How do I know when I'm in REM sleep?

I see everyone talking about what goes on with their OSA when they are in REM sleep. How do I know by looking at the charts whaqt I am in REM sleep
Thanks in advance.

You can't with any certainty. Anything else is just a guess unless you are hooked up to an EEG machine.

I have sufficient information (brain waves, pulse, oxygen, breathing) to deduce it with reasonable accuracy, better than any one of those channels of information, but not with certainty.

Without EEG info, and someone who knows how to read that data, you can't really tell with certainty when you're in REM sleep. However, REM speep occurs in periods of 20-40 minutes 4 or 5 times each night and you can infer REM sleep from looking at your flow rate data. During REM sleep your breathing rate and flow rate data will be very irregular. If you have a recording pulse oximeter you can also see REM sleep as periods during which your SPO2 level will fluctuate rapidly.

Slartybartfast wrote:Without EEG info, and someone who knows how to read that data, you can't really tell with certainty when you're in REM sleep. However, REM speep occurs in periods of 20-40 minutes 4 or 5 times each night and you can infer REM sleep from looking at your flow rate data. During REM sleep your breathing rate and flow rate data will be very irregular. If you have a recording pulse oximeter you can also see REM sleep as periods during which your SPO2 level will fluctuate rapidly.

Anybody else have views in which they're confident as to how to infer from the detailed data charts, oximetry, etc., what portions of their sleep graphs are probably REM periods?

Slartybartfastand, MaxDarkside and others have now sufficiently encouraged me; so my order is in now for a pulse oximeter. For those interested, $95 will get you a Contec Oximeter Model CMS-50F, described at http://www.door2doorshop.com/wholesale/ ... 0F-36.html , online from http://www.door2doorshop.com as other members have noted.

Max, do you have anything to add re. how you interpret respiratory rate, flow rate and oximetry data? Or any of the other detailed graphs produced by ResScan or SleepyHead software to help identify when you think you've had REM sleep periods?

During my sleep studies, the sensors near the outside corners of my eyes were supposed to detect rapid eye movement. (REM)

Papit wrote:Max, do you have anything to add re. how you interpret respiratory rate, flow rate and oximetry data? Or any of the other detailed graphs produced by ResScan or SleepyHead software to help identify when you think you've had REM sleep periods?

I don't have a lot of time just at the moment, explaining takes some time, but have a look at what I review each day. #1 is my ResMed data, #2 is the pulse-ox, #3 is the Zeo brainwaves from the serial port hack (there is a LOT of info in those squiggles) and #4 is the Zeo classification of sleep stages, which is quite wrong at times, but if you use it as a guide, and look up at the brain waves and then the pulse ox and then at the respiration, you can see "the story" of what is going on (at least I can pretty clearly). The differentiation between Light/REM/Wake is not always easy, because you could be sliding into REM from Light when you Wake and it becomes a bit of a jumble. Most of my "Wakes" are during "REM", but "Wake" is also tipped off in the other series as a strong HR jump, or a reset on the ResMed, or ... I will be coming up with a mechanism using our software (powerful stuff) to count wakes using such anomalies in the data, perhaps correlated with what you don't see... IR camera to confirm.

All charts come from our software, except the Zeo bar chart at the bottom. I have it all time synchronized, selectable by any date/time range, any variable, stored in a real-time database, with a whole suite of analytical tools that can be used from clustering to predictive analytics, to anomaly detection to causal analytics, to performance optimization (my personal performance).

I don't have a lot of time just at the moment, explaining takes some time, but have a look at what I review each day. #1 is my ResMed data, #2 is the pulse-ox, #3 is the Zeo brainwaves from the serial port hack (there is a LOT of info in those squiggles) and #4 is the Zeo classification of sleep stages, which is quite wrong at times, but if you use it as a guide, and look up at the brain waves and then the pulse ox and then at the respiration, you can see "the story" of what is going on (at least I can pretty clearly). The differentiation between Light/REM/Wake is not always easy, because you could be sliding into REM from Light when you Wake and it becomes a bit of a jumble. Most of my "Wakes" are during "REM", but "Wake" is also tipped off in the other series as a strong HR jump, or a reset on the ResMed, or ... I will be coming up with a mechanism using our software (powerful stuff) to count wakes using such anomalies in the data, perhaps correlated with what you don't see... IR camera to confirm. . . .

Thanks, Max. That's a great looking set of chart from more than one source that you've been able to sync together on the time scale axis. Kudos. I see you recently discussed the Zeo sourced brainwave graphs on your blog, so I'll go there some more to learn how you make use of that data re. when REM might be occurring.

When you have some time, sure would like to get 'walked through' how you use #1, those ResMed detailed graphs, and #2, the pulse oximetry graphs that you included in your post for helping you determine when REM might be occurring.

Also, I assume you developed that pie chart in the upper right corner that's superimposed on the Item #1 ResMed charts. How are the values of the 'pieces' of the pie calculated?

I will write up more on the REM and other sleep stages stuff at the blog. The pie chart comes from Zeo's web site, based on the 5 second data the Zeo produces.

Zeo's web site, based on the 5 second data the Zeo produces.

Shouldn't that be 5 minute data? There is also another firmware revision and Java JAR file that allows for decoding 30 second epoch data. I use them in my own analysis. More informative than the 5 minute data and easier to use than the real time data. Initially I wondered about an odd breathing pattern of increasing and decreasing intensity. Once I got the Zeo and decoded it's data I came to realize that during REM I often exhibit Periodic Breathing.

http://en.wikipedia.org/wiki/Sleep

The above link has a little hypnogram showing normal cycles of sleep stages, just below the little kid picture.
I am documented (per sleep study) much, much worse in regards to my OSA in REM sleep.
53 AHI in REM sleep vs 12 AHI in non REM sleep and sleep position was not a factor.
I can pretty much eyeball my cpap reports and spot what probably are REM stage events because they pretty much line up with the time frames shown on that hypnogram. Never have much of anything show up till about 90 minutes after sleep onset and then a few events happen and always in the wee hours of the morning is when I might have the most activity.

Is it perfect eyeballing? No, but it is sufficient for my needs. I see no need to strap on another gizmo to my head. I figure as long as the events are properly treated I don't care when they occurred because I can't control REM anyway.

Now of course if your sleep stage cycles aren't normal for some reason, eyeballing probably isn't going to be much help but in the absence of any other gizmo to hopefully aid with more specific sleep stage information.....you can get a close approximation as to REM stages in your cpap reports by comparing normal sleep cycle times with your reports. It's not perfect but then...other than EEG...nothing is.