Manual physical balance assistance of therapists during gait training of stroke survivors: characteristics and predicting the timing.
Juliet A M Haarman, Erik Maartens, Herman van der Kooij and 3 others
PMID 29197402WHAT IT FOUND
In four patients not used to build it, a lower-back sensor algorithm caught 81% of therapist balance assistance events, and 87% of its alerts were real.
It missed shoulder corrections. This is a small engineering step, not evidence that therapy changes.
Key findings
01When tested on four patients not used to build it, the lower-back acceleration algorithm caught 81% of therapist balance-assistance events, and 87% of its alerts were correct.
02Therapists provided most assistance at the side of the pelvis (78% of events), with roughly 80% of total force in the frontal plane, a median duration of 1.1 s, and a median impulse of 9.4 Ns over that plane.
03The median distance walked between balance-assistance events was 11.5 m, and the median number of events per measurement set was 3.
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
Only eight male stroke survivors and seven therapists were measured, so the force and timing numbers are descriptive and may not apply to women or to patients who cannot walk 10 m with supervision. The supplied methods and results measured therapist assistance and algorithm performance, not patient improvement in balance or walking. The algorithm was built on four patients and tested on four others, and individual detection scores varied from 67% to 100%. Some therapist assistance was given to the trunk or shoulder, where the hip sensors could not measure force, so force summaries are incomplete. The authors state that time synchronization between the force and motion systems had a small inaccuracy, and therapist reaction time could delay the assist relative to the acceleration peak. The authors describe the dataset as a first step, not fixed numbers, and say more subjects and tasks are needed.
Declared interests
The supplied text says the study was funded by ZonMw and does not list author conflicts of interest.
The easy way to misread this
Do not read this as evidence that manual balance assistance improves stroke recovery or that a robot can safely decide when to help. The study measured only eight male stroke survivors during supervised gait training, and the algorithm did not catch all therapist assists, especially when help was given to the shoulder or trunk.