Anyone tried to use AI to detect RERAs?

[tan]

Is that a RERA on the screen?



Rather than tediously scrolling through hours of flow rate data, maybe it's now possible to teach a machine to analyze data and flag RERAs? I am sure I am not the first one with such an idea.

[Dog Slobber]

The Resmed Airsense 10 AutoSet and AutoSet for Her, (perhaps other models) do *attempt* to flag RERAs.

I never found the RERA flagging to be accurate, reliable or beneficial. I've since moved on to an AirCurve, and don't miss not having the RERA flagging.

[palerider]

Is that a RERA on the screen?



Rather than tediously scrolling through hours of flow rate data, maybe it's now possible to teach a machine to analyze data and flag RERAs? I am sure I am not the first one with such an idea.

Since you stopped breathing, it's an apnea. But it doesn't *count* because it wasn't at least 10 seconds.

[Geer1]

Some of the breathes before pause look a little suspect then afterwords they look more sleep like. I am thinking the pause is sleep transition (if over 10 seconds would be called a sleep transition apnea).

The arousal either happened before the data you posted or possibly right before the cessation. Not enough data to know if it was a breathing related arousal or just spontaneous arousal. You can’t really tell without eeg (unless breathing gets obviously flow limited etc for 10 seconds prior to arousal then it is a safer guess of being a rera.

[chunkyfrog]

Can you hiccup in your sleep?

[tan]

Can you hiccup in your sleep?

Maybe, question is whether the increased flow, after a short pause, is an indication of sleep-disordered breathing, or is it something to be ignored, or it's still arousal but not breath-related, but I guess it's impossible to tell without EEG.

Thanks, everyone, for chiming in.

[greatunclebill]

sleep transition apnea

LOL... good one.

[djams]

The Resmed Airsense 10 AutoSet and AutoSet for Her, (perhaps other models) do *attempt* to flag RERAs.

I never found the RERA flagging to be accurate, reliable or beneficial. I've since moved on to an AirCurve, and don't miss not having the RERA flagging.

Interesting. RERA's were a major component of my SDB, and I ignored then for quite a while. If they meant anything, they'd show up on the machine's sleep report, I thought. Then I watched the sleep disordered breathing video that palerider posts pretty frequently and it prompted me to take a look.

Here is an example of what I'd see regularly while getting my settings tweaked. Three minutes of struggling to breathe, with a big arousal at the end. Looks very much like a hypopnea. I speculate that it didn't meet the criteria for flow reduction. But not sure. I'd have 10-20 of these a night. As you point out, not all of them would look as legit as this one. But OA's+H's+RERA's is what made me accept the advice here to keep raising min pressure.

So I wouldn't tell folks to ignore RERA's, were very real in my case.

RERA.png (17.01 KiB) Viewed 19834 times

[palerider]

The Resmed Airsense 10 AutoSet and AutoSet for Her, (perhaps other models) do *attempt* to flag RERAs.

I never found the RERA flagging to be accurate, reliable or beneficial. I've since moved on to an AirCurve, and don't miss not having the RERA flagging.

Interesting. RERA's were a major component of my SDB, and I ignored then for quite a while. If they meant anything, they'd show up on the machine's sleep report, I thought. Then I watched the sleep disordered breathing video that palerider posts pretty frequently and it prompted me to take a look.

Here is an example of what I'd see regularly while getting my settings tweaked. Three minutes of struggling to breathe, with a big arousal at the end. Looks very much like a hypopnea. I speculate that it didn't meet the criteria for flow reduction. But not sure. I'd have 10-20 of these a night. As you point out, not all of them would look as legit as this one. But OA's+H's+RERA's is what made me accept the advice here to keep raising min pressure.

So I wouldn't tell folks to ignore RERA's, were very real in my case.

What I see there is increasing flow limitation (the chair shapes on the inhalations) then an arousal, recovery breaths, possibly some moving around, and more recovery breaths.

[Geer1]

So I wouldn't tell folks to ignore RERA's, were very real in my case.
RERA.png

I don't believe Dog Slobber was saying that you should ignore RERA's, just that you should take the Autoset flagged RERA events with a grain of salt and that you need to review your data to see if what is flagged as a RERA is actually a RERA (like in your case).

I have now had four RERA's flagged by my autoset and three of them are questionable (like the following example). I believe there have been many other more obvious RERA's not flagged by the machine as well, they simply can't flag RERA's accurately as it requires EEG data which is not available on these machines.

RERA.PNG (14.02 KiB) Viewed 19819 times

[palerider]

flag RERA's accurately as it requires EEG data which is not available on these machines.

Thank you, captain oblivious.

[Geer1]

What I see there is increasing flow limitation (the chair shapes on the inhalations) then an arousal, recovery breaths, possibly some moving around, and more recovery breaths.

Thank you, captain oblivious.