Automatic gait events detection with inertial measurement units: healthy subjects and moderate to severe impaired patients.
Cyril Voisard, Nicolas de l'Escalopier, Damien Ricard and 1 others
PMID 38890696WHAT IT FOUND
An algorithm using foot-worn sensors accurately detected gait events in healthy people and in patients with multiple sclerosis or post-stroke foot deformity.
It matched the gold standard without needing extra equipment, offering a practical tool for clinical gait analysis.
Key findings
01The new algorithm achieved an F1-score of 100% in healthy subjects, 99.4% in patients with multiple sclerosis, and 96.3% in patients with equino varus foot.
02The median absolute error for detecting toe-off was 8 ms in healthy subjects, 23 ms in multiple sclerosis patients, and 27 ms in patients with equino varus foot.
03The method outperformed a state-of-the-art algorithm in all groups, with significantly lower median errors for toe-off and heel-strike detection.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for OTs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study only analyzed straight-line walking, not turns or real-world environments. The gold standard (GAITRite) has limitations in detecting events in highly pathological gaits, such as toe-strike in foot drop. Sensors were placed on the foot, which may miss certain acceleration patterns compared to shank-mounted sensors. The study did not include all neurological conditions, focusing only on multiple sclerosis and post-stroke hemiplegia.
The easy way to misread this
Do not assume this algorithm works for all gait pathologies or in free-living environments. It was validated only on straight-line walking in a hospital setting, and its accuracy relies on the specific definitions of heel-strike and toe-off used here, which may differ from other clinical interpretations.