Rehabilitation Exergames: Use of Motion Sensing and Machine Learning to Quantify Exercise Performance in Healthy Volunteers.
Reza Haghighi Osgouei, David Soulsby, Fernando Bello
PMID 32808932WHAT IT FOUND
Two machine learning methods scored healthy volunteers' exercise movements against a physiotherapist's reference with similar trends.
However, the scores ignored speed and were heavily influenced by how researchers defined 'worst performance'. These metrics are not yet validated for patients.
Key findings
01The two algorithms produced similar overall trends when scoring movement similarity, with no significant difference between them.
02The scoring methods were sensitive to range of motion but ignored the speed of the exercise; two people moving at different speeds with the same range would get the same score.
03The final similarity scores depended heavily on the researchers' choice of 'worst performance' (defined as no movement) to set the scale boundaries.
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 was conducted entirely on healthy volunteers, not on patients with injuries or movement impairments. The 'worst performance' benchmark used to create the 0% score was an arbitrary choice by the researchers (no movement), which directly dictates the scale of all other scores. The scoring algorithms ignored the speed of the exercise, meaning they cannot distinguish between a controlled movement and a rushed one if the range of motion is identical. The study did not validate the similarity scores against any existing clinical assessment tools or patient outcomes.
Declared interests
None declared.
The easy way to misread this
Do not assume these similarity scores reflect clinical quality of movement. The study used healthy volunteers, and the algorithms ignored movement speed and were calibrated based on an arbitrary 'no movement' baseline, not on clinical standards of good technique.