Movealytics
Movealytics AI-powered motion analysis
Learn More

Research-driven motion analysis with a strong clinical foundation

Movealytics is a translational platform grounded in peer-reviewed research demonstrating the feasibility of markerless motion analysis using smartphone video.

Evidence supporting accessible markerless motion analysis

The evidence base combines external research on smartphone-based markerless motion analysis with validation studies from the Movealytics team spanning pose estimation, stride detection, automated scoring, upper-extremity assessment, and video quality screening.

External markerless motion analysis evidence

Published studies increasingly support the clinical usefulness of single-camera smartphone-based markerless motion analysis for selected gait measures and follow-up applications.

  • Good agreement has been reported for core spatiotemporal gait measures in several clinical populations.
  • Recent work suggests video-based systems can detect within-person change over time, supporting rehabilitation follow-up.
  • The evidence is strongest for gait; complex 3D biomechanics still require more specialized systems.
View references
  1. Stenum J, et al. (2021) Applications of pose estimation in human health and performance across the lifespan. Sensors. Nov 3;21(21):7315.
  2. Stenum J, et al. (2024) Clinical gait analysis using video-based pose estimation: Multiple perspectives, clinical populations, and measuring change. PLOS Digital Health. Mar 26;3(3):e0000467.
  3. Zoeller CS, et al. (2026) Video-Based Motion Capture Smartphone Apps for Testing Human Motor Performance Skills: Scoping Review. JMIR mHealth and uHealth. Feb 19;14(1):e65474.

Pose estimation validation

The Movealytics research program established the technical foundation for smartphone-based motion analysis by evaluating pose models and converting 2D keypoints into clinically useful events and segments.

  • OpenPose showed stronger performance than alternative pose engines in handheld clinical video comparisons.
  • Handheld smartphone gait video could be converted into movement-plane detection, foot-event detection, and stride segmentation.
  • These studies support ordinary 2D smartphone video as a practical basis for markerless clinical motion analysis.
View references
  1. Zhang F, et al. (2021) Comparison of OpenPose and HyperPose artificial intelligence models for analysis of hand-held smartphone videos. 2021 IEEE International Symposium on Medical Measurements and Applications (MeMeA) 2021 Jun 23, 1–6.
  2. Mroz S, et al. (2021) Comparing the Quality of Human Pose Estimation with BlazePose or OpenPose. IEEE 4th International Conference on Bio-Engineering for Smart Technologies (BioSMART 2021), December, Paris, pp. 33–36.
  3. Ramesh SH, et al. (2023) Automated Stride Detection from OpenPose Keypoints Using Handheld Smartphone Video. 2023 IEEE Sensors Applications Symposium (SAS), July, Ottawa.

Clinical scoring automation

Movealytics translates pose detection into clinically interpretable scoring tools for gait and upper-extremity function, with supporting validation and video-quality screening research.

  • Automated EVGS validation has shown strong performance for several gait parameters and robust view/direction detection.
  • A dedicated video-quality framework helps determine whether a recording is suitable for automated analysis.
  • Automated MMS scoring from smartphone video showed strong agreement with expert scoring in upper-extremity assessment.
View references
  1. Harini-Ramesh S, et al. (2023) Development and Validation of a Remote Video Acquisition and Analysis Protocol Using Artificial Intelligence (AI) for Evaluation of Gait. AACPDM 77th Annual Meeting. September, Chicago.
  2. Somasundaram I, et al. (2025) Automated Implementation of the Edinburgh Visual Gait Score (EVGS). Sensors, 25(10), 3226.
  3. Jeeva RA, et al. (2025) Automated Video Quality Assessment for the Edinburgh Visual Gait Score (EVGS). Methods and Protocols. 8(4), 71.
  4. Su C, et al. (2025) Automated Assessment of Upper Extremity Function with the Modified Mallet Score using Single Plane Smartphone Video. Sensors. 25(15), 1619.

Selected publications and presentations

Mroz et al., 2021 – IEEE BioSMART

Comparison of BlazePose and OpenPose pose estimation algorithms for movement analysis.

Zhang et al., 2021 – IEEE MeMeA

Comparison of OpenPose and HyperPose models for handheld smartphone video analysis.

Somasundaram et al., 2025 – Sensors

Improved automated EVGS implementation and validation on real patient videos.

Arumugam Jeeva et al., 2025 – Methods and Protocols

Automated video quality assessment algorithms for EVGS scoring reliability.

Built on a deep foundation of translational research

The Movealytics founding team collectively brings over 1,000 peer-reviewed publications and conference presentations, with extensive experience translating biomechanics research into clinical technologies.

Meet the Team

The team's research has been published in leading peer-reviewed journals and presented at international conferences in biomedical engineering, rehabilitation medicine, and clinical biomechanics — spanning pose estimation, automated clinical scoring, gait analysis, and markerless motion capture.