A machine learning approach for the design optimization of a multiple magnetic and inertial sensors wearable system for the spine mobility assessment.
Dalia Y Domínguez-Jiménez, Adriana Martínez-Hernández, Gustavo Pacheco-Santiago and 3 others
PMID 39501297WHAT IT FOUND
A machine-learning method identified a smaller wearable sensor set that still classified trunk movements in patients with ankylosing spondylitis, but it did not test clinical diagnosis or patient outcomes.
Key findings
01The configuration using sensors 1, 5, 6, 8, 9, 10, 13, 14, and 15 had AUC values of 0.963 for movement 1, 0.944 for movement 2, and 0.852 for movement 3.
02The unconstrained optimization produced higher AUC values than the constrained optimization.
03The selected configurations had fewer redundant sensor pairs than using all fifteen sensors.
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 12 patients with ankylosing spondylitis were analyzed, and no healthy control group was included. The sample was 10 men and 2 women, and the authors note that female anatomy may produce different range-of-motion patterns. The analysis tested only anterior hip flexion, trunk lateral flexion, and trunk axial rotation, not the cervical movements used in the original system. Sensor sliding during axial trunk rotation produced noise, which the authors say can affect classification for that movement. The authors state that further research with more patients with ankylosing spondylitis and a healthy control group is needed to validate the system in clinical spine evaluation.
Declared interests
Funding was from DGAPA-PAPIIT UNAM and Secretaria de Educación, Ciencia, Tecnología e Innovación. No other conflicts of interest are declared.
The easy way to misread this
Do not conclude that this wearable system improves spine mobility assessment or patient care. The study tested sensor placement and movement classification in patients with ankylosing spondylitis, and it did not show that the reduced system improves diagnosis, treatment decisions, or patient outcomes.