Among the many sleep tracking devices B1 provides some potentially interesting sleep data, especially sleep interruptions (see image below).
Those interruptions for most people are a trip to the restroom but in my case I wondered if they could be an alternate way to video for identifying my sleepwalking.
As the graph below show, B1 very rarely sees my sleepwalking events as a sleep interruptions (data in purple).
Overall only 5.4% of my sleepwalking events are seen as a sleep interruptions by B1. And for a BCSL at level 5 (the level at which I am going out of bed) the detection rate is 45%.
If I include the "no data" category (when the device temporarily stops recording my sleep, which could be from the software not recognizing a sleep interruption) the rates are 12.2% and 63% respectively.
Interestingly basis feature "toss and turn" does a slightly better job at detecting my sleepwalking, as this count is decreased by clonazepam (see graph below), which decreases my sleepwalking about 50%.
Also the overall the number of "tosses and turns" correlates with the number of sleepwalking events, their intensity and their complexity (see graphs below).
Unfortunately it has a high and variable false positive rate, as graph A shows (for example when 20 to 30 tosses and turns are measured only about 6 sleepwalking events happened), and as such is not a reliable measure of my sleepwalking.
No comments:
Post a Comment