Enhanced gait tracking measures for individuals with stroke using leg-worn inertial sensors.

Forty-one participants were recruited, but Table 1 includes 39 after two participants withdrew. The analysis used 116 sessions collected before October 1, 2023, not every visit from every participant. Some assessments were skipped when a physical therapist judged them unsafe or inappropriate. Participants used prescribed assistive devices, orthotics and therapist assistance as needed. Two sensor brands were used because one was discontinued, and they sampled at 500 Hz and 100 Hz. The normal gait template for dynamic time warping was built from healthy controls, not from the stroke participants. Data were imbalanced across time points, and limited subgroup sample sizes reduced statistical power. Some participants were absent after hospital discharge, so fewer events were recorded at later time points. The sensors were on the lower legs, while the mat measured plantar pressure under the foot, so shank movement may not match foot contact exactly. Swing time and single support time estimates were not reliable. The study tested algorithm measurement performance, not whether the algorithm changes clinical decisions or patient outcomes.

The supplied text names the National Institute on Disability, Independent Living, and Rehabilitation Research as the funder. It does not state author conflicts of interest.

Do not treat this as proof that leg-worn sensors can replace the instrumented mat. The enhanced algorithm still estimated swing time and single support time poorly, and the study tested measurement performance, not patient outcomes.