analysis was limited to the pelvis and lower body; upper-body kinematics were not assessed, so the study cannot speak to how clothing choice might affect upper-extremity data. Many participants chose to wear socks during testing, and neither condition allowed shoes, even though gait assessments in applied settings are often conducted with a child's own footwear or orthotic devices, which the authors note may limit how directly these findings generalize to assessments conducted with shoes on. Survey response patterns were also affected by participant age, with 46% of respondents under age 8, and some partial or non-responses, which the authors note may have introduced some uncertainty into the perception-survey findings. The convenience sample recruited for this study may be inherently biased, since many participants and families had previously taken part in related research or were already familiar with the assessment environment. The authors identify building on these clothing recommendations with further usability research specifically in children with mobility-related differences as a clear next step.
Citation: Rande, A., Sharma, M., Robu, I., Kuntze, G., Ursulak, G., Outerleys, J., Laende, E. K., Manocha, R. H. K., Pathy, R., & Condliffe, E. G. (2026). What not to wear: Examining the usability of markerless motion capture for pediatric populations. Journal of Biomechanics, 203, 113349. https://doi.org/10.1016/j.jbiomech.2026.113349
Link: Read the full study
Can Children Wear Their Own Clothing to a Markerless Gait Assessment?
This study shows that letting children wear their own, unrestricted clothing during a Theia3D gait assessment produced kinematic differences under 4 degrees compared with a conventional marker-based-style setup, while being faster and clearly preferred by participants and caregivers alike. If your work involves making pediatric gait assessment more accessible and you want practical, evidence-based guidance on clothing for a markerless setup, talk to our team about what this study could mean for your assessment workflow.