Automated Detection of Repetitive Motor Behaviors as an Outcome Measurement in Intellectual and Developmental Disabilities.
Kristin H Gilchrist, Meghan Hegarty-Craver, Robert B Christian and 3 others
PMID 29164446WHAT IT FOUND
In 15 usable participants, sensors caught 80% of rocking events, with 1.5 false alarms per hour.
Hand-flapping detection produced many false alarms when limited to flapping. This is a measurement tool.
Key findings
01The rocking algorithm detected 80% of rocking events across five participants who rocked, while false positives were 17 over 11 hours of recording for all 15 usable participants.
02The shirt collar sensor produced 19 false positives over 11 hours for all 15 usable participants, and overall detection was only slightly reduced compared with the chest sensor.
03Hand flapping detection produced many false positives when limited to flapping, but counting any repetitive hand stereotypy reduced false positives to 17 over 11 hours for all 15 usable participants.
STILL TO COME
How it was doneWhat they found
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 15 of 20 enrolled participants had usable video, and only five rocked and four flapped. The session was in a clinic, and activities were chosen to evoke repetitive behaviors, not to observe everyday behavior. The algorithms need testing in more real-life activities and settings. Hand flapping detection produced many false positives when limited to flapping, and it could not separate flapping from other repetitive hand stereotypy. Video timing for short flapping events was too imprecise to compare time spent flapping. All participants were male.
Declared interests
The authors declared no conflicts of interest.
The easy way to misread this
Do not conclude that these sensors are ready for clinical use. The study tested detection in 15 usable participants, only 5 rocked and 4 flapped, and used clinic activities chosen to provoke behaviors.