+- Apnea Board Forum - CPAP | Sleep Apnea (https://www.apneaboard.com/forums)
+-- Forum: Public Area (https://www.apneaboard.com/forums/Forum-Public-Area)
+--- Forum: Main Apnea Board Forum (https://www.apneaboard.com/forums/Forum-Main-Apnea-Board-Forum)
+--- Thread: Algorithm used to classify apneas (/Thread-Algorithm-used-to-classify-apneas)
Algorithm used to classify apneas - fredkahan - 11-13-2020
I am interested to understand how my ResMed (and I assume all xPAP machines) classifies from the information it gathers the nature of the breathing anomalies encountered. I assume that display software like OSCAR simply reports the conclusion rendered by the machine, rather than attempting deduction on its own. Am I correct in recalling that the machine imposes deliberate pressure oscillations as it encounters problematic readings as part of its analysis and subsequent corrective measures? I'm sure others have asked this question as well. Where can I read about this process? What additional information or analysis is provided by the proprietary ResScan app? Thankyou.
Classifying apneas: an additional query about the use of oximetry data - fredkahan - 11-13-2020
Further to my previous query, can oximetry data contribute to analysis. Does ResScan accept oximetry data, and make something of it?
RE: Algorithm used to classify apneas - Sleeprider - 11-13-2020
Fredkahan, I merged your two threads here. It's often better to keep related questions and therapy together rather than in separate threads.
You are correct that OSCAR merely reports the data provided by the machine. Your machine determines apnea events by several methods. Your machine measures your respiratory flow rate and analyzes if there is a partial reduction in flow (flow limitation or hypopnea) or a cessation of flow (apnea). It uses a "forced air oscillating technique" (FOT) which is a 1-cm fluctuation in pressure at 4 times per second to determine if the airway is closed (obstructive apnea) or open (central apnea). All of this data can be graphed and summarized.
Resscan is capable of reporting oximetry data from a Resmed pulse oximeter. It does not work with any other brands. Oscar allows other less expensive devices to be used.
You should request the clinical manual for your Resmed Airsense 10 Autoset from the forum. The clinical manual discusses all of these measurements and the machine response in detail. Use the CPAP Setup Manual link at the top of this page and follow the instructions to request a manual. https://www.apneaboard.com/adjust-cpap-pressure/change-cpap-pressure-settings-adjusting-your-machine-with-a-clinician-setup-manual
RE: Algorithm used to classify apneas - GuyScharf - 11-13-2020
I am interested to understand how my ResMed (and I assume all xPAP machines) classifies from the information it gathers the nature of the breathing anomalies encountered. I assume that display software like OSCAR simply reports the conclusion rendered by the machine, rather than attempting deduction on its own.
That is correct. OSCAR just reports what the machine said.
Am I correct in recalling that the machine imposes deliberate pressure oscillations as it encounters problematic readings as part of its analysis and subsequent corrective measures?
Yes, ResMed machines use a high frequency pulse to distinguish between an OA and a CA. I don't know that the pulse is related to the decision algorithm about adjusting pressure, but the apnea does play into that algorithm.
I'm sure others have asked this question as well. Where can I read about this process?
The AirSense 10 clinician manual does provide an introduction to apneas and how the CPAP machine reponds. But it is not very specific. I would imagine that the details of the algorithm are proprietary.
Also, you might want to look at ResMed's page: https://www.resmed.com/ap/en/consumer/support/devices/the-advantages-of-autoset.html
RE: Algorithm used to classify apneas - fredkahan - 11-13-2020
Thankyou, I really appreciate the depth of your response and will follow-up on the resources you describe.
As a newcomer to treatment, still in my compliance period, I'm frustrated by my enforced inability to experiment with masks and other tricks to overcome my still too high AHI. I wouldn't want a well intended experiment to fail and disqualify that night from the requisite 22/30 sessions in the compliance period!
My high AHI's are dominated by "unclassified apneas" occurring during major air leaks. My oximetry data never reflects an underlying problem, and I'm tempted to ignore them. I'm sure I'm wrong! Thus my interest in understanding "classification."
Thanks again.
RE: Algorithm used to classify apneas - GuyScharf - 11-13-2020
The machine calls an apnea "unclassified" when, because of high leak rate, it cannot determine whether the apnea is an OA or CA. I presume the high flow of air as the machine attempts to maintain pressure in the presence of large leaks results in the high frequency pulse not being detected reliably.
So I recommend that your immediate focus be to reduce the leak rate so the machine can properly classify any apneas.
RE: Algorithm used to classify apneas - multicast - 11-14-2020
Maybe of help: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4629962/
RE: Algorithm used to classify apneas - fredkahan - 11-14-2020
Terrific! Just what I wanted. Thank you so much.
RE: Algorithm used to classify apneas - happydreams - 11-14-2020
This is very helpful! It is one of the better discussions of how various PAP machines work and their basic algorithms. Now I understand better how my bi-level works and when some of the controls are indicated. Just added this article to my collection.