“`html
Study Summary
Researchers developed and tested a machine learning algorithm using smartwatches to estimate physical activity intensity. The algorithm categorized activity levels into sedentary, light, moderate, and vigorous based on metabolic equivalent (MET) values. The study involved 24 adults and 18 children.
Key Findings
- For adults, the algorithm showed high accuracy in classifying sedentary, moderate, and vigorous activities, with an area under the ROC curve (AUC) of 0.96, 0.88, and 0.86 respectively.
- The mean absolute error (MAE) between true and estimated METs was 0.75, indicating a strong estimation accuracy.
- For children, the algorithm demonstrated comparable accuracy, with AUC values of 0.98, 0.89, and 0.85 for sedentary, moderate, and vigorous activities respectively.
- The correlation coefficient and intraclass correlation (ICC) also showed high agreement between true and estimated METs in both adults and children.
Practical Implications
The developed model provides a highly accurate way to estimate physical activity intensity in both adults and children using smartwatches. This has potential applications in health monitoring and areas such as myopia prevention and control.
“`
