Resources · Evidence

The science behind smartphone movement analysis

Objective measurement only earns trust if it agrees with the instruments clinicians already rely on. Here is the peer-reviewed evidence that markerless, video-based movement analysis can — and where Kinetically stands on its own validation.

Kinetically turns a smartphone camera into standardized movement metrics usingmarkerless, computer-vision pose estimation — the same family of methods now studied extensively in the clinical and biomechanics literature. Rather than ask you to take the approach on faith, this page summarizes what independent, peer-reviewed research has found about its accuracy and reliability.

What this evidence is — and isn't. The studies below validate thescientific method Kinetically is built on: markerless, video-based movement analysis. They are independent research, not evaluations of the Kinetically product. Kinetically's own head-to-head validation against reference standards is in preparation (see below) — until it is published, we do not attribute any specific accuracy figure to the product.

Peer-reviewed evidence

What the published research shows

A selection of independent, peer-reviewed studies on smartphone and markerless movement analysis — across the gait, rehabilitation, and neurology contexts Kinetically serves.

< 3°mean joint-angle error vs. laboratory motion capture, from a handheld smartphonePeiffer et al., 2025
Smartphone capture · biomechanics

A handheld smartphone paired with computer-vision pose estimation measured whole-body movement with mean joint-angle errors under 3° and pelvis-translation errors of a few centimeters versus laboratory motion capture — and tracked clinically meaningful change in prospective clinical use.

Peiffer JD, Shah K, Djuraskovic I, et al. (2025). Portable Biomechanics Laboratory: Clinically Accessible Movement Analysis from a Handheld Smartphone arXiv preprint 2507.08268. doi:10.48550/arXiv.2507.08268

Gait · chronic conditions

A systematic review of smartphone-based gait analysis in chronic health conditions reported good-to-very-good validity against established reference systems in clinical settings — with precision more variable in unsupervised home environments.

Bea T, Chaabene H, Freitag C, et al. (2025). Psychometric Characteristics of Smartphone-Based Gait Analyses in Chronic Health Conditions: A Systematic Review Journal of Functional Morphology and Kinesiology 10(2):133. doi:10.3390/jfmk10020133

Rehabilitation · clinical measurement

A systematic review in the Journal of NeuroEngineering and Rehabilitation found markerless motion capture is a practical and increasingly validated alternative to marker-based laboratories for clinical measurement, while noting accuracy varies by joint, plane, and task.

Lam WWT, Tang YM, Fong KNK (2023). A systematic review of the applications of markerless motion capture (MMC) technology for clinical measurement in rehabilitation Journal of NeuroEngineering and Rehabilitation 20:57. doi:10.1186/s12984-023-01186-9

Reliability · spatiotemporal gait

AI-driven markerless capture showed excellent within- and between-session reliability for spatiotemporal gait parameters and sagittal-plane kinematics in healthy adults; out-of-plane (frontal/transverse) measures were flagged for further refinement.

Schoenwether B, Ripic Z, Nienhuis M, et al. (2025). Reliability of artificial intelligence-driven markerless motion capture in gait analyses of healthy adults PLOS ONE 20(1):e0316119. doi:10.1371/journal.pone.0316119

Neurology · stroke recovery

In stroke survivors, AI markerless motion capture showed good-to-excellent validity for gait speed, stride length, and stance time versus reference measurement — supporting its use for key spatiotemporal metrics, with some parameters (e.g., stride width) interpreted with caution.

Alammari B, Schoenwether B, Ripic Z, et al. (2025). Validity of AI-Driven Markerless Motion Capture for Spatiotemporal Gait Analysis in Stroke Survivors Sensors 25(17):5315. doi:10.3390/s25175315

In practice

How Kinetically applies these methods

  1. 01

    Standardized capture

    Guided, repeatable protocols for recognized tests — Timed Up & Go, gait speed, sit-to-stand, balance, range of motion — so each capture is comparable over time.

  2. 02

    Dense landmark tracking

    1,200+ body landmarks tracked per frame from ordinary smartphone video — no wearables, mats, or force plates — the markerless approach the literature evaluates.

  3. 03

    Metrics mapped to standards

    Outputs map to recognized clinical measures (e.g., standard gait variables; MDS-UPDRS motor items where applicable) so results speak the language clinicians and payers already use.

  4. 04

    Trend, don't just snapshot

    The same standardized test is re-captured across visits and at home, so change is measured against a patient's own baseline — where objective tracking adds the most value.

Kinetically's product-specific validation is in preparation. We're finalizing the agreement, reliability, and limits-of-agreement analyses — with citations — that compare Kinetically's measurements directly against instrumented reference standards. If you're a clinician, researcher, or payer evaluating the platform and want our current validation materials, reach out and we'll share what's available under review.

Request our validation materials
FAQ

Common questions about the evidence

Is smartphone-based movement analysis clinically valid?

A growing body of peer-reviewed research finds that smartphone and markerless, video-based movement analysis can reach good-to-very-good agreement with laboratory reference systems for core gait and mobility metrics. Recent work reported mean joint-angle errors under 3° versus laboratory motion capture from an ordinary handheld smartphone. Accuracy varies by movement, plane, and capture conditions, so results are strongest for standardized, well-instrumented tests.

How accurate is markerless motion capture compared to a motion lab?

In 2025 research, handheld smartphone capture with computer-vision pose estimation produced mean joint-angle errors below 3° and pelvis-translation errors of a few centimeters against gold-standard marker-based motion capture. Spatiotemporal gait parameters (gait speed, stride length, stance time) tend to show good-to-excellent validity and reliability; out-of-sagittal-plane and width-based measures are less precise and are interpreted with more caution.

Does Kinetically have its own published validation studies?

Kinetically is built on the same peer-reviewed, computer-vision movement-analysis methods validated in the literature, and is preparing its own published validation materials comparing the platform against instrumented reference standards. Until those are published, Kinetically does not claim specific product accuracy figures; clinicians, researchers, and payers evaluating the platform can request the current validation materials directly.

Which movement metrics have the strongest evidence base?

The strongest evidence is for spatiotemporal gait metrics — gait speed, stride and step length, stance and swing time — and for sagittal-plane joint kinematics. These are exactly the standardized measures (e.g., Timed Up & Go, gait speed, sit-to-stand) that Kinetically captures and trends over time.

References

  1. Peiffer JD, Shah K, Djuraskovic I, et al. (2025). Portable Biomechanics Laboratory: Clinically Accessible Movement Analysis from a Handheld Smartphone. arXiv preprint 2507.08268. doi:10.48550/arXiv.2507.08268
  2. Bea T, Chaabene H, Freitag C, et al. (2025). Psychometric Characteristics of Smartphone-Based Gait Analyses in Chronic Health Conditions: A Systematic Review. Journal of Functional Morphology and Kinesiology 10(2):133. doi:10.3390/jfmk10020133
  3. Lam WWT, Tang YM, Fong KNK (2023). A systematic review of the applications of markerless motion capture (MMC) technology for clinical measurement in rehabilitation. Journal of NeuroEngineering and Rehabilitation 20:57. doi:10.1186/s12984-023-01186-9
  4. Schoenwether B, Ripic Z, Nienhuis M, et al. (2025). Reliability of artificial intelligence-driven markerless motion capture in gait analyses of healthy adults. PLOS ONE 20(1):e0316119. doi:10.1371/journal.pone.0316119
  5. Alammari B, Schoenwether B, Ripic Z, et al. (2025). Validity of AI-Driven Markerless Motion Capture for Spatiotemporal Gait Analysis in Stroke Survivors. Sensors 25(17):5315. doi:10.3390/s25175315

Citations are to independent, peer-reviewed research on smartphone and markerless movement analysis. They establish the validity of the method category; they are not evaluations of the Kinetically platform. Kinetically supports — and does not replace — clinical judgment.

Get started

See movement, measured.

Start free and turn a smartphone video into objective clinical metrics in minutes.

Start FreeExplore the platform