Quantitative assessment of the 30-second sit-to-stand test using computer vision technology for physical therapists.
Chin-Ying Hsu, Ying-Po Hsu, Tien-Lung Sun and 1 others
Smartphone video of the 30-second sit-to-stand test can distinguish normal from weak older adults by tracking how the body moves, not just counting repetitions.
In 32 adults followed over six months, movement-quality measures separated the groups with accuracy of 0.943 to 0.971.
Key findings
1The Weak group showed greater extremal centre-of-mass displacement, altered distributional properties, and lower temporal complexity (Complexity Index) than the Normal group, while the Normal group showed no significant change over six months.
2Machine-learning classifiers using only centre-of-mass-derived features (repetition count excluded as an input) achieved area under the curve of 0.993 to 0.997 and accuracy of 0.943 to 0.971 in separating Normal from Weak participants across both sessions.
3The Weak group exhibited reduced distal-joint entropy and relatively greater proximal involvement, consistent with a shift toward proximal compensation when distal control is insufficient.
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 32 of the 55 recruited adults completed both sessions; attrition was due to loss of contact, health issues, or inability to attend. The Normal and Weak labels came from the repetition count itself, so the model is predicting a threshold it was built around rather than an independent clinical outcome. No independent cohort has tested these results, and the authors explicitly call the performance preliminary. Body height and weight were not recorded, and the centre-of-mass trajectories were not scaled to individual body size. No laboratory-grade motion capture or force plate was used to verify the smartphone-based measurements.
Declared interests
Funded by Tao Yuan General Hospital, Ministry of Health and Welfare. No other funding or commercial conflicts are declared in the text.
The easy way to misread this
Do not read the 0.943 to 0.971 accuracy as evidence this is ready for clinical use. The sample was 32 adults, the Normal and Weak labels were defined by the repetition count, and no independent cohort has validated these results. The authors call the performance preliminary.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →