A single lumbar IMU paired with a nonlinear analytics pipeline can detect gait reorganization during intensive stroke rehabilitation imperceptible to standard clinical measures, in a proof-of-concept case series of three people with chronic stroke.
All three participants showed a 52-62% increase in Wolf's Lyapunov exponent during Stroke Boot Camp (SBC) Week 2, the most consistent nonlinear change across participants.
Entropy responses were heterogeneous across participants, but SampEn and RCMSE showed greater longitudinal sensitivity than conventional spatiotemporal metrics.
The nonlinear gait changes were detected by the IMU during intensive rehabilitation and were described as imperceptible to standard clinical measures.
A single lower-back IMU enabled continuous multi-day free-living gait monitoring across five sessions from one week pre-intervention to three months post-intervention.
A validated Python-based pipeline was used to process free-living data into linear and nonlinear gait biomarkers.
The authors position the work as a reproducible methodological foundation for larger controlled trials.
This small pilot study followed three people with long-term effects of stroke through a two-week intensive rehabilitation program. Each wore a small motion sensor on the lower back during everyday life, from a week before the program until three months after. The researchers analyzed not only standard walking measures such as speed and step timing but also measures of how complex and variable the walking pattern is over time. All three participants showed a 52-62% increase in one measure of walking dynamics (the Wolf Lyapunov exponent) during the second week of the program. Other complexity measures, such as entropy, changed differently from person to person, but they tended to pick up changes over time better than conventional walking measures. Standard clinical tests, given at each session, did not reveal these changes. This research suggests that a single wearable sensor worn at home could track subtle changes in how people walk during stroke rehabilitation, beyond what clinic tests show. However, with only three participants and no control group, the findings are preliminary. They mainly show that the approach is feasible and justify larger controlled studies to understand what the changes mean for recovery.
Soangra R, Duff S, McKenzie A. (2026). In-Home Free-Living Gait Monitoring Using a Single Lumbar IMU Reveals Nonlinear Complexity Changes During Intensive Stroke Rehabilitation.. Sensors (Basel, Switzerland). https://doi.org/10.3390/s26206467