Can an ordinary video measure 10-meter walk speed?

Gait speed observation in a clinical setting

MedMeters literature review prepared 31 October 2026. The year in the article heading refers to the original study publication, not an earlier MedMeters posting date.

r = .92–.98correlation with video-analysis reference · findings from external research

Clinical question

How well can a deep-learning video pipeline estimate speed during a 10-meter walk in a real-world outdoor setting?

Study design

Thirty-two healthy participants across three age groups walked at slow, comfortable and fast speeds. A YOLOv11x-based system estimated segment speeds, which were compared against the Tracker video-analysis software.

What the researchers found

Correlations ranged from 0.92 to 0.98 across evaluated segments and speeds; Bland–Altman analyses indicated good agreement in the tested setup.

Why this matters for MedMeters

The study supports the concept of a calibrated video gait-speed protocol that can generate structured mobility measures without a dedicated gait laboratory.

What this does not prove

This was a small healthy cohort and a comparison against another video-analysis tool—not proof of clinical interchangeability with all gait measurement standards or accuracy in patients using walking aids.

Original peer-reviewed publication

Validation of a Deep Learning-Based Markerless Video System for Gait Speed Assessment During the 10-Meter Walk Test in Outdoor Clinical Settings
Kamnardsiri et al. · JFMK · 2026
DOI: 10.3390/jfmk11030357

External scientific evidence summary. Not medical advice. MedMeters is a concept under development and has not been evaluated in these studies. Clinical claims, validation and regulatory status must be established separately.