How does a CPAP know the difference between CSA and OSA?

[Dutch]

On a different apnea forum we have a discussion about how a CPAP detects the difference between a central apnea, and an obstructive apnea.
A long time ago I picked up somewhere that a Resmed could detect the patients heartbeat.
Is a heartbeat detected, then it would be a central, is there no heartbeat, because the airway is blocked, then it would be an obstructive apnea.
I tried to find that source again. But it's years ago I red about it.Probably it doesn't exist anymore.
The only thing I found is Respironics' CPAP trying to detect a central with little puffs of pressure.

Please could anybody give me a clue?

[PsychoMike]

Hi Dutch,

Not quite about the heartbeat

The ResMed machines use air pulses to see if the airway is open. It sends out it's pulses and looks at what happens to the pressure. With OSA, the airway is blocked so the pressure will rise much more (smaller volume) with each pulse than if the airway is open (CSA event with a much larger volume that includes the airway and lungs).

We're still only talking little puffs (not like it's trying to blow you up like a balloon), but the pulses are large enough and algorithm is sensitive enough to figure out whether it is OSA related or CSA related (so it can boost the pressure on the OSA ones and not the CSA ones)

[Dutch]

Hi PsychoMike.

Thanks for your answer.
Sounds plausible.
But still I remember a reaction from a YouTuber on that review of a Resmed machine, who called it a 'jaw dropping technology'.
But zero hits on Google also... I must have seen it in my dreams.

[justMongo]

Hi Dutch,

Not quite about the heartbeat

The ResMed machines use air pulses to see if the airway is open. It sends out it's pulses and looks at what happens to the pressure. With OSA, the airway is blocked so the pressure will rise much more (smaller volume) with each pulse than if the airway is open (CSA event with a much larger volume that includes the airway and lungs).

We're still only talking little puffs (not like it's trying to blow you up like a balloon), but the pulses are large enough and algorithm is sensitive enough to figure out whether it is OSA related or CSA related (so it can boost the pressure on the OSA ones and not the CSA ones)

Close enough for government work.

After 4 seconds of no flow, it modulates the pressure with a 1cm-H2O sine wave and listens to the reflection. It's more of a sonar like principle.

In the highly detailed data on ResScan, you can actually see the pressure modulation on the pressure graph during an apnea event.

It's really quite brilliant.

[zonk]

Phillips Respironcis: Detection accuracy of obstructed airway versus clear airway apneas
http://www.healthcare.philips.com/main/h...apneas.wpd

ResMed: Central Sleep Apnea Detection and the Enhanced AutoSet Algorithm
http://www.resmed.com/au/assets/document...-paper.pdf

[SnuffySleeper]

Phillips Respironcis: Detection accuracy of obstructed airway versus clear airway apneas
http://www.healthcare.philips.com/main/h...apneas.wpd

ResMed: Central Sleep Apnea Detection and the Enhanced AutoSet Algorithm
http://www.resmed.com/au/assets/document...-paper.pdf

Holy moly, that's a hidden gem. Thank you sir!

[SleepWrangler]

The only thing I found is Respironics' CPAP trying to detect a central with little puffs of pressure.

I've been researching central apnea too. SleepyHead software is careful to use the term Clear Airway (CA) event and not the term Central Apnea event. Both have the same acronym unfortunately. I may be confused but truly detecting Central Apnea requires more sensors than our machines have available.

[SnuffySleeper]

The only thing I found is Respironics' CPAP trying to detect a central with little puffs of pressure.

I've been researching central apnea too. SleepyHead software is careful to use the term Clear Airway (CA) event and not the term Central Apnea event. Both have the same acronym unfortunately. I may be confused but truly detecting Central Apnea requires more sensors than our machines have available.

Yes, Sleepyhead/Resscan report a lot of CA's that should be hypopneas and so on and so forth. The cpap machines have a hard time indicating exactly what kind of event it is because it only has wav forms and using air resistance as the main reporting trigger. Still it gives you a good idea of whats going on, but to truly get detailed you need a sleep study and being hooked up to all those electrodes with a camera on you...

One of the many things I learned my last doctors visit after i expressed to my doctor the amount of CA's I had and he said it's probably just the machine saying they are CA events when they should be hypopneas and after we compared the wav forms of my machine to the sleep study I am sure my Doctor is right. A higher pressure on the cpap machine also gives a more accurate event as it has more air resistance to work with so the algorithm can detect the specific event. Me with my low pressure of 8, I get a lot of CA events.

[SleepWrangler]

Me with my low pressure of 8, I get a lot of CA events.

Ahhh. It appears I have very similar results. Did the detailed data from your machine change your doctor's opinion of your prescription? What is your conclusion?

[SnuffySleeper]

Ahhh. It appears I have very similar results. Did the detailed data from your machine change your doctor's opinion of your prescription? What is your conclusion?

My doctor was more concerned with the overall AHI of the machine data than the actual details of daily events. I think to him if the machine reported a CA, hypopnea, obstructive..etc it didn't matter. The machine just recognized "something" happened and to get a better detail he relied on the sleep study stats. I had 48 AHI on my sleep study with the majority being hypopneas and 1.89 being central apneas.

I did my own personal experiment (would not recommend this) up the pressure and took a long nap. After studying the data all my CA's turned into hypopneas in Sleepyhead giving a more aligned result with my actual sleep study results. After reading a lot on the subject it's because of the added air resistance the machine has to read.

My personal opinion is that the Cpap machines are great for a general overview, but it's not great at detecting specific event details but it tries it's best (which is better than nothing). The main point is that "something" happened and if it becomes a trend we should let the doctor know to get a sleep study done.

With me, my doctor wasn't really concerned with the CA events the machine was reporting, but I am going for another sleep study in two weeks to see if my pressure is optimal, so we'll see how that goes.