In-lab versus at-home activity recognition in ambulatory subjects with incomplete spinal cord injury.
Mark V Albert, Yohannes Azeze, Michael Courtois and 1 others
PMID 28166824WHAT IT FOUND
In incomplete spinal cord injury, a waist accelerometer recognized instructed home activities much better when trained on at-home movement data (85.9%) than when trained only on lab data (54.6%).
Key findings
01In-lab activity data were predicted with 91.6% accuracy (89.9–93.2).
02Classifiers trained on in-lab data recognized at-home activities with 54.6% accuracy (51.6–57.6).
03Training on at-home data improved at-home recognition to 85.9% accuracy (83.6–87.9).
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 included only 13 ambulatory subjects with incomplete spinal cord injury. Activities were instructed and performed in fixed sequences, not normal free-living movement. Labels were made by one researcher from accelerometer data and expected activity order, without video validation or test-retest reliability. The group had a wide range of mobility impairments, and joystick-controlled wheelchairs made wheeling hard to distinguish from sitting. The study tested algorithm classification accuracy, not whether activity tracking improves therapy or patient outcomes.
The easy way to misread this
Do not conclude that wearable activity tracking improves mobility outcomes in incomplete spinal cord injury. The study measured how well algorithms classified instructed activities, not whether tracking changed therapy or patient function.