Indirect measurement of anterior-posterior ground reaction forces using a minimal set of wearable inertial sensors: from healthy to hemiparetic walking.
Dheepak Arumukhom Revi, Andre M Alvarez, Conor J Walsh and 2 others
PMID 32600348WHAT IT FOUND
Three wearable sensors on the pelvis, thigh and shank, plus a person-specific model, closely estimated walking propulsion and braking forces compared with forceplates in 10 healthy and 5 post-stroke walkers.
The method still needs forceplate training.
Key findings
01The indirect measurement of the anterior-posterior ground reaction force time series strongly approximated forceplate measurements in healthy and post-stroke walkers, with error of 4.62%bw (percent of body weight) in healthy walkers and 2.64%bw in post-stroke walkers.
02Removing thigh IMU parameters substantially weakened reconstruction, increasing error by 155% in healthy walkers and 125% in post-stroke walkers.
03For the post-stroke paretic limb, estimated peak propulsion magnitude remained accurate in validation, with error of 1.07%bw and agreement of 0.98.
STILL TO COME
How it was doneWhat they foundWhat it means for PTs
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 included only 10 healthy individuals and 5 individuals with chronic post-stroke hemiparesis, so accuracy may not hold for larger or different groups. All post-stroke participants were more than 6 months after stroke and could walk without assistance, so it does not show performance in acute or more impaired walkers. Two healthy participants had right-limb data unusable because a sensor moved, so secure mounting and sensor drift matter. The model was subject-specific and required forceplate data for training, so it is not a standalone direct measurement. Comparisons were limited to strides with full forceplate contact and no turns, so accuracy on all steps was not directly validated. The approach may not suit people with higher walking variability or forces outside the range studied, such as acute post-stroke patients. The lower limb sensors had to be mounted so one plane matched the segment sagittal plane; incorrect placement could produce wrong angles.
Declared interests
No conflicts of interest or funding declaration appears in the supplied text. The journal metadata lists NIH extramural research support.
The easy way to misread this
Do not conclude that these sensors directly measure ground reaction forces or that they are ready for clinical use. The estimates were produced by a subject-specific model trained on forceplate data and tested in 10 healthy and 5 post-stroke walkers on a laboratory track.