Decision trees for detection of activity intensity in youth with cerebral palsy
Trost, 2016
Description
Two count-based decision trees are presented for classifying activity intensity of children with cerebral palsy wearing an ActiGraph at the hip. The average accuracy was 76.7-82.9% for vertical axis counts and 75.6-84.2% for vector magnitude counts.
Development/Validation
Sample: 51 (27 males) children with cerebral palsy, mean age was 12.4±3.3 years, 12.4±3.3 years, and 12.7±3.1 years for Gross Motor Function Classification System (GMFCS) 1, 2, and 3, respectively
Setting: Not reported
Activities: Activities of daily living, self-paced walking, stationary
Criterion: Cosmed K4b2 (VO2)
Accelerometer(s): ActiGraph GT3X on right hip
Validation approach: Leave-one-out cross-validation
Instructions
Two decision trees were available- one for vertical axis counts/15-s and one for vector magnitude (square root of the sum of the squared counts in each axis) counts/15-s. Models also incorporate the Gross Motor Function Classification System (GMFCS) rating of impairment.
Vertical Axis (counts/15-s)
Vector Magnitude (counts/15-s)
Source Information
Trost, S. G., Fragala-Pinkham, M., Lennon, N., & O'Neil, M. E. (2016). Decision trees for detection of activity intensity in youth with cerebral palsy. Medicine and Science in Sports and Exercise, 48(5), 958. https://doi.org/10.1249/MSS.0000000000000842 Link to Paper
Corresponding author: Stewart Trost, s.trost@qut.edu.au
Contact
Kimberly Clevenger at accelerometerrepository@gmail.com